<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
<channel>
  <title>Gerris guides</title>
  <link>https://gerriscorp.com/guides/</link>
  <description>Technical SEO, AI search, migration, email, and privacy guides by Chris Abraham</description>
  <language>en-us</language>
  <lastBuildDate>Sat, 14 Nov 2026 12:00:00 +0000</lastBuildDate>
  <atom:link href="https://gerriscorp.com/rss.xml" rel="self" type="application/rss+xml"/>
  <item>
    <title>Why reputation management only works when the behavior changes</title>
    <link>https://gerriscorp.com/guides/reputation-management-requires-change/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/reputation-management-requires-change/</guid>
    <pubDate>Sat, 14 Nov 2026 12:00:00 +0000</pubDate>
    <description>Reputation work can push bad results down and build a better record. It cannot outrun the behavior that caused the damage. Change first, then it holds.</description>
    <content:encoded>&lt;h1&gt;Why reputation management only works when the behavior changes&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-11-14"&gt;November 14, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Online reputation management can change what a search for your name turns up. It can push bad results down, crowd them out with accurate material, and build a record that ranks above the damage. It cannot outrun the behavior that caused the problem. If the behavior continues, new complaints and fresh articles keep arriving, and each round of cleanup buys less than the round before it. I have argued in print since 2013 that the money only works after the client changes. This guide explains what the work can do, what it cannot do, and the order to do things in.&lt;/p&gt;

&lt;h2&gt;What reputation management can do&lt;/h2&gt;
&lt;p&gt;Reputation management is defensive SEO: the same craft that ranks a page for a product, pointed at a name. Done properly, it finds everything a search turns up about you, answers what deserves an answer, corrects what is factually wrong at the source where that is possible, and publishes enough accurate material on properties you control to push the bad results down where almost nobody looks.&lt;/p&gt;
&lt;p&gt;The building matters because silence is not neutral. I put it this way in a 2019 interview: "Google abhors a vacuum. When a community doesn't own its narrative, its narrative is vulnerable." If you publish nothing, the record of you is whatever other people happened to leave behind: an old lawsuit, an angry review, a namesake's arrest. Fill the vacuum and the picture changes: "If you create enough content, you can own the first two pages of Google."&lt;/p&gt;
&lt;p&gt;Notice what is missing from that list: deletion. Honest criticism almost never comes down, and a demand letter rarely removes a true story. The realistic tools are displacement and dilution. You bury the bad result; you rarely delete it. It stays online, one needle in a growing stack of needles, too deep for a searcher to reach. A bad result sitting at position three is a crisis. The same result at position thirty is a footnote.&lt;/p&gt;

&lt;h2&gt;What it cannot do&lt;/h2&gt;
&lt;p&gt;The work cannot make a false front true. As I said in the same interview: "Your reputation online can be curated, but it has to be the truth." A polished record built on behavior that has not changed is paint over a leak. My comparison for that is water: "Online reputation damage is like water damage. It'll seep through if you try to paper it over." The stain comes back because the pipe is still leaking, and the pipe is the conduct that produced the complaints in the first place.&lt;/p&gt;
&lt;p&gt;It also cannot stop the flow. Every new bad review, every fresh article, every court filing is new material, and search engines prefer new material. A campaign can bury last month's story. It cannot bury next month's, because next month's has not been written yet.&lt;/p&gt;

&lt;h2&gt;The law of diminishing returns&lt;/h2&gt;
&lt;p&gt;I worked this out in print in 2013, in an essay titled "Reputation management is not about dressing up mistakes." One of its sections is headed, without decoration, "The law of diminishing returns." The argument: brute force, backed by enough money, can hold a first page for a while, but Google normalizes the campaign, and "the newest, freshest, and most relevant real, organic, authentic, and popular content will bubble to the top." So the same monthly spend buys a smaller improvement every month, for as long as the underlying behavior keeps supplying that fresh content.&lt;/p&gt;
&lt;p&gt;In the same essay I gave the two situations their names. If you have been done wrong, by error or by malice, the job is finite; I call that scenario "the clean." If your own conduct keeps producing the problem, the job has no end; I call that "repair and repeat." Of the second job I wrote: "Online reputation management does work, but it's an endless game of whack-a-mole."&lt;/p&gt;
&lt;p&gt;The essay's opening still says the whole thing as plainly as I know how: "If you want to fix your online reputation, writing big checks to an online reputation agency isn't enough over the long haul—you'll also really need to get your own house in order."&lt;/p&gt;

&lt;h2&gt;Own it first: mea maxima culpa marketing&lt;/h2&gt;
&lt;p&gt;When the fault is yours, the first move is not content. The first move is to own the mistake in public, specifically, and to make amends where amends can be made. I coined a name for the full-strength version in 2013: mea maxima culpa marketing. In the 2019 interview I described it in one breath: "It's mea maxima culpa marketing. You show you're flexible, you listen."&lt;/p&gt;
&lt;p&gt;The apology has to be real, because a performed one backfires. From the 2013 essay: if the tone of contrition "seems hollow or false, you'll probably just make it worse, much worse." Amends are concrete: the refund issued, the job finished, the record corrected, the customer answered in the same place the complaint stands. That is not garnish on the campaign. It is the first piece of new evidence, and every later piece points back to it.&lt;/p&gt;

&lt;h2&gt;The turnaround has a name&lt;/h2&gt;
&lt;p&gt;The change-first argument in this guide dates to my writing of 2013. The word I use for the turnaround itself is newer. It is metanoia, a Greek word for a change of mind deep enough to change direction: stop doing the thing, repair what can be repaired, and become the person the new record describes. A record built for someone who has not turned is a costume, and a costume costs more to maintain every year. A record built for someone who has turned is simply documentation.&lt;/p&gt;

&lt;h2&gt;What you do first&lt;/h2&gt;
&lt;p&gt;If the problem is your own conduct, the order of operations is fixed, and the search work comes last:&lt;/p&gt;
&lt;ol&gt;
  &lt;li&gt;Stop the behavior that keeps producing complaints. If customers keep reporting the same failure, the failure is a business process, not a run of bad luck.&lt;/li&gt;
  &lt;li&gt;Make amends where amends are possible: refunds issued, work finished, records corrected, and direct answers to the people affected.&lt;/li&gt;
  &lt;li&gt;Fix the process underneath, so the same failure stops recurring on a schedule.&lt;/li&gt;
  &lt;li&gt;Own what is already public: answer the serious criticism where it stands, once, with facts and with the change attached.&lt;/li&gt;
  &lt;li&gt;Only then start the search work in the next section.&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;One exception, and it matters. If the damage was done to you rather than by you—a false claim, a mistaken identity, a story that belongs to someone else—skip this section. That is the clean, a finite job with a finish line. As I wrote in 2013: "If you've simply been besmirched, write a check to me."&lt;/p&gt;

&lt;h2&gt;How the work runs once the change is real&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;Audit the record. Search your name and your company the way a stranger, a customer, or a hiring manager would, and list what ranks on the first two pages.&lt;/li&gt;
  &lt;li&gt;Correct at the source. Factual errors get correction requests backed by evidence; outdated pages get updated or answered.&lt;/li&gt;
  &lt;li&gt;Respond where a response helps: once, in place, with the amends and the change stated plainly.&lt;/li&gt;
  &lt;li&gt;Build the record. Your own site, completed profiles, and steady publishing of accurate material fill the vacuum with the truth.&lt;/li&gt;
  &lt;li&gt;Tend it in small daily doses. Managing a reputation is like getting fitter: "Spend about 20 minutes a day, walk more, take the stairs at first. If you try too hard in a weekend, you might break something."&lt;/li&gt;
  &lt;li&gt;Monitor and repeat. Check the first two pages every month. Moving results is the braces; keeping them moved is the retainer, worn on a schedule. Because the behavior stopped feeding the problem, each month's gains hold instead of expiring, and the record compounds.&lt;/li&gt;
&lt;/ol&gt;

&lt;p class="related"&gt;I handle this work for clients as &lt;a href="https://gerriscorp.com/services/reputation/"&gt;reputation and perception management&lt;/a&gt;. The privacy side of the same problem is covered in my guide to &lt;a href="https://gerriscorp.com/guides/personal-information-removal/"&gt;removing your personal information from data brokers and search&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://www.argentum.org/expert-shares-best-practices-to-protect-your-online-reputation/"&gt;Argentum: Expert shares best practices to protect your online reputation&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://biznology.com/2013/02/mea-maxima-culpa-marketing/"&gt;Biznology: Mea maxima culpa marketing&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://chrisabraham.com/blog/reputation-management-is-not-about-dressing-up-mistakes"&gt;Chris Abraham: Reputation management is not about dressing up mistakes&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Programmatic SEO: what it is, where it works, and why Google ignores it</title>
    <link>https://gerriscorp.com/guides/programmatic-seo/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/programmatic-seo/</guid>
    <pubDate>Mon, 02 Nov 2026 12:00:00 +0000</pubDate>
    <description>What programmatic SEO is, where database driven pages earn their place, why Google leaves them unindexed, and the playbook that fixes a set at scale.</description>
    <content:encoded>&lt;h1&gt;Programmatic SEO: what it is, where it works, and why Google ignores it&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-11-02"&gt;November 2, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Programmatic SEO publishes pages from a database instead of an editor: one template, thousands of rows of structured data, and a page for every row, each aimed at a long-tail search. Built on real data, it produces some of the most useful pages on the web. Built on swapped variables and nothing else, it produces pages Google crawls once and refuses to index.&lt;/p&gt;

&lt;h2&gt;What is programmatic SEO?&lt;/h2&gt;
&lt;p&gt;Programmatic SEO is a publishing method with two working parts. The first is a data layer: a structured table of entities, such as products, properties, places, routes, or practitioners, where every row carries that entity's attributes. The second is a template layer: page designs that render one row each, under a predictable URL pattern. Point the templates at the table and the site gains a page for every row, each one matched to a query pattern such as {product} price, {service} in {city}, or {city} to {city} flights.&lt;/p&gt;
&lt;p&gt;Every content management system renders templates, so templating alone is not the method. Programmatic SEO begins when the database decides how many pages exist and which searches each page targets. A single build can create a few hundred pages or a few million, and nothing in the build checks whether each page deserves to exist. That check is the whole job, and it belongs to the people publishing, because Google performs its own version of it later and publishes the verdict in Search Console.&lt;/p&gt;

&lt;h2&gt;Where does programmatic SEO work?&lt;/h2&gt;
&lt;p&gt;It works where every row holds facts that are true of that entity alone, facts a rival cannot reproduce by renaming a variable. The durable examples share that property:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Property listings.&lt;/strong&gt; Price, beds, baths, photographs, floor area, days on market, and sale status. Two listings on the same street differ in ways a buyer acts on.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Product databases.&lt;/strong&gt; Specifications, price, stock, compatibility, manuals, and owner reviews. A parts catalog page answers questions no article could answer as well.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Location pages carrying genuine local data.&lt;/strong&gt; The address, the hours, the staff, the services that branch actually offers, local inventory, and reviews of that branch. A location page earns its URL with the branch's real record.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Pages computed from live data.&lt;/strong&gt; Prices, comparisons, schedules, and conversions drawn from feeds that change, so the page a visitor reads today differs from last month's because the facts moved.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;The test takes ten seconds. Hide the names, read two pages from the set side by side, and ask whether a visitor could tell them apart and learn something different from each. If the answer is yes, the set is publishing. If the pages collapse into one page wearing different nouns, Google reaches the same conclusion.&lt;/p&gt;

&lt;h2&gt;Why do programmatic pages fail to get indexed?&lt;/h2&gt;
&lt;p&gt;Indexing is a selection, and a template set that differs only in swapped variables gives Google nothing to select. The failure modes stack:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Near duplication.&lt;/strong&gt; Google crawls a sample of the set, compares the pages, and finds the same text arranged around different names and numbers. The rest of the set waits in "Discovered, currently not indexed" or falls into "Crawled, currently not indexed," the status the &lt;a href="https://gerriscorp.com/guides/crawled-not-indexed/"&gt;guide to Crawled, currently not indexed&lt;/a&gt; takes apart.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;No unique value per URL.&lt;/strong&gt; An indexed page has to be a good answer to something. A page whose only unique tokens are a city and a product name answers nothing its siblings fail to answer.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Weak internal linking.&lt;/strong&gt; Generated sets often sit deep in the site, reachable only through paginated archives or a sitemap file. Pages that nothing links to read as pages nobody values, and crawlers budget their visits accordingly.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Crawl waste.&lt;/strong&gt; Combinations, parameters, and sort orders multiply URLs far beyond the entity count, and Googlebot spends visits on variants while new or changed real pages wait. The &lt;a href="https://gerriscorp.com/guides/faceted-navigation/"&gt;faceted navigation guide&lt;/a&gt; covers the catalog version of this trap.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Site-level drag.&lt;/strong&gt; Thousands of weak URLs feed the quality signals Google computes for the whole domain and the crawl rate it grants. A flood in one template can slow indexing for pages that were never part of the flood.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;What does Google's scaled content abuse policy say?&lt;/h2&gt;
&lt;p&gt;In March 2024, Google added scaled content abuse to its &lt;a href="https://developers.google.com/search/docs/essentials/spam-policies"&gt;spam policies for Google web search&lt;/a&gt;. The policy names the practice plainly: generating many pages primarily to manipulate search rankings, with little or no value added for the people who land on them, is spam. The method of production makes no difference. Automation, human writers, and a mixture of the two all fall under the same rule, and the examples Google gives include bulk-generated pages published without added value, material scraped or stitched from other sources, and many sites used to hide a scaled operation.&lt;/p&gt;
&lt;p&gt;The policy judges the output and its purpose, never the technique. A database behind a page is ordinary publishing, and has been since the first catalog went online. A database behind ten thousand pages that exist to catch queries, carrying nothing a visitor could use, is the abuse. Sites that cross the line risk demotion or removal under the spam policies, and the cleanup takes far longer than the publishing did. Any programmatic project should be planned against that standard from the first template, because retrofitting value into a published flood is the expensive direction.&lt;/p&gt;

&lt;h2&gt;How do you fix a programmatic site Google will not index?&lt;/h2&gt;
&lt;p&gt;Work at the level of the template and the dataset, never page by page. A set of fifty thousand URLs has perhaps six problems, one per template, and fixing templates fixes the set.&lt;/p&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Audit each template against demand and data.&lt;/strong&gt; For every page type, write down the searches it targets and the facts unique to each row. A type with no real demand and no unique data is a removal candidate, however many pages it has already published.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Feed the survivors real data.&lt;/strong&gt; Live prices and availability, local facts, photographs, reviews, computed comparisons: material a competitor cannot produce by renaming a variable. Where the data cannot be obtained, the honest move is to shrink the set until it matches the data on hand.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Chunk the content.&lt;/strong&gt; Break each template into blocks that each answer one question: the specification table, the price block, the local details, the questions buyers actually ask, each under a descriptive heading. Readers scan blocks, and answer engines extract from blocks. A wall of prose with variables poured in defeats both.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Match schema to the page type.&lt;/strong&gt; Product pages carry Product data, listings carry the listing type, locations carry LocalBusiness, and hubs carry ItemList, always describing content the visitor can see. The &lt;a href="https://gerriscorp.com/guides/structured-data/"&gt;structured data guide&lt;/a&gt; sorts the types worth using from the ones Google retired.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Rebuild internal linking as hub and spoke.&lt;/strong&gt; A hub for each category or region links down to every spoke beneath it; spokes link back to the hub and across to true siblings. Keep the set a few clicks from the home page, keep the sitemap limited to pages meant for the index, and treat &lt;a href="https://gerriscorp.com/services/internal-linking/"&gt;internal linking architecture&lt;/a&gt; as part of the publishing system rather than decoration added later.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Prune what cannot justify itself.&lt;/strong&gt; Delete the combinations nobody searches for, consolidate near-duplicates into the strongest URL with redirects, and noindex the variants visitors need but engines do not, such as sort orders and filter states. Pages nothing links to and nobody needs can go with a 410. A smaller set in which every page carries real facts recrawls faster and indexes better than a large set diluted by filler.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Earn real links.&lt;/strong&gt; Publish the interesting edges of the dataset: rankings, research, tools, and local data that journalists, partners, and customers cite. Links into the hubs lift the whole set, while a large inventory with no links pointing at it asks Google to take the site's word for its importance.&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;Measure the result by template in Search Console's Page indexing report. Movement shows over weeks as Google recrawls the set, and on a pruned site the strongest pages usually get picked up first.&lt;/p&gt;

&lt;section class="faq"&gt;
&lt;h2&gt;Questions about programmatic SEO&lt;/h2&gt;
&lt;h3&gt;Is programmatic SEO against Google's rules?&lt;/h3&gt;
&lt;p&gt;No. The spam policies ban scaled content abuse: many pages produced to manipulate rankings with little value added. A database-driven page carrying real facts about a real entity is ordinary publishing, and some of the most heavily indexed pages on the web are built exactly that way.&lt;/p&gt;
&lt;h3&gt;How many pages is it safe to generate?&lt;/h3&gt;
&lt;p&gt;The policies set no number. Safety scales with data: a million listings backed by live records index well, while a thousand city pages carrying swapped names struggle. Count the unique facts on each page rather than counting pages.&lt;/p&gt;
&lt;h3&gt;Why does Google crawl my pages and still refuse to index them?&lt;/h3&gt;
&lt;p&gt;Crawling is cheap and indexing is a judgment. Google samples the set, finds pages that differ mainly in variables, and declines to spend index space on the rest. The &lt;a href="https://gerriscorp.com/guides/crawled-not-indexed/"&gt;guide to Crawled, currently not indexed&lt;/a&gt; gives the full method for confirming the cause on a specific site.&lt;/p&gt;
&lt;h3&gt;Should thin pages be noindexed or deleted?&lt;/h3&gt;
&lt;p&gt;Delete pages nobody needs and nothing links to; a 410 tells Google the removal was deliberate. Noindex pages visitors still need, such as filtered views, so they stay reachable while leaving the index. Consolidate pages that split one topic into near copies into the strongest URL and redirect the rest.&lt;/p&gt;
&lt;h3&gt;Does adding schema get programmatic pages indexed?&lt;/h3&gt;
&lt;p&gt;No. Schema describes a page; it cannot give a page value. Accurate, type-matched structured data helps engines understand and display a page that already deserves indexing, and wasted schema on thin pages changes nothing.&lt;/p&gt;
&lt;h3&gt;How long does recovery take after a prune?&lt;/h3&gt;
&lt;p&gt;Expect the first movement in weeks and the full picture across a quarter. Google recrawls at its own pace, strongest pages first. Track the Page indexing report template by template and resist republishing the pruned set while the recrawl is under way.&lt;/p&gt;
&lt;/section&gt;

&lt;p class="related"&gt;I run this diagnosis as a fixed-fee &lt;a href="https://gerriscorp.com/services/indexing-audit/"&gt;indexing audit&lt;/a&gt;. See also &lt;a href="https://gerriscorp.com/case-studies/programmatic-indexing/"&gt;the programmatic indexing case study&lt;/a&gt; and the &lt;a href="https://gerriscorp.com/guides/faceted-navigation/"&gt;faceted navigation guide&lt;/a&gt; for the catalog version of the crawl problem.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/essentials/spam-policies"&gt;Google Search Central: Spam policies for Google web search&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/large-site-managing-crawl-budget"&gt;Google Search Central: Managing crawl budget for large sites&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/fundamentals/creating-helpful-content"&gt;Google Search Central: Creating helpful, reliable, people-first content&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Podcasting as an investment in being found by AI search</title>
    <link>https://gerriscorp.com/guides/podcast-investment/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/podcast-investment/</guid>
    <pubDate>Sat, 24 Oct 2026 12:00:00 +0000</pubDate>
    <description>A podcast is an asset that keeps paying. Each episode's notes are copied word for word by hundreds of directories, and AI answers quote them for years to come.</description>
    <content:encoded>&lt;h1&gt;Podcasting as an investment in being found by AI search&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-24"&gt;October 24, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Most owners who start a podcast treat it as content: record an episode, promote it for a week, move on to the next one. That undersells the asset. Done properly, a podcast is closer to buying a building than to renting a billboard. The recording is a single afternoon of work, but the words attached to it are copied onto hundreds of trusted sites, and those copies are what AI answer engines read when your customers ask questions.&lt;/p&gt;

&lt;h2&gt;One recording, hundreds of copies&lt;/h2&gt;
&lt;p&gt;A podcast runs on a feed, and the feed is a syndication machine. Your host sends every episode to the podcast directories: Apple Podcasts, Spotify, the Podcast Index, Pocket Casts, Castbox, TuneIn, and hundreds of smaller apps and sites that most owners have never heard of. Each directory republishes the episode with its show notes, word for word, exactly as you wrote them. Your business name, your links, and your claims, in your own phrasing, end up living on all of them at once.&lt;/p&gt;
&lt;p&gt;That matters more than it used to, because the directories are no longer the only readers. Search engines index those pages, and the AI answer engines, Google's AI answers, ChatGPT, and Perplexity among them, read the same copies when they work out what to say about a business. When a customer asks an AI assistant who to call or what a company believes, the answer is assembled from text like this. An owner who has recorded fifty episodes has fifty topics on which hundreds of sites repeat his own words back to the machines. An owner who has recorded none is quoted from whatever other people wrote about him, or not quoted at all.&lt;/p&gt;

&lt;h2&gt;Why it compounds instead of expiring&lt;/h2&gt;
&lt;p&gt;Compare the asset to the usual alternatives. Advertising stops the day the payments stop. A social post is buried within days. A single article can rank for a season and then slip as competitors publish over it. A podcast back catalog behaves differently: every episode stays listed in every directory, keeps collecting listeners, and keeps standing as a permanent, citable statement of what you know and what you sell. Each new episode adds another page of quotable text and strengthens the catalog behind it.&lt;/p&gt;
&lt;p&gt;The production cost is paid once, and it is paid in a currency owners already spend. The raw material of a business podcast is the conversations you have anyway: the questions customers ask every week, the explanations you give on sales calls, the stories you tell about work that went right and work that went wrong. Record those conversations once and they keep answering for you, in your voice, for years. That is what makes the show a citation factory rather than a content chore: the factory runs on talk that was already happening.&lt;/p&gt;

&lt;h2&gt;Proof, from my own show&lt;/h2&gt;
&lt;p&gt;I host a podcast, The Chris Abraham Show, and the mechanism above is not theory to me. On October 24, 2026, a searcher asked Google what Chris Abraham thinks about a television show, the Canadian detective series Private Eyes, and Google's AI answered by quoting my podcast's show notes nearly word for word: the passage where I explained that I came for Jason Priestley and the detective stories and stayed for the patience, the apologies, and the father and daughter relationships. The same notes sit on Apple Podcasts, on Podcast Addict, and on the show's own website, all carrying the same text from the same feed.&lt;/p&gt;
&lt;p&gt;Nobody pitched that answer and nobody paid for the placement. The notes were written to be quotable, the feed carried them everywhere a podcast can be listed, and the engine did the rest. Note the subject, too: a question with no commercial value at all, answered from my own words because my words were the ones available to quote. Now run the same machinery on questions in your market, the ones your buyers actually ask before they spend money.&lt;/p&gt;

&lt;h2&gt;What doing it right involves&lt;/h2&gt;
&lt;p&gt;Three things separate a show that becomes an asset from a show that is just audio.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Show notes written to be quoted.&lt;/strong&gt; Spell out full names of people, companies, and products. Link to the canonical page on your own site. State your claims plainly and completely, because a directory or an AI engine will repeat them exactly as written. Vague notes produce vague answers.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;An owned show site with transcripts.&lt;/strong&gt; Platform pages are thin and platform notes are capped; Spotify cuts show notes off at about 4,000 characters. A site on your own domain carries the full notes and the full transcript of every episode, in text the search and AI crawlers can read completely, and it is the one copy no platform can take away. Here is &lt;a href="https://gerriscorp.com/guides/podcast-website/"&gt;how that site gets built&lt;/a&gt;.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Monthly measurement.&lt;/strong&gt; Ask the questions your buyers ask, of Google, ChatGPT, and Perplexity, every month, and write down what the answers say. That is how you watch citations arrive, and how you catch a wrong claim early enough to correct it at the source. The method is in &lt;a href="https://gerriscorp.com/guides/measure-ai-search/"&gt;the guide to measuring AI search&lt;/a&gt;.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;The practitioner details, episode pages, schema, and indexing, are in &lt;a href="https://gerriscorp.com/guides/podcast-seo/"&gt;the podcast SEO guide&lt;/a&gt;. This page is the owner's version of the same argument: the technique is simple, and the returns belong to whoever bothers to set it up properly and keep recording.&lt;/p&gt;

&lt;h2&gt;Who should buy this asset&lt;/h2&gt;
&lt;p&gt;Honesty about the time frame: a podcast is a slow asset. If you need leads this month, advertising buys them faster, and I will tell you so on a call. A catalog pays out over seasons and years, to a business its owner intends to keep. If that is the business you are building, few marketing assets treat you better, because this one gets more valuable the longer you hold it and cannot be outbid by a competitor with a larger monthly budget.&lt;/p&gt;

&lt;p&gt;I set up podcast sites, show notes, transcripts, and AI answer measurement for clients, and every engagement starts with a check of what the AI answers currently say about the business. If you want that check on yours, or you want a show of your own working this way, &lt;a href="https://gerriscorp.com/contact/"&gt;ask me to look at it&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://podcastindex.org/"&gt;Podcast Index&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://podcasters.apple.com/"&gt;Apple Podcasts for Podcasters&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Your Google business reports are getting a new scoreboard</title>
    <link>https://gerriscorp.com/guides/google-business-profile-reports/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/google-business-profile-reports/</guid>
    <pubDate>Sat, 24 Oct 2026 12:00:00 +0000</pubDate>
    <description>Google is testing rebuilt performance reports for business profiles, with five new sections. I explain what each one tells an owner and what to check monthly.</description>
    <content:encoded>&lt;h1&gt;Your Google business reports are getting a new scoreboard&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-24"&gt;October 24, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;If you run a local business, your Google Business Profile is the listing that shows your hours, your phone number, your reviews, and your spot on the map when a customer searches nearby. Google is now testing a rebuilt set of performance reports for that listing: five new sections that tell you far more than the old strip of numbers did. I wrote about the news on &lt;a href="https://chrisabraham.net/2026/10/google-rebuilds-the-scoreboard.html"&gt;my weblog&lt;/a&gt; the day it broke. This guide is the owner version: what each section means, how to read the numbers without fooling yourself, and what to do about them each month.&lt;/p&gt;

&lt;h2&gt;What Google is changing&lt;/h2&gt;
&lt;p&gt;The old Performance tab was thin: a count of views, calls, direction requests, and website clicks, take it or leave it. The test version, reported by Search Engine Roundtable on October 23, 2026 from screenshots and video sent in by two practitioners who received it early, is a real reporting suite. It arrived one day after Google launched a new home portal for business profiles, which says this is a program, not a stray experiment. The reporter who has covered this beat longer than anyone called it a very big improvement on the current reports.&lt;/p&gt;
&lt;p&gt;Calibrate before celebrating: it is a test. It may reach your account next week or next quarter, and Google has retired better ideas than this one. Nobody should rebuild a reporting habit this week on the strength of screenshots. But the direction is unmistakable. Google is investing in telling owners, in detail and with advice attached, what their listings are doing. An owner who learns to read the new scoreboard early gets the benefit of that investment first.&lt;/p&gt;

&lt;h2&gt;The five new sections, in plain language&lt;/h2&gt;
&lt;h3&gt;Overview&lt;/h3&gt;
&lt;p&gt;The front page of the new reports, and the section to read first. At the top, Google places its own recommendations for your profile: the facts to complete, the photos to add, the details it thinks are costing you customers. Read it as a to do list written by the referee. The referee works for Google, so check the advice against your own judgment, but ignoring homework from the outfit that ranks you is rarely wise.&lt;/p&gt;
&lt;h3&gt;Engagement&lt;/h3&gt;
&lt;p&gt;What people did once your listing caught their eye: the calls placed, the messages sent, the bookings started, the website visits, the direction requests, gathered in one place instead of scattered. This is the section that comes closest to money, because each entry is a customer who moved from looking to acting.&lt;/p&gt;
&lt;h3&gt;Reviews&lt;/h3&gt;
&lt;p&gt;Your reviews, broken out in detail instead of flattened into a single star average. An average hides everything that matters. The detail view shows the recent reviews themselves, the patterns inside them, what customers praise, what they complain about, and whether each one got a reply. Reputation is most of the conversion game for a local business, and this section turns it from a feeling into a ledger.&lt;/p&gt;
&lt;h3&gt;Audience&lt;/h3&gt;
&lt;p&gt;Who is finding you. The audience section describes the people your listing reaches, including the split between people meeting your business for the first time and people coming back to it. Owners ask two questions more than any others: who is finding me, and is it the right crowd? This is the first time the profile reports have tried to answer the first one directly.&lt;/p&gt;
&lt;h3&gt;Search trends&lt;/h3&gt;
&lt;p&gt;The actual words people typed that brought your listing up, and how those words change over time. This is your market describing your business in its own language, which is rarely the language on your sign. If searchers keep finding you under words for work you no longer do, or fail to find you under the words for the work you want, this section shows it before the phone goes quiet.&lt;/p&gt;

&lt;h2&gt;Why your clicks and calls can fall while business holds up&lt;/h2&gt;
&lt;p&gt;Here is the reading skill the new reports demand, because the old habit will mislead you. A customer who asks Google a question increasingly gets the answer from an AI summary at the top of the results: your hours, your phone number, whether you are any good, read straight off your profile and your reviews. That customer got what they needed and never clicked anything. The action happened; it just never registered on your listing.&lt;/p&gt;
&lt;p&gt;An analysis published in October 2026 followed three established florists in the United States and found profile interactions down 26 percent against the same period a year earlier, with website clicks down 30 percent, phone calls down 21 percent, and direction requests down 15 percent, on profiles that were active and well kept. Three florists are a signal, not a census, so read those numbers as direction rather than proof. The direction matches what I see across local accounts: the measurable actions on a profile are getting scarcer because the AI answers absorb them first. Your profile now matters as the source the machine reads, not only as the page the customer visits. Judge your marketing by the phone, the bookings, and the till alongside the report, never by one falling line on a chart.&lt;/p&gt;

&lt;h2&gt;Fifteen minutes a month: the owner routine&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Save the numbers first.&lt;/strong&gt; Screenshot or export this month's figures before anything else. Baselines are the first casualty of every dashboard migration, and nobody can prove growth against a number that no longer exists.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Read the Overview recommendations.&lt;/strong&gt; Do each one, or hand it to whoever manages your listing with a date attached.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Read every new review and answer each one.&lt;/strong&gt; The new detail view is about to make neglected reviews visible to anyone who looks, starting with you.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Compare against the same month last year.&lt;/strong&gt; Seasonality makes last month a liar for most local businesses. The honest comparison is a year back.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Read Search trends against your price list.&lt;/strong&gt; The words bringing people to your door should match the work you actually want to sell.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Ask the machines about your business.&lt;/strong&gt; Once a month, put your name into Google, ChatGPT, and Perplexity and note what the AI answers say. I keep &lt;a href="https://gerriscorp.com/guides/measure-ai-search/"&gt;a guide to measuring AI search&lt;/a&gt; for exactly this habit.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;What to ask whoever manages your listing&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Have the new reports reached my account yet, and if so, what do the Overview recommendations say?&lt;/li&gt;
  &lt;li&gt;What were my calls, bookings, and direction requests this month, set against the same month last year?&lt;/li&gt;
  &lt;li&gt;Has every review been answered, and what are reviewers praising and complaining about this month?&lt;/li&gt;
  &lt;li&gt;Which search words are bringing my listing up, and do they match the work I sell?&lt;/li&gt;
  &lt;li&gt;What do the AI answers say about my business this month, and is every word of it true?&lt;/li&gt;
  &lt;li&gt;What changed on the profile this month, and why?&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;A manager who can answer those six questions in writing, briefly, once a month, is earning the fee. A manager who sends a screenshot of a views chart is not. That standard is the whole of &lt;a href="https://gerriscorp.com/services/google-business-profile/"&gt;my Google Business Profile service&lt;/a&gt;: the work reported in writing, with numbers, on a rhythm an owner can hold me to.&lt;/p&gt;

&lt;h2&gt;When to worry, and when not to&lt;/h2&gt;
&lt;p&gt;Worry when the rating slides across several months rather than dipping once. Worry when reviews sit unanswered, when wrong facts surface in the AI answers about your hours, your address, or what you sell, and when Search trends drift toward work you do not offer. Worry most when every section falls at once and the phone goes quiet with them, because that pattern is a business problem and not a measurement problem.&lt;/p&gt;
&lt;p&gt;Do not worry when clicks or calls ease downward while bookings and the register hold up: the machine answered on the spot, and the customer still came. Do not worry that your account lacks the new sections yet, since a test rolls out account by account. And do not worry over one soft month in a seasonal trade, which is weather, not climate.&lt;/p&gt;

&lt;section class="faq"&gt;
&lt;h2&gt;Questions owners ask about the new reports&lt;/h2&gt;
&lt;h3&gt;My calls are down. Is my business in trouble?&lt;/h3&gt;
&lt;p&gt;Not on that evidence alone. AI answers now hand out your phone number, your hours, and your rating before a customer ever reaches your listing, so some calls and clicks simply stop registering. Put the report next to your bookings and your till. If those hold while the chart falls, the likeliest story is that the answers absorbed the actions. If everything falls together, that is a real problem and worth a proper diagnosis.&lt;/p&gt;
&lt;h3&gt;Do I have the new reports yet?&lt;/h3&gt;
&lt;p&gt;Open your profile's Performance section and look for the five sections named in this guide. If they are not there, the test has not reached your account. Google rolls tests out gradually and quietly; even the trade reporter who broke the story did not have it on the day he published. Nothing is wrong with your account.&lt;/p&gt;
&lt;h3&gt;What should I do this week?&lt;/h3&gt;
&lt;p&gt;Three things. Save your current Performance numbers somewhere safe, because the old figures may not survive the migration. Answer any review that is still waiting for a reply. And check that the facts on your profile, hours, phone, address, services, are exactly right, because those facts are what the AI answers repeat.&lt;/p&gt;
&lt;/section&gt;

&lt;p&gt;Your Google Business Profile is a core of my practice at Gerris, and the deal I make with every owner is simple: the work gets reported in writing, with numbers, on a rhythm you can hold me to. A richer scoreboard makes that deal easier to keep and easier to check. If you want the new reports read for your business and the findings put plainly, &lt;a href="https://gerriscorp.com/contact/"&gt;ask me to look at it&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://www.seroundtable.com/google-business-profiles-new-performance-reports-42367.html"&gt;Search Engine Roundtable&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://chrisabraham.net/2026/10/google-rebuilds-the-scoreboard.html"&gt;Chris Abraham&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>When a Customer Asks for the Best, a Machine Answers</title>
    <link>https://gerriscorp.com/guides/ai-answers-review-searches/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/ai-answers-review-searches/</guid>
    <pubDate>Fri, 23 Oct 2026 12:00:00 +0000</pubDate>
    <description>Google's AI Overviews now answer most review searches. What the machines say about untended businesses, and the audit, sprint, and measurement that fix it.</description>
    <content:encoded>&lt;h1&gt;When a Customer Asks for the Best, a Machine Answers&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-23"&gt;October 23, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Somebody in your service area is, right now, asking Google a question like "best [your trade] near me" or "who should I call for [your service]." A year ago they got a list of businesses and made their own shortlist. Today they get a written answer: two or three names, a sentence about each, and a recommendation. If your name is not in that paragraph, you were never in the running, and no amount of scrolling changes it, because most people never scroll.&lt;/p&gt;
&lt;p&gt;This is not a prediction. Search Engine Land, reporting Semrush's tracking data, found that Google's AI Overviews appeared for 88% of commercial keywords in August 2026, nearly double their share in September 2025. Of the five search types Semrush studied, review searches—exactly the "best," "top-rated," and "near me" questions above—moved the furthest. The searches that end in a phone call are the searches the machines took over first.&lt;/p&gt;
&lt;h2&gt;What the answer says about a business nobody is tending&lt;/h2&gt;
&lt;p&gt;An AI answer is only as good as the record it is built from: your website, your Google Business Profile, your listings, and your reviews, reconciled by a system that cannot call you to check. When that record is untended, the answer does not come back blank. It comes back wrong, in predictable ways:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;&lt;strong&gt;It names a competitor&lt;/strong&gt; whose facts happen to agree with each other across sources, because agreement is what the system reads as truth.&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;It describes the business you used to be&lt;/strong&gt;—services you dropped, a location you closed, hours you changed—because old listings never die on their own.&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;It blends you with somebody else&lt;/strong&gt;: a similarly named business across town or across the country, your details and theirs averaged into a company that does not exist.&lt;/li&gt;
&lt;li&gt;&lt;strong&gt;It leaves you out entirely&lt;/strong&gt; on the searches that carry the most money, while happily citing you for trivia.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;None of this requires an enemy or a glitch. It is what an untended record produces, automatically, every day.&lt;/p&gt;
&lt;h2&gt;Why service businesses sit in the blast radius&lt;/h2&gt;
&lt;p&gt;Review and recommendation searches are buying searches. The person asking has a problem, a budget, and a short afternoon. The machine's paragraph is not the top of their research; for most of them it is the whole of it. A business can have loyal customers, strong reviews, and a good website, and still lose that customer before the first click, because the answer was assembled from a profile nobody has opened in three years and a directory listing nobody claimed.&lt;/p&gt;
&lt;p&gt;The good news is that the same mechanism cuts the other way. The answer rewards the business whose record is consistent, complete, and claimed. That is work. It is also, for once, work a smaller business can finish before a bigger competitor bothers to start.&lt;/p&gt;
&lt;h2&gt;What Gerris does about it&lt;/h2&gt;
&lt;p&gt;&lt;strong&gt;The AI Visibility Audit—$750 flat.&lt;/strong&gt; We run your money searches through Google AI Overviews, ChatGPT, Perplexity, and Gemini, and document what the answers say: where you are named, where you are missing, where you are described wrongly, and where a competitor stands in your place. We check your name, address, and phone consistency across your site, your Google Business Profile, and your listings. You receive a written report, fixes in priority order, yours to keep whoever does the work.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;The Claim Sprint—$1,500 flat.&lt;/strong&gt; We fix the layer the audit exposes: the technical foundation (crawl, Search Console, titles, broken links) and the claim layer (entity schema, llms.txt, a canonical page that states what your business is, FAQ schema), so the machines can quote you instead of reconstructing you. Items that belong with your developers come back as a precise written fix list.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;AI Presence, Measured—$500/month.&lt;/strong&gt; Answers drift. We re-run your search set monthly and deliver a written receipt: what changed, what the answers say now, and what gets fixed next.&lt;/p&gt;
&lt;h2&gt;Your Google Business Profile is doing more talking than you think&lt;/h2&gt;
&lt;p&gt;For local and service businesses, the Google Business Profile is one of the primary sources the AI answers are built from. A suspended profile, a duplicate listing, a wrong service area, or a category that undersells you does not stay a maps problem; it propagates into every answer the machines give about you. Gerris's profile work—service-area configuration, reinstatement support, and cleanup—is run under the same direction as our search work for one reason: your profile, your website, and your listings have to tell one story, because one story is what the machines believe.&lt;/p&gt;
&lt;h2&gt;Start with the audit&lt;/h2&gt;
&lt;p&gt;You do not need to guess what the machines say about your business, and you do not need a retainer to find out. Start with the $750 audit, read it, and decide from evidence. &lt;strong&gt;&lt;a href="https://gerriscorp.com/contact/"&gt;Contact Gerris&lt;/a&gt; to start with the audit.&lt;/strong&gt;&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>B2B catalog SEO: help the right buyer choose the right part</title>
    <link>https://gerriscorp.com/guides/b2b-catalog-seo/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/b2b-catalog-seo/</guid>
    <pubDate>Tue, 20 Oct 2026 12:00:00 +0000</pubDate>
    <description>Help technical buyers find specifications, compare parts, and request a useful quote. A B2B catalog SEO guide with a pilot and measurement plan.</description>
    <content:encoded>&lt;h1&gt;B2B catalog SEO: help the right buyer choose the right part&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-20"&gt;October 20, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A catalog can attract visitors while leaving the sales team with the wrong inquiries. The useful question is whether a buyer can establish that a part fits the application, understand the buying terms, and send enough information for a meaningful quote. I start there, then work backward through the content and technical SEO.&lt;/p&gt;
&lt;h2&gt;Separate the buyer questions before adding pages&lt;/h2&gt;
&lt;p&gt;An engineer checking compatibility, a purchasing manager sourcing a known part, and a hobbyist buying one replacement can use similar language. Their needs differ. Review actual inquiries with sales before treating every visit as equally valuable. Record which product families bring repeat business, which requests cannot be served, and which information sales repeatedly has to ask for.&lt;/p&gt;
&lt;p&gt;Use that evidence to choose a small pilot family. A list of twenty commercially important parts can teach more than a rewrite of five thousand descriptions. Pick products with reliable source data, a sales owner who can review the copy, and enough existing demand to make observation worthwhile. Do not select solely by search volume.&lt;/p&gt;
&lt;h2&gt;Make the specification understandable and inspectable&lt;/h2&gt;
&lt;p&gt;Put the information needed for an initial decision into readable page text. For a hypothetical fitting, that could include thread standard, connection size, material, dimensions, and the conditions attached to a stated pressure rating. The exact fields must come from the manufacturer's approved documentation. An SEO writer should never infer compatibility or a safety rating from a similar product.&lt;/p&gt;
&lt;p&gt;A practical product template includes:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;The manufacturer and exact part identifier, including punctuation that changes the identifier.&lt;/li&gt;
&lt;li&gt;A concise description that distinguishes this item from adjacent variants.&lt;/li&gt;
&lt;li&gt;A labeled specification table with units and the applicable conditions.&lt;/li&gt;
&lt;li&gt;A current datasheet, its revision where available, and any limits on the information shown.&lt;/li&gt;
&lt;li&gt;Ordering information such as pack quantity, minimum order, or the need to confirm availability.&lt;/li&gt;
&lt;li&gt;A quote or purchase route that carries the product context into the next step.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Keep downloadable drawings and PDFs where they help the buyer. Add enough HTML context that someone can understand the item before downloading. Check whether tabs, selectors, or scroll effects hide the essential text from a plain fetch. My &lt;a href="https://gerriscorp.com/services/crawler-visibility/"&gt;crawler visibility service&lt;/a&gt; addresses that rendering problem.&lt;/p&gt;
&lt;h2&gt;Choose which pages deserve separate addresses&lt;/h2&gt;
&lt;p&gt;A product, a product family, an application, and a filter combination are different editorial decisions. Give an individual part a useful page when buyers need that item's specific information. Use a family page to explain selection among genuinely related options. An application page should answer a real use question with expert review, rather than repeat a category description under a new heading.&lt;/p&gt;
&lt;p&gt;Avoid creating a page for every possible combination of specifications just because the database permits it. Sample filter URLs, identify duplicates, and agree which combinations deserve discovery. Check internal links, canonical signals, and sitemap entries together. See the &lt;a href="https://gerriscorp.com/guides/faceted-navigation/"&gt;faceted navigation guide&lt;/a&gt; for the underlying crawl problem.&lt;/p&gt;
&lt;p&gt;For obsolete parts, ask whether the page still helps owners identify an installed component. A discontinued notice and a documented replacement can be useful. Do not redirect to a superficially similar item whose compatibility is unknown. Preserve the distinction between a verified replacement and an alternative that needs engineering review.&lt;/p&gt;
&lt;h2&gt;Give technical buyers a usable quote route&lt;/h2&gt;
&lt;p&gt;A generic contact form can lose the part number the visitor just selected. Test the route from the actual product page. The form should preserve the relevant context and explain what information helps sales respond: quantity, timing, destination, and application details where appropriate. Ask the commercial team which fields are essential rather than making every field mandatory.&lt;/p&gt;
&lt;p&gt;State genuine constraints before submission. If the business only supplies trade accounts or has a minimum order, let buyers know. That may reduce inquiry volume while improving fit. A qualification rule should describe how the company actually operates, not discourage an unfamiliar buyer who could become a valuable account.&lt;/p&gt;
&lt;h2&gt;Use structured data without inventing a transaction&lt;/h2&gt;
&lt;p&gt;Structured data should describe the visible product accurately. Google's &lt;a href="https://developers.google.com/search/docs/appearance/structured-data/product-snippet"&gt;product snippet documentation&lt;/a&gt; explains eligibility and required properties. Check the applicable requirements before promising an enhanced search appearance. A quote-only catalog must not manufacture prices, reviews, or stock status to satisfy a validator. Valid descriptive markup and eligibility for a particular search feature are separate questions.&lt;/p&gt;
&lt;h2&gt;Measure the pilot against commercial fit&lt;/h2&gt;
&lt;p&gt;Before editing, save the selected URLs, their current content, indexing observations, relevant query groups, and available inquiry counts. Define a qualified inquiry with sales in writing. That might require a supported product, a serviceable destination, and enough quantity or account potential to justify follow-up. Keep rejected inquiries visible in the review so the definition cannot quietly change to flatter the results.&lt;/p&gt;
&lt;p&gt;After release, verify the pages, links, forms, and analytics events. Review the same product family over an agreed period that reflects the sales cycle. Compare visibility, inquiries, qualified opportunities, and eventual orders where records allow it. Annotate promotions, stock changes, and seasonal effects. A small sample can indicate a direction without proving that SEO caused every order.&lt;/p&gt;
&lt;h2&gt;How I take this on&lt;/h2&gt;
&lt;p&gt;My &lt;a href="https://gerriscorp.com/services/b2b-catalog-seo/"&gt;B2B catalog SEO service&lt;/a&gt; starts with a paid diagnostic and a bounded product family. I map the content gaps, specify template changes, coordinate expert review, and verify implementation. The deliverable is a repeatable page standard and an evidence trail for the pilot, with a decision about whether the rest of the catalog warrants the same work.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>A website handover needs more than a zip file and a password</title>
    <link>https://gerriscorp.com/guides/website-handover/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/website-handover/</guid>
    <pubDate>Thu, 15 Oct 2026 12:00:00 +0000</pubDate>
    <description>A website handover checklist for owners: account control, reproducible builds, hosting, DNS, data, backups, renewals, and proof a new maintainer can operate it.</description>
    <content:encoded>&lt;h1&gt;A website handover needs more than a zip file and a password&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-15"&gt;October 15, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;The contractor says the website is yours. Can someone new rebuild it, publish a correction, renew the domain, and restore the data without calling that contractor? Those are practical handover questions, and they can be answered while everyone is still available to help. Here is the inventory I would want before accepting responsibility for an inherited site.&lt;/p&gt;

&lt;h2&gt;Start with the account that controls each layer&lt;/h2&gt;
&lt;p&gt;A domain registration, a DNS zone, a hosting subscription, and a code repository can belong to four different accounts. List them separately. For each, record the business owner, the people with administrative access, the billing contact, and the recovery route. The person who can edit a page is not necessarily the person who can recover the account.&lt;/p&gt;
&lt;p&gt;The &lt;a href="https://gerriscorp.com/downloads/website-handover-inventory.csv"&gt;website handover inventory worksheet&lt;/a&gt; provides rows for these layers and room to record evidence. Put account identifiers in the private working copy, but keep passwords, recovery codes, tokens, and private keys in an approved secrets manager. The worksheet should say where access is managed, not become another loose credentials file.&lt;/p&gt;

&lt;h2&gt;Prove the source matches the site&lt;/h2&gt;
&lt;p&gt;Ask which repository and revision produced the current deployment. Keep the build instructions, runtime requirements, dependency files, and a list of required configuration variable names with the project. A useful handover lets a new maintainer start from the recorded revision in a fresh workspace and reproduce the intended result.&lt;/p&gt;
&lt;p&gt;Try that build before treating the instructions as complete. If it needs a file that exists only on somebody's laptop, an unrecorded account, or a package that cannot be retrieved, the trial has found an unresolved dependency. Record the dependency and its owner rather than quietly borrowing the departing developer's working directory.&lt;/p&gt;

&lt;h2&gt;Separate a repository transfer from an access review&lt;/h2&gt;
&lt;p&gt;GitHub documents that secrets, deploy keys, and webhooks can remain associated with a transferred repository. Receiving ownership therefore does not by itself settle who can still act on the project. Review the actual connections, collaborators, automation accounts, and deployment credentials as part of the handover.&lt;/p&gt;
&lt;p&gt;Agree the timing of access changes with the people responsible for the live service. An old key may still be powering the production build. Replace and test necessary connections before withdrawing the previous ones, then record that the old access has been removed. This sequence keeps an ownership improvement from becoming an accidental outage.&lt;/p&gt;

&lt;h2&gt;Inventory the things outside Git&lt;/h2&gt;
&lt;p&gt;A working site may depend on a database, uploaded media, scheduled jobs, a mail provider, search indexes, an external form service, or storage controlled by a different account. Include the dependencies the visitor never sees. A nightly import that stops running can make the site wrong for weeks before anybody notices that a background job was missed.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Content and files:&lt;/strong&gt; where editable content, original artwork, and visitor uploads live.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Recurring jobs:&lt;/strong&gt; what runs, on whose service, and who receives a failure alert.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Forms and notifications:&lt;/strong&gt; where submissions are stored and how delivery is verified.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Analytics:&lt;/strong&gt; which properties and accounts contain the business's reporting history.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Licenses:&lt;/strong&gt; which commercial components need renewal or a new owner.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Restore something before calling it a backup&lt;/h2&gt;
&lt;p&gt;List what each backup includes, its schedule, where it is stored, and who can retrieve it. Then restore the necessary components into a separate test environment. A repository backup is useful, but the application may also need data and media stored elsewhere. GitHub's backup documentation explains repository backup methods; the running service needs its own inventory around them.&lt;/p&gt;
&lt;p&gt;Choose a small, recognizable check after the restore: a particular article, a known uploaded file, or a test record that should exist. Record the backup date and the result. Keep test notifications and outbound integrations disabled while checking the restored copy so a rehearsal does not contact real customers.&lt;/p&gt;

&lt;h2&gt;Write down the domain and email dependencies&lt;/h2&gt;
&lt;p&gt;A website move often shares a DNS zone with the company's email. Save the current records and identify which serve the site, which serve mail, and which verify ownership for other services. A new maintainer needs to know why a record exists before deciding that an unfamiliar entry is obsolete.&lt;/p&gt;
&lt;p&gt;Record domain renewal responsibility, certificate handling, redirects, and any old hostnames that still serve files or forward visitors. The domain is not fully handed over if its renewal notice goes only to an address nobody in the business can read. The &lt;a href="https://gerriscorp.com/services/email-and-dns/"&gt;email and DNS service&lt;/a&gt; covers that layer when it needs hands on work.&lt;/p&gt;

&lt;h2&gt;Ask the receiving person to do one routine job&lt;/h2&gt;
&lt;p&gt;Have the new maintainer make an agreed harmless change on a preview, inspect the result, and explain the release and rollback process. It is a much better test of the documentation than asking whether the folder of instructions looks complete. The exercise also reveals which decisions still depend on knowledge the departing contractor has never written down.&lt;/p&gt;
&lt;p&gt;Record unresolved items, their owners, and the next date to review them. A handover can be accepted with clearly identified followup work, but a missing recovery route should not disappear inside a general statement that everything is finished.&lt;/p&gt;

&lt;h2&gt;Keep the inventory useful after handover&lt;/h2&gt;
&lt;p&gt;Review it when hosting changes, a new supplier joins, or an integration is added. It becomes the starting point for the next contractor and a much calmer starting point for an outage. My &lt;a href="https://gerriscorp.com/case-studies/app-takeover/"&gt;application takeover case&lt;/a&gt; shows the kind of configuration and operational work that an inherited project can require.&lt;/p&gt;
&lt;p&gt;If the documentation is incomplete or nobody knows what is running, &lt;a href="https://gerriscorp.com/services/rescue/"&gt;project rescue and takeover review&lt;/a&gt; is an existing way to hire me for the inventory and a practical next plan. A second opinion before you accept a supplier's handover also fits &lt;a href="https://gerriscorp.com/services/advisory/"&gt;strategic advisory&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/repositories/creating-and-managing-repositories/transferring-a-repository"&gt;GitHub Docs: Transferring a repository&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/repositories/archiving-a-github-repository/backing-up-a-repository"&gt;GitHub Docs: Backing up a repository&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Test an AI pilot with the work it is supposed to do</title>
    <link>https://gerriscorp.com/guides/ai-pilot-tests/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/ai-pilot-tests/</guid>
    <pubDate>Thu, 15 Oct 2026 12:00:00 +0000</pubDate>
    <description>Turn an AI demonstration into a testable pilot: real examples, expected results, permissions, retries, failure records, and a decision your team can explain.</description>
    <content:encoded>&lt;h1&gt;Test an AI pilot with the work it is supposed to do&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-15"&gt;October 15, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A polished demonstration answers the question the demonstrator prepared. Your working day contains the requests they did not prepare: a missing date, two customers with the same name, a task submitted twice, or a report that sounds finished when the underlying action failed. A pilot needs examples of those situations before anyone can decide whether to rely on it.&lt;/p&gt;

&lt;h2&gt;Write one plain job description&lt;/h2&gt;
&lt;p&gt;Begin with a sentence that has an input and an observable result. For example: given an approved meeting note, produce draft tasks with the correct owner and due date. That is narrow enough to test. A promise to make the team more productive is useful as an ambition, but it leaves the test writer guessing which behavior is required.&lt;/p&gt;
&lt;p&gt;Record the tool version, its connected services, and the account permissions used in the trial. Give the pilot an owner who can decide whether an output is acceptable. The person writing the test should not have to infer the answer from whatever the assistant happened to produce.&lt;/p&gt;

&lt;h2&gt;Collect ordinary examples before clever ones&lt;/h2&gt;
&lt;p&gt;Ask the future users for several requests they would really make. Preserve the shorthand, unfinished sentences, and naming conventions that appear in daily work. Add a small number of unusual cases deliberately, rather than building the whole collection from puzzles. The ordinary examples show whether the workflow is useful; the exceptions show where it needs help.&lt;/p&gt;
&lt;p&gt;I learned this with my own terminal assistant. A phrase I had used repeatedly was being silently ignored. Turning the exact request into a test made the problem visible and kept the correction from depending on a conversation I would later forget. The build history is described in the &lt;a href="https://gerriscorp.com/case-studies/built-with-claude-code/"&gt;assistant case study&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;Define the result before running the case&lt;/h2&gt;
&lt;p&gt;For a draft task, the owner must refer to the right person, an absent date must remain absent or trigger a question, and no live task should be created if the request only asks for a draft. For a summary, separate facts that must be present from wording that can vary. This avoids the easy mistake of approving a convincing answer because it resembles the last convincing answer.&lt;/p&gt;
&lt;p&gt;Use the &lt;a href="https://gerriscorp.com/downloads/ai-pilot-test-cases.csv"&gt;AI pilot test case worksheet&lt;/a&gt; as a starting point. Its rows describe invented examples. Fill in the expected behavior, the observed result, and the evidence. Keep real client data in the approved project workspace rather than in a public example file.&lt;/p&gt;

&lt;h2&gt;Include a few failures on purpose&lt;/h2&gt;
&lt;table&gt;
&lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Case&lt;/th&gt;&lt;th scope="col"&gt;Question to settle&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;&lt;th scope="row"&gt;Missing information&lt;/th&gt;&lt;td&gt;Does the assistant ask, preserve the gap, or invent a value?&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;th scope="row"&gt;Ambiguous identity&lt;/th&gt;&lt;td&gt;Can it stop before choosing the wrong customer?&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;th scope="row"&gt;Duplicate request&lt;/th&gt;&lt;td&gt;Does a retry repeat an action already completed?&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;th scope="row"&gt;Unavailable service&lt;/th&gt;&lt;td&gt;Is failure reported without claiming the action succeeded?&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;th scope="row"&gt;Instructions inside source material&lt;/th&gt;&lt;td&gt;Does quoted content stay separate from the user's request?&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;th scope="row"&gt;Unapproved action&lt;/th&gt;&lt;td&gt;Does the workflow wait at the agreed decision point?&lt;/td&gt;&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;Exercise service failures in an isolated trial or with a simulated dependency. Do not create a live outage to see how the pilot reacts. The aim is a reproducible example, not a surprise for the people depending on the service.&lt;/p&gt;

&lt;h2&gt;Test the permission boundary as behavior&lt;/h2&gt;
&lt;p&gt;OWASP's discussion of excessive agency recommends limiting tool capabilities and permissions, with human approval for consequential actions. Translate that principle into a case you can observe. If a workflow is allowed to prepare a message but requires approval to send it, inspect both the draft and the sending record. The assistant saying that it waited is only one part of the evidence.&lt;/p&gt;
&lt;p&gt;A written instruction helps explain the rules. An account that lacks unnecessary permissions helps enforce them. Review both when the pilot connects to another service. A harmless drafting exercise can become a different system when somebody adds a connection that can publish, delete, or spend.&lt;/p&gt;

&lt;h2&gt;Save evidence that outlives the session&lt;/h2&gt;
&lt;p&gt;Keep the inputs, outputs, timestamps, relevant service records, and the acceptance decision together. When a test runs in GitHub Actions, downloadable workflow artifacts can hold its reports. Set retention deliberately and store sensitive evidence only where the team has approved it. A screenshot of the final answer rarely explains a retry, a hidden error, or a partial action.&lt;/p&gt;
&lt;p&gt;My &lt;a href="https://gerriscorp.com/case-studies/data-cleanup/"&gt;outreach tracker audit&lt;/a&gt; is a useful reminder here. The tracker described activity, but checking the actual sending records changed what could honestly be reported. A pilot should make that comparison easier, not bury the trail under a confident paragraph.&lt;/p&gt;

&lt;h2&gt;Score separate questions separately&lt;/h2&gt;
&lt;p&gt;Record correctness, completeness, unauthorized actions, and manual correction effort independently. A mostly useful summary and a message sent to the wrong person should not average into an acceptable score. Decide in advance which failures stop the pilot and which can be tolerated with a documented workaround.&lt;/p&gt;
&lt;p&gt;Repeat cases where results vary and retain the individual outcomes. If only one run was observed, say so. A small collection can reveal a repeatable failure without proving a general success rate. The report should state what was tested and the limits of that evidence.&lt;/p&gt;

&lt;h2&gt;Keep the collection when the pilot ends&lt;/h2&gt;
&lt;p&gt;Use the same examples when a prompt changes, a new model is proposed, or a connector gains another permission. Add real failures as the team encounters them. Remove obsolete cases through an explicit decision so the collection continues to describe the job people are actually doing.&lt;/p&gt;
&lt;p&gt;I can help define and run that review through &lt;a href="https://gerriscorp.com/services/ai-workflow-testing/"&gt;AI workflow testing&lt;/a&gt;. For the earlier buying decision, the &lt;a href="https://gerriscorp.com/guides/ai-vendor-pitch/"&gt;AI vendor questions&lt;/a&gt; help establish what the supplier should demonstrate before a pilot begins.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://genai.owasp.org/llmrisk/llm062025-excessive-agency/"&gt;OWASP Gen AI Security Project: Excessive Agency&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/actions/how-tos/manage-workflow-runs/download-workflow-artifacts"&gt;GitHub Docs: Downloading workflow artifacts&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Diagnosing Core Web Vitals with MCP and an AI coding assistant</title>
    <link>https://gerriscorp.com/guides/core-web-vitals-mcp/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/core-web-vitals-mcp/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>How to diagnose LCP, INP, and CLS with an AI coding assistant through MCP: Chrome DevTools traces, real-user CrUX data, and Screaming Frog at scale.</description>
    <content:encoded>&lt;h1&gt;Diagnosing Core Web Vitals with MCP and an AI coding assistant&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt; · Updated &lt;time datetime="2026-10-15"&gt;October 15, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Fixing a slow page used to mean running a test, reading a waterfall, opening the code, and going back and forth between the three. With the Model Context Protocol, an AI coding assistant such as Claude Code can drive Chrome, record a performance trace, read the result, and open the component responsible, all in one session. Here is how I set it up, what each piece does, and where a person still has to check the work.&lt;/p&gt;
&lt;p&gt;OpenAI Codex and Claude Code are my primary coding assistants for repository work. This guide keeps a Claude Code command as a concrete setup example; choose your assistant and follow its current MCP documentation before connecting browser tools.&lt;/p&gt;

&lt;h2&gt;The three numbers you're chasing&lt;/h2&gt;
&lt;p&gt;Core Web Vitals measure loading, responsiveness, and visual stability. Google's web.dev thresholds for a good experience are a Largest Contentful Paint within 2.5 seconds, an Interaction to Next Paint of 200 milliseconds or less, and a Cumulative Layout Shift of 0.1 or less, measured at the 75th percentile of page loads on mobile and desktop separately. A page passes when it meets all three. The &lt;a href="https://gerriscorp.com/guides/core-web-vitals/"&gt;Core Web Vitals guide&lt;/a&gt; explains each metric and its usual causes; this guide is about the tooling.&lt;/p&gt;

&lt;h2&gt;What MCP adds&lt;/h2&gt;
&lt;p&gt;MCP is a standard way for an AI assistant to call outside tools. An MCP server wraps a tool, such as a browser or a crawler, and the assistant can ask it to do things and read back the results. In Claude Code, you register a server with &lt;code&gt;claude mcp add&lt;/code&gt;, and its tools become available in the session. The assistant still needs you to tell it what to look at and to check its conclusions, but the tedious loop of test, read, and search the code collapses into one conversation.&lt;/p&gt;

&lt;h2&gt;Chrome DevTools MCP: lab traces on demand&lt;/h2&gt;
&lt;p&gt;Google's Chrome DevTools MCP server gives an assistant control of a real Chrome browser and the DevTools performance panel. Its documented setup for Claude Code is a single command:&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;claude mcp add chrome-devtools --scope user npx chrome-devtools-mcp@latest&lt;/code&gt;&lt;/pre&gt;
&lt;p&gt;The tools that matter for Core Web Vitals:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;performance_start_trace&lt;/strong&gt; and &lt;strong&gt;performance_stop_trace&lt;/strong&gt; record a trace, optionally reloading the page and bypassing the cache to measure a cold load. The tool is described as finding frontend performance issues and Core Web Vitals.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;performance_analyze_insight&lt;/strong&gt; opens a specific finding from the trace, such as the LCP breakdown, which shows where the time before the largest element appeared actually went.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;emulate&lt;/strong&gt; throttles the CPU and network, with presets from Slow 3G to Fast 4G, and sets a mobile viewport, so a fast desktop doesn't hide a slow phone experience.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;list_console_messages&lt;/strong&gt; and &lt;strong&gt;list_network_requests&lt;/strong&gt; show errors and the heaviest resources.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;lighthouse_audit&lt;/strong&gt; runs Lighthouse for accessibility, SEO, and best practices. Its documentation notes that it excludes performance, which is what the trace tools are for.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Two settings are worth knowing. The performance tools can send the traced URL to Google's CrUX API to show real-user data beside the lab result; the &lt;code&gt;--no-performance-crux&lt;/code&gt; flag turns that off. And Google collects usage statistics from the server by default, which &lt;code&gt;--no-usage-statistics&lt;/code&gt; disables.&lt;/p&gt;

&lt;h2&gt;Real-user data: CrUX, and a change to PageSpeed Insights&lt;/h2&gt;
&lt;p&gt;A lab trace shows one load on one machine. Search uses field data from real Chrome users, collected in the Chrome UX Report. The CrUX API returns that data for a page or a whole origin, with a Google Cloud API key set up for it.&lt;/p&gt;
&lt;p&gt;One change to plan around: Google's PageSpeed Insights API documentation says Google plans to stop including CrUX real-user data in that API, and recommends the CrUX API or the CrUX History API instead. Any script or dashboard that pulls field data from PageSpeed Insights should move now.&lt;/p&gt;

&lt;h2&gt;Screaming Frog: every template, not one page&lt;/h2&gt;
&lt;p&gt;Traces are for depth. For breadth, Screaming Frog's PageSpeed Insights integration, under Configuration &amp;gt; API Access &amp;gt; PageSpeed Insights, pulls Lighthouse metrics and opportunities for every URL during a crawl. It can also run Lighthouse locally, which avoids the API's daily query limit but has no CrUX data.&lt;/p&gt;
&lt;p&gt;Screaming Frog version 24 also runs as an MCP server, so an assistant can start crawls and read reports and bulk exports directly. It's a paid licence feature and needs database storage mode. Optional tools let the assistant run Node.js scripts, and they're off until you enable them in the MCP server settings.&lt;/p&gt;
&lt;p&gt;The combination is where the time savings live: crawl every URL, group the slow ones by template, trace one representative page per template, and fix the template once.&lt;/p&gt;

&lt;h2&gt;A working session, step by step&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Start from field data.&lt;/strong&gt; Check the Core Web Vitals report in Search Console or the CrUX API to see which metric fails and on which group of pages.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Pick representative URLs,&lt;/strong&gt; one per failing template.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Trace each on emulated mobile.&lt;/strong&gt; Ask the assistant to throttle to a mid-range phone and record a cold load.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Read the insight, then the code.&lt;/strong&gt; For a slow LCP, the breakdown shows whether the time goes to the server, to fetching the hero image, or to rendering. With the repository open, the assistant can find the component that loads the image and propose a change, such as preloading it or removing its lazy loading.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Test on a preview.&lt;/strong&gt; Trace the same page on the branch's preview deployment and compare.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Confirm in the field.&lt;/strong&gt; CrUX data is a rolling 28-day window, so real-user numbers improve gradually after release.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;Where to keep a hand on the wheel&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Lab is not field.&lt;/strong&gt; A trace on a developer's machine, even throttled, is one sample. Treat it as a diagnosis and judge the outcome by field data.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Check every proposed fix.&lt;/strong&gt; An assistant can suggest deferring a script that the page needs, or a preload that competes with the real LCP image. Measure before and after.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Mind what the browser can see.&lt;/strong&gt; The DevTools server's own documentation warns that it exposes the browser's content to the assistant. Use a separate browser profile, never one signed in to anything sensitive.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;INP needs interaction.&lt;/strong&gt; A page load trace says little about responsiveness. Have the assistant click, type, and open menus during a trace, and compare with field INP.&lt;/li&gt;
&lt;/ul&gt;

&lt;p class="related"&gt;This is how I run speed work on code based sites as part of &lt;a href="https://gerriscorp.com/services/site-speed/"&gt;site speed and Core Web Vitals&lt;/a&gt; and &lt;a href="https://gerriscorp.com/services/seo-engineering/"&gt;SEO engineering&lt;/a&gt;. For the habits that keep an AI assistant safe in a client codebase, see &lt;a href="https://gerriscorp.com/guides/ai-coding-assistants/"&gt;using AI coding assistants on a business website&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://github.com/ChromeDevTools/chrome-devtools-mcp"&gt;Chrome DevTools MCP (README)&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://github.com/ChromeDevTools/chrome-devtools-mcp/blob/main/docs/tool-reference.md"&gt;Chrome DevTools MCP tool reference&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://code.claude.com/docs/en/mcp"&gt;Anthropic: Connect Claude Code to tools via MCP&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://web.dev/articles/vitals"&gt;web.dev: Web Vitals&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/speed/docs/insights/v5/get-started"&gt;Google: PageSpeed Insights API&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developer.chrome.com/docs/crux/api"&gt;Chrome for Developers: CrUX API&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.screamingfrog.co.uk/seo-spider/user-guide/configuration/"&gt;Screaming Frog: SEO Spider configuration (MCP server and PageSpeed Insights)&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Repositories plus super intelligence: the SEO superpower nobody told you to learn</title>
    <link>https://gerriscorp.com/guides/seo-superpower/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/seo-superpower/</guid>
    <pubDate>Sun, 11 Oct 2026 12:00:00 +0000</pubDate>
    <description>Repositories and AI coding assistants help SEOs ship fixes. How Git, OpenAI Codex, Claude Code, and GitHub Copilot connect search findings to changes in code.</description>
    <content:encoded>&lt;h1&gt;Repositories plus super intelligence: the SEO superpower nobody told you to learn&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-11"&gt;October 11, 2026&lt;/time&gt; · Updated &lt;time datetime="2026-10-15"&gt;October 15, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;On September 29, 2026, President Trump signed an executive order replacing "Artificial Intelligence" with "Super Intelligence" across the executive branch. Call it AI, SI, or a very fast intern; for anyone in SEO, the name matters less than what it does when you pair it with the other half of the equation: the code repository. Git plus a capable coding assistant is a real superpower. It takes you far outside the confines of being a Yoast monkey or a Screaming Frog jockey.&lt;/p&gt;

&lt;h2&gt;Half one: the repository&lt;/h2&gt;
&lt;p&gt;Every serious website now lives in version control, almost always Git, which Linus Torvalds and the Linux kernel community built in 2005 after losing free use of their previous tool. Before Git, teams used CVS and then Subversion. Today the code sits on GitHub, GitLab, or Bitbucket, and every change arrives as a commit, gets proposed as a pull request, is reviewed, and can be rolled back.&lt;/p&gt;
&lt;p&gt;For SEO, that means the things you care about are files. Title patterns, canonical logic, hreflang, structured data, redirects, robots rules, sitemaps, internal linking modules, and lazy loading behavior are all written down somewhere in the repository. If you can't read it, you're guessing about it from the outside.&lt;/p&gt;

&lt;h2&gt;Half two: super intelligence that reads code&lt;/h2&gt;
&lt;p&gt;Coding assistants have moved past autocomplete. GitHub says Copilot can now answer questions about a repository, plan and make changes, run tests in a cloud environment, and review pull requests. Anthropic's Claude Code does the same from a terminal, and OpenAI Codex helps write, review, and debug code. Point one at a codebase and ask where the canonical tag comes from, and it will find the component, the function feeding it, and the edge case that breaks it, in minutes.&lt;/p&gt;

&lt;h2&gt;Why each half fails alone&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;A repository without AI&lt;/strong&gt; is a wall of unfamiliar code. Most SEOs never get past the folder list, so they file a ticket and wait.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;AI without a repository&lt;/strong&gt; is copy and paste from a chat window: snippets with no history, no review, and no way to know whether the fix ever shipped or what else it broke.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Together,&lt;/strong&gt; the assistant reads and drafts, Git records and protects, and the SEO supplies the one thing neither has: knowing what search engines and AI answer engines actually need.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;What the superpower looks like in practice&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Diagnosis down to the line.&lt;/strong&gt; A crawl tells you product pages render 300 words. The repository tells you a scroll triggered component never mounts for a crawler. I found exactly that once, and it's the difference between "Google doesn't like the page" and a fix a developer can ship the same day. The details are in my &lt;a href="https://gerriscorp.com/guides/javascript-seo/"&gt;JavaScript SEO guide&lt;/a&gt;.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Migrations with receipts.&lt;/strong&gt; Redirect maps written as config, tested against the old URL list before launch, and reviewed like any other code.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Fixes at template scale.&lt;/strong&gt; One pull request repairs every page built from that template, today and in the future.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Your own tools.&lt;/strong&gt; I direct OpenAI Codex and Claude Code like a small development team; Claude Code built me a command line assistant with a full test suite, moved a legacy blog off dead software, and built this website. &lt;a href="https://gerriscorp.com/case-studies/built-with-claude-code/"&gt;See my Codex and Claude Code workflow.&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;How to start, if you're an SEO&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Learn six Git words:&lt;/strong&gt; clone, branch, commit, diff, pull request, merge. That's enough to be useful and safe.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Ask for read access&lt;/strong&gt; to a client's repository. Read-only is an easy yes for most teams.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Have the assistant give you the tour.&lt;/strong&gt; Ask it to explain the structure, where templates live, and where every SEO element is generated.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Make your first pull request tiny,&lt;/strong&gt; such as a title pattern or a missing canonical, and let the developer review it.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Write the rules down&lt;/strong&gt; in a project memory file and turn them into automated checks, so the assistant follows them every time.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Check everything.&lt;/strong&gt; The assistant is confidently wrong often enough that verification is the job. My method for doing this safely is in &lt;a href="https://gerriscorp.com/guides/ai-coding-assistants/"&gt;how to use AI coding assistants on a business website&lt;/a&gt;.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;What this means for your business&lt;/h2&gt;
&lt;p&gt;If you run a company with a website, the question to ask your SEO is no longer whether they know the plugin. It's whether they can read the code and send a pull request. An SEO with both halves of the superpower costs you fewer developer hours, finds the cause, and leaves a reviewed history of every change.&lt;/p&gt;

&lt;h2&gt;Hire the superpower&lt;/h2&gt;
&lt;p&gt;This is how I work every day: in the repository, through pull requests, with super intelligence doing the heavy reading under my direction and my name on the result. I can be your &lt;a href="https://gerriscorp.com/services/technical-lead/"&gt;fractional technical lead&lt;/a&gt;, turn a diagnosis into &lt;a href="https://gerriscorp.com/services/developers/"&gt;developer ready specs&lt;/a&gt; and verify they shipped, or &lt;a href="https://gerriscorp.com/services/ai-automation/"&gt;build you small AI tools safely&lt;/a&gt;. &lt;a href="https://gerriscorp.com/contact/"&gt;Tell me about your site&lt;/a&gt; and I'll tell you what the code is doing to your search visibility.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.openai.com/api/docs/guides/code-generation"&gt;OpenAI: Codex for code generation&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.whitehouse.gov/fact-sheets/2026/09/fact-sheet-president-donald-j-trump-inaugurates-the-era-of-super-intelligence/"&gt;The White House: Fact Sheet: President Donald J. Trump Inaugurates The Era of Super Intelligence (September 29, 2026)&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://git-scm.com/book/en/v2/Getting-Started-A-Short-History-of-Git"&gt;Git: A short history of Git&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://code.claude.com/docs/en/overview"&gt;Anthropic: Claude Code overview&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/copilot/concepts/copilot-surfaces/copilot-on-github"&gt;GitHub Docs: GitHub Copilot on GitHub.com&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Build a podcast website with Codex or Claude Code, GitHub Pages, and Cloudflare</title>
    <link>https://gerriscorp.com/guides/podcast-website/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/podcast-website/</guid>
    <pubDate>Thu, 08 Oct 2026 12:00:00 +0000</pubDate>
    <description>Build a podcast website with OpenAI Codex or Claude Code, GitHub Pages, and Cloudflare: a page for every episode, stable links, and daily feed updates.</description>
    <content:encoded>&lt;h1&gt;Build a podcast website with Codex or Claude Code, GitHub Pages, and Cloudflare&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-08"&gt;October 8, 2026&lt;/time&gt; · Updated &lt;time datetime="2026-10-15"&gt;October 15, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A podcast that lives only on Spotify and Apple is a tenant. The platforms own the pages, the search results, and the rules, and a platform can close a door overnight, the way Google Podcasts did in 2024. A website on your own domain is the one property you control. Here is the setup I build: it costs the price of a domain, it updates itself, and Codex or Claude Code does the typing.&lt;/p&gt;

&lt;h2&gt;What you end up with&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;One page per episode, at an address that never changes, with the cover art, the player, the show notes, and the transcript.&lt;/li&gt;
  &lt;li&gt;A home page that features the newest episode, an archive of every episode, and an about page with every place to subscribe.&lt;/li&gt;
  &lt;li&gt;Structured data, a sitemap, and descriptions written for search engines and AI assistants.&lt;/li&gt;
  &lt;li&gt;A daily job that reads your podcast feed and adds new episodes on its own.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;The audio stays with your host. Each player streams from the address in your feed, so the host still counts every play and your download numbers stay whole.&lt;/p&gt;

&lt;h2&gt;Step 1: register the domain at cost&lt;/h2&gt;
&lt;p&gt;Cloudflare Registrar sells domains at the wholesale price the registry charges, plus the ICANN fee, with no markup on renewals. Privacy for the registration record comes included. A .com runs about ten dollars a year. The catch is that the domain must use Cloudflare's DNS, which suits this setup anyway. Pick a name that matches the show exactly, and register it before you announce anything.&lt;/p&gt;

&lt;h2&gt;Step 2: have Codex or Claude Code build the generator&lt;/h2&gt;
&lt;p&gt;OpenAI Codex and Claude Code are coding assistants I use to work directly with project files. Describe the site to either assistant, then have it draft the code, run the build, and work through the errors under your review. The prompt that matters most names your feed and your rules. A good one reads close to this:&lt;/p&gt;
&lt;blockquote&gt;&lt;p&gt;Build a static site from this podcast RSS feed, one page per episode, in Python. Keep each episode's address forever once it's assigned. Cache the cover art and transcripts in the repository, never the audio. Use the feed's audio URL exactly as given. Give every page a unique title near 60 characters and a description near 155. Add PodcastEpisode structured data and a sitemap.&lt;/p&gt;&lt;/blockquote&gt;
&lt;p&gt;Then review what it builds the way you'd review a contractor's work. Open a few episode pages, read the titles, click the players, and ask for changes in plain English. The first full build of a show with a few hundred episodes takes an afternoon, most of it spent downloading artwork.&lt;/p&gt;

&lt;h2&gt;Step 3: publish on GitHub Pages&lt;/h2&gt;
&lt;p&gt;GitHub Pages hosts static sites free from a public repository. Published sites can be up to 1 GB, with a soft limit of 100 GB of traffic a month, which covers a podcast site comfortably because the audio never touches it. Put a file named &lt;code&gt;CNAME&lt;/code&gt; containing your domain in the repository, and turn on Pages for the main branch.&lt;/p&gt;

&lt;h2&gt;Step 4: point the domain, then turn on HTTPS&lt;/h2&gt;
&lt;p&gt;In Cloudflare's DNS, add the four GitHub Pages A records and four AAAA records for the bare domain, and a CNAME for www pointing at your github.io address. Leave every record set to DNS only, the gray cloud, while GitHub issues the certificate. When the "Enforce HTTPS" box in the repository's Pages settings becomes clickable, tick it. If the certificate never starts, remove the custom domain in the Pages settings and add it again; GitHub skips the request when the domain was saved before its DNS existed.&lt;/p&gt;

&lt;h2&gt;Step 5: make it run itself&lt;/h2&gt;
&lt;p&gt;Ask your coding assistant for a GitHub Actions workflow that runs the generator every morning, commits only when something changed, and fails loudly when the feed can't be read. Three details make it dependable:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Keepalive.&lt;/strong&gt; GitHub disables scheduled workflows in public repositories after 60 days without activity. A step that re-enables the workflow through the GitHub API on every run keeps the schedule alive through a quiet season.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Late transcripts.&lt;/strong&gt; Hosts often add transcripts days after an episode goes up. Recheck every episode that lacks one, on every run.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;A sanity check.&lt;/strong&gt; If the feed suddenly lists far fewer episodes than the site has, stop and publish nothing.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Step 6: tell the search engines&lt;/h2&gt;
&lt;p&gt;Add the site to Google Search Console and Bing Webmaster Tools and submit the sitemap. Wire in IndexNow so Bing and the other engines that share it hear about each new episode the morning it appears. Put the site's address in your podcast host's settings, too, so Apple and Spotify link listeners back to it.&lt;/p&gt;

&lt;h2&gt;What it costs&lt;/h2&gt;
&lt;p&gt;The domain is the recurring site cost in this setup; hosting, HTTPS, and the daily job fit within the free allowances described above. Budget separately for whichever coding assistant you use. Check its current plan and usage limits before starting.&lt;/p&gt;

&lt;h2&gt;Where AI saves the most time&lt;/h2&gt;
&lt;p&gt;The generator itself is routine code. The slow parts are the judgment calls: a title that reads well at 60 characters, a description for an episode whose notes are two words, alt text for hundreds of cover images. An assistant drafts those in minutes, and you spend your time reviewing instead of typing. For the working method, specs, review, and keeping secrets out of the repository, see &lt;a href="https://gerriscorp.com/guides/ai-coding-assistants/"&gt;how to use AI coding assistants on a website safely&lt;/a&gt;. For the DNS side, see &lt;a href="https://gerriscorp.com/guides/cloudflare-seo/"&gt;Cloudflare settings that help or hurt SEO&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Want this built for your show, or for any feed you publish? &lt;a href="https://gerriscorp.com/contact/"&gt;Tell me about it&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.openai.com/api/docs/guides/code-generation"&gt;OpenAI: Codex for code generation&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/pages/getting-started-with-github-pages/github-pages-limits"&gt;GitHub Docs: GitHub Pages limits&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/pages/configuring-a-custom-domain-for-your-github-pages-site/managing-a-custom-domain-for-your-github-pages-site"&gt;GitHub Docs: Managing a custom domain for your GitHub Pages site&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.cloudflare.com/registrar/"&gt;Cloudflare Docs: Cloudflare Registrar&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/actions/managing-workflow-runs-and-deployments/managing-workflow-runs/disabling-and-enabling-a-workflow"&gt;GitHub Docs: Disabling and enabling a workflow&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.anthropic.com/en/docs/claude-code/overview"&gt;Anthropic: Claude Code overview&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Podcast SEO: get every episode found in Google and AI search</title>
    <link>https://gerriscorp.com/guides/podcast-seo/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/podcast-seo/</guid>
    <pubDate>Thu, 08 Oct 2026 12:00:00 +0000</pubDate>
    <description>How to make podcast episodes findable in Google, Bing, and AI answers: episode pages on your own domain, transcripts in the HTML, schema, and alt text.</description>
    <content:encoded>&lt;h1&gt;Podcast SEO: get every episode found in Google and AI search&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-08"&gt;October 8, 2026&lt;/time&gt; · Updated &lt;time datetime="2026-10-15"&gt;October 15, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Search engines index text. A podcast is audio wrapped in a feed that podcast apps read, and the pages that platforms generate for each episode are thin, interchangeable, and owned by someone else. If you want people who search a topic you covered to find the episode where you covered it, each episode needs a real page of text on a domain you control.&lt;/p&gt;

&lt;h2&gt;Give every episode its own page&lt;/h2&gt;
&lt;p&gt;One episode, one address, forever. The page carries the title, the date, the running time, the cover art, a player, the full show notes, and links to listen in the major apps. Keep the address stable even if you later edit the title on your host, because every link and every search ranking points to that address.&lt;/p&gt;
&lt;p&gt;The player can stream straight from the audio file listed in your feed. The host keeps counting downloads, and your page keeps the search value.&lt;/p&gt;

&lt;h2&gt;Put the transcript in the HTML&lt;/h2&gt;
&lt;p&gt;The transcript is the single biggest SEO asset a podcast has: thousands of words, in your own phrasing, on exactly the topics you know. Most hosts now generate transcripts automatically. Publish the text on the episode page as real HTML. A collapsible section is fine, since the text is still in the page, and it keeps the player and notes near the top. A transcript hidden inside a player widget or a downloadable file is invisible to search.&lt;/p&gt;

&lt;h2&gt;Write titles and descriptions for searchers&lt;/h2&gt;
&lt;p&gt;Episode titles written for loyal listeners, like inside jokes or bare episode numbers, mean something only to regulars. Each page needs a title tag of about 60 characters that says what the episode covers, and a meta description of about 150 that a searcher would click. The on-page heading can stay playful. Write a unique pair for every episode; duplicate titles across hundreds of pages are a common reason an archive underperforms.&lt;/p&gt;

&lt;h2&gt;Mark it up with schema&lt;/h2&gt;
&lt;p&gt;Schema.org has types built for this. Mark each page as a &lt;code&gt;PodcastEpisode&lt;/code&gt; with its name, date, duration, description, and image, link the audio as an &lt;code&gt;AudioObject&lt;/code&gt;, and connect every episode to one &lt;code&gt;PodcastSeries&lt;/code&gt; on the home page with links to the show on each platform. Season and episode numbers have their own properties too. This is how search engines and AI systems learn that hundreds of pages belong to one show by one host.&lt;/p&gt;

&lt;h2&gt;Describe your cover art&lt;/h2&gt;
&lt;p&gt;Custom covers per episode help in image search and social shares, but only if the alt text says what each image shows. Describe the picture in a sentence, quote any words that appear in it, and skip the "image of" lead-in. Screen reader users and text-only browsers get a caption; search engines get context.&lt;/p&gt;

&lt;h2&gt;Get indexed quickly&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Sitemap:&lt;/strong&gt; list every episode page with its date, and submit it in Google Search Console and Bing Webmaster Tools.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;IndexNow:&lt;/strong&gt; ping Bing and the other participating engines the moment a new episode page goes live.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Internal links:&lt;/strong&gt; link each episode to the previous and next ones, and keep a full archive page, so crawlers can reach the whole catalog.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Feed link:&lt;/strong&gt; put your site's address in your host's show settings, so the apps link listeners back to it.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Make it readable by AI assistants&lt;/h2&gt;
&lt;p&gt;People increasingly ask ChatGPT, Claude, Perplexity, or Google's AI Mode what a podcast said about something. Those systems quote pages they can read. Allow their crawlers in robots.txt, keep the text in plain HTML, and consider an &lt;code&gt;llms.txt&lt;/code&gt; file that lists every episode with a link. See &lt;a href="https://gerriscorp.com/guides/ai-assistants/"&gt;how AI assistants find and describe a business&lt;/a&gt; for the wider picture.&lt;/p&gt;

&lt;h2&gt;Own the property&lt;/h2&gt;
&lt;p&gt;Directories close. When Google Podcasts went dark in 2024, every show listed there lost that doorway at once. A site on your own domain outlasts every directory, collects the search value of every episode, and keeps working while you sleep. It can also update itself: a small script that reads your feed each morning adds new episodes and late transcripts without you touching it. Here is &lt;a href="https://gerriscorp.com/guides/podcast-website/"&gt;how to build one with Codex or Claude Code, GitHub Pages, and Cloudflare&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Starting from scratch? Read &lt;a href="https://gerriscorp.com/guides/podcast-for-your-brand/"&gt;what it really takes to start a podcast for your brand&lt;/a&gt;, or &lt;a href="https://gerriscorp.com/contact/"&gt;ask me to look at your show's search visibility&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://schema.org/PodcastEpisode"&gt;Schema.org: PodcastEpisode&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/google-images"&gt;Google Search Central: Image SEO best practices&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/sitemaps/build-sitemap"&gt;Google Search Central: Build and submit a sitemap&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.indexnow.org/"&gt;IndexNow&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Starting a podcast to build your brand: what it really takes</title>
    <link>https://gerriscorp.com/guides/podcast-for-your-brand/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/podcast-for-your-brand/</guid>
    <pubDate>Thu, 08 Oct 2026 12:00:00 +0000</pubDate>
    <description>What a podcast does for a personal or business brand, the commitment behind it, and the cheap gear and AI tools that take the friction out of every episode.</description>
    <content:encoded>&lt;h1&gt;Starting a podcast to build your brand: what it really takes&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-08"&gt;October 8, 2026&lt;/time&gt; · Updated &lt;time datetime="2026-10-15"&gt;October 15, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A podcast puts a founder's or an expert's actual voice in someone's ears for half an hour, by choice, while they drive or walk the dog. Few marketing channels earn that much attention. I recommend one to every client with something to say, along with a warning: the hard part is showing up, week after week, for a year.&lt;/p&gt;

&lt;h2&gt;What a podcast does for a brand&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Trust at scale.&lt;/strong&gt; Listeners hear how you think, how you handle a tangent, and what makes you laugh. By the time they call, they already know you.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;A searchable record of expertise.&lt;/strong&gt; Every episode, with its notes and transcript, becomes evidence that you know your subject, for prospects, for search engines, and for the AI assistants people now ask about you.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Raw material.&lt;/strong&gt; One recording becomes show notes, an article, social clips, a newsletter item, and a transcript full of quotable lines.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;The real cost is commitment&lt;/h2&gt;
&lt;p&gt;A podcast rewards consistency long before it rewards polish. Pick a rhythm you can keep for a year, weekly or every other week, and protect it on the calendar like a client meeting. Many shows stop after a handful of episodes, so simply continuing puts you ahead of a crowd. Growth is slow, and the archive is the asset: the fiftieth episode makes the first forty-nine easier to find.&lt;/p&gt;

&lt;h2&gt;A pocket recorder or a phone is enough&lt;/h2&gt;
&lt;p&gt;A small digital voice recorder, the kind &lt;a href="https://www.amazon.com/dp/B082QL6KLG?tag=gerris-20" rel="sponsored noopener"&gt;Sony sells&lt;/a&gt; for less than a nice dinner (I record on a Sony ICD-UX570), records clean audio, fits in a jacket pocket, and charges from a plain USB port. Buy two or three and keep one in every bag, so an idea never waits for the studio. A smartphone held close in a quiet room works just as well. A carpeted bedroom beats an echoing kitchen with an expensive microphone every time.&lt;/p&gt;
&lt;p&gt;Noise used to be the reason people bought gear. Adobe Podcast's Enhance Speech, which has a free version, now takes a noisy recording and brings it close to studio quality in a minute or two. Audacity, free on every computer, handles the edit: trim the false starts and the long pauses, and leave the personality in.&lt;/p&gt;

&lt;h2&gt;Hosting and distribution are free&lt;/h2&gt;
&lt;p&gt;A podcast host stores the audio and publishes the RSS feed that every app reads. Spotify for Creators, formerly Anchor, hosts shows free, publishes them on Spotify, and gives you the feed to list on Apple Podcasts and every other directory. Record on whatever you carry, upload the file, write the notes, schedule it, and every app picks it up.&lt;/p&gt;

&lt;h2&gt;YouTube is a separate channel&lt;/h2&gt;
&lt;p&gt;YouTube can import an audio podcast from its RSS feed, but it turns each episode into a video over a still image, and those videos live only on YouTube and YouTube Music. YouTube audiences expect faces, cuts, and visuals, and they discover shows through YouTube's own recommendations. Treat it as its own channel with its own effort, and keep publishing the audio feed everywhere else.&lt;/p&gt;

&lt;h2&gt;Spend money on the parts that repeat&lt;/h2&gt;
&lt;p&gt;An intro, an outro, and the show's main cover art appear on every episode, so they repay a professional. Freelancers on Fiverr produce all three for a modest fee. For individual episodes, ChatGPT turns a title and a mood into a custom cover in minutes, and a distinct cover per episode makes each one easier to spot in a feed and easier to share.&lt;/p&gt;

&lt;h2&gt;Let AI clear away the chores&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Cleanup:&lt;/strong&gt; AI enhancement rescues recordings made in imperfect rooms.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Transcripts:&lt;/strong&gt; hosts and apps generate them automatically. Publish them, because they make every episode searchable.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Show notes:&lt;/strong&gt; paste the transcript into an assistant, ask for a summary, chapter points, and a title, then edit for your voice.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Cover art:&lt;/strong&gt; an image model gives every episode its own look.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;New formats:&lt;/strong&gt; Google's NotebookLM turns notes or an episode into a two-host audio discussion, an easy bonus episode for listeners who enjoy that style.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Your opinions stay yours. AI removes the busywork that makes people skip a week, and skipped weeks are what end shows.&lt;/p&gt;

&lt;h2&gt;Own a home for it&lt;/h2&gt;
&lt;p&gt;Your host's pages belong to the host. Give the show its own website on its own domain, with a page for every episode, and the archive works for your brand in search for years. Read &lt;a href="https://gerriscorp.com/guides/podcast-seo/"&gt;how to get every episode found in Google and AI search&lt;/a&gt;, and &lt;a href="https://gerriscorp.com/guides/podcast-website/"&gt;how to build the site with Codex or Claude Code and GitHub Pages&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;Want help turning a show into search visibility for your brand? &lt;a href="https://gerriscorp.com/contact/"&gt;Tell me about it&lt;/a&gt;.&lt;/p&gt;
&lt;p&gt;As an Amazon Associate I earn from qualifying purchases.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://creators.spotify.com/"&gt;Spotify for Creators&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://podcast.adobe.com/enhance"&gt;Adobe Podcast: Enhance Speech&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.audacityteam.org/"&gt;Audacity&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/youtube/answer/13525207"&gt;YouTube Help: Deliver podcasts using an RSS feed&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://notebooklm.google.com/"&gt;Google NotebookLM&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to use AI coding assistants on a business website safely</title>
    <link>https://gerriscorp.com/guides/ai-coding-assistants/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/ai-coding-assistants/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>A practical workflow for OpenAI Codex, Claude Code, and GitHub Copilot on business websites: written specs, project context, automated checks, and human review.</description>
    <content:encoded>&lt;h1&gt;How to use AI coding assistants on a business website safely&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt; · Updated &lt;time datetime="2026-10-15"&gt;October 15, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;AI coding assistants such as OpenAI Codex, Claude Code, GitHub Copilot, and Cursor can now do real work on real websites: migrations, redirect maps, templates, structured data, content checks across hundreds of pages. They work fast and they work confidently, including when they're wrong. The method below is how I get the speed without the damage.&lt;/p&gt;

&lt;h2&gt;Start with a written spec&lt;/h2&gt;
&lt;p&gt;An assistant does exactly what it understands you to want, at full speed. Before any code, write down the goal, the constraints, and what done looks like. "Move the blog to the new platform" is a wish. A spec says: every old URL keeps working or redirects to its closest match, titles and dates carry over, images are copied and not hotlinked, the build fails if any internal link points to a missing page, and nothing ships until the redirect list has been checked. The assistant can help draft this, and a person decides it.&lt;/p&gt;

&lt;h2&gt;Give the project a memory&lt;/h2&gt;
&lt;p&gt;Claude Code reads a &lt;code&gt;CLAUDE.md&lt;/code&gt; file at the start of every session, and other tools have their own equivalents. Put the things that must never be forgotten there: the build command, the deploy process, the voice rules for copy, the files that must never be committed, the decisions already made and why. This site's project file holds its style rules and a list of names that must never appear on the site. The assistant reads it every time; a new contractor would need the same briefing.&lt;/p&gt;

&lt;h2&gt;Turn the rules into checks&lt;/h2&gt;
&lt;p&gt;Instructions get forgotten. Automated checks don't. Every rule that can be tested by a script should be. The build for this site refuses to finish if a page title falls outside 50 to 60 characters, if a meta description is the wrong length, if two pages share a title, if any page links to one that doesn't exist, or if two pages share too much of their wording. When the assistant writes a new page, the build tells it what's wrong before I ever see the draft. The same approach works for any site: link checkers, HTML validators, accessibility tests, and unit tests for anything with logic in it.&lt;/p&gt;

&lt;h2&gt;Small changes, reviewed&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Use version control.&lt;/strong&gt; Every change goes through Git, on a branch, so anything can be undone and every change can be read as a diff.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Keep each change small enough to read.&lt;/strong&gt; A change that touches three files gets reviewed properly. One that touches three hundred gets skimmed.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Read the diff, not the summary.&lt;/strong&gt; An assistant's description of what it changed is a claim. The diff is the evidence.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Watch for scope creep.&lt;/strong&gt; Assistants like to tidy things nobody asked them to touch. Reject unrelated changes and ask again.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Keep secrets and production out of reach&lt;/h2&gt;
&lt;p&gt;Never paste passwords, API keys, or customer data into a conversation, and never commit them to a repository. Give the assistant access to a copy of the site, a staging environment, or a local build, and keep deploys to production a deliberate human step, or an automated pipeline that runs only after the checks pass. For anything destructive, such as deleting files, rewriting history, or changing DNS, require explicit confirmation every time. Inventory the DNS records before touching any of them; a careless change can stop a company's email.&lt;/p&gt;

&lt;h2&gt;Verify against the real world&lt;/h2&gt;
&lt;p&gt;Assistants are well read and out of date in places. They'll describe a platform setting that moved last year, or a search feature Google retired, with total confidence. For anything that depends on a third party's current behavior, check the vendor's documentation or test it directly. After deploying, check the live result: fetch the pages, test the redirects with &lt;code&gt;curl&lt;/code&gt;, run PageSpeed Insights, and look at the page in a browser. "The code looks right" isn't the same as "the site works."&lt;/p&gt;

&lt;h2&gt;Where assistants are strongest&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Bulk, rule-based work.&lt;/strong&gt; Building a redirect map from hundreds of old URLs, rewriting templates across a site, generating structured data from page content, converting content between platforms.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Audits at scale.&lt;/strong&gt; Checking every page for missing titles, broken links, duplicate descriptions, or images without alt text, then fixing them.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Small internal tools.&lt;/strong&gt; Scripts that pull reports, clean exports, or automate a weekly chore.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Explaining unfamiliar code.&lt;/strong&gt; Reading an inherited theme or plugin and describing what it actually does.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Where to slow down&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Judgment calls.&lt;/strong&gt; Which pages to merge, which claims a page should make, what a client would or wouldn't want public. The assistant can propose; a person decides.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Security and payments.&lt;/strong&gt; Authentication, checkout, and anything handling personal data needs an expert review regardless of who wrote it.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Long sessions.&lt;/strong&gt; As a conversation grows, early instructions carry less weight. Restate the important rules, or keep them in the project memory file where they're read fresh.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Someone has to be accountable&lt;/h2&gt;
&lt;p&gt;The assistant writes code; a person owns the outcome. For a business site that means someone who can read the diff, understands search and infrastructure, and will notice when a confident change is wrong. That's the role I play for clients: I write the spec, set up the checks, direct the assistant or the developers, and verify what ships.&lt;/p&gt;

&lt;p class="related"&gt;See how this site and other projects were built in the &lt;a href="https://gerriscorp.com/case-studies/built-with-claude-code/"&gt;Codex and Claude Code case study&lt;/a&gt;, and what I build for clients under &lt;a href="https://gerriscorp.com/services/ai-automation/"&gt;AI tools and automation&lt;/a&gt; and &lt;a href="https://gerriscorp.com/services/technical-lead/"&gt;fractional technical lead&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.openai.com/api/docs/guides/code-generation"&gt;OpenAI: Codex for code generation&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.anthropic.com/en/docs/claude-code/overview"&gt;Anthropic: Claude Code overview&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Glossary of SEO, AI search, and technical website terms</title>
    <link>https://gerriscorp.com/guides/glossary/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/glossary/</guid>
    <pubDate>Mon, 05 Oct 2026 12:00:00 +0000</pubDate>
    <description>Plain definitions of technical SEO, AI search, site speed, and email terms: AEO, GEO, canonical tags, Core Web Vitals, DMARC, IndexNow, llms.txt, and many more.</description>
    <content:encoded>&lt;h1&gt;Glossary of SEO, AI search, and technical website terms&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-05"&gt;October 5, 2026&lt;/time&gt; · Updated &lt;time datetime="2026-10-15"&gt;October 15, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Plain definitions of the terms that come up in my work, written so each one makes sense on its own. The &lt;a href="https://gerriscorp.com/site-index/"&gt;index&lt;/a&gt; lists every page where a term comes up.&lt;/p&gt;

&lt;dl class="glossary"&gt;
  &lt;dt id="301-redirect"&gt;301 redirect&lt;/dt&gt;
  &lt;dd&gt;A permanent redirect from one URL to another. Search engines transfer the old URL's ranking signals to the new one. The 308 status code is also permanent; 302 and 307 are temporary.&lt;/dd&gt;
  &lt;dt id="acceptance-criteria"&gt;Acceptance criteria&lt;/dt&gt;
  &lt;dd&gt;The plain statements a finished change has to satisfy before anyone calls it done, written before the work starts. They turn a request into something a developer can build and a reviewer can check.&lt;/dd&gt;
  &lt;dt id="aeo"&gt;AEO (answer engine optimization)&lt;/dt&gt;
  &lt;dd&gt;Shaping content so it directly answers the questions people ask, in a form that search features and AI assistants can quote.&lt;/dd&gt;
  &lt;dt id="ai-mode"&gt;AI Mode&lt;/dt&gt;
  &lt;dd&gt;Google Search's conversational, AI-generated search experience, which answers questions with a written response and links, built on Google's search index.&lt;/dd&gt;
  &lt;dt id="ai-overviews"&gt;AI Overviews&lt;/dt&gt;
  &lt;dd&gt;The AI-generated summaries that appear at the top of many Google search results, with links to the pages they draw on.&lt;/dd&gt;
  &lt;dt id="ai-washing"&gt;AI washing&lt;/dt&gt;
  &lt;dd&gt;Claiming artificial intelligence a product doesn't really use, or overstating what its AI does. The SEC brought its first AI washing cases in March 2024; see &lt;a href="https://gerriscorp.com/guides/ai-vendor-pitch/"&gt;how to vet an AI vendor's pitch&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="alt-text"&gt;Alt text&lt;/dt&gt;
  &lt;dd&gt;A text description of an image in the HTML alt attribute, read aloud by screen readers and used by search engines to understand the image.&lt;/dd&gt;
  &lt;dt id="amaras-law"&gt;Amara's law&lt;/dt&gt;
  &lt;dd&gt;Roy Amara's observation that people overestimate a technology's effect in the short run and underestimate it in the long run. It's the first test in &lt;a href="https://gerriscorp.com/guides/hype-or-shift/"&gt;hype or real shift&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="bimi"&gt;BIMI&lt;/dt&gt;
  &lt;dd&gt;Brand Indicators for Message Identification: a DNS record that lets supporting inboxes show a brand's logo beside authenticated email. It requires DMARC enforcement.&lt;/dd&gt;
  &lt;dt id="bing-webmaster-tools"&gt;Bing Webmaster Tools&lt;/dt&gt;
  &lt;dd&gt;Microsoft's free service for monitoring a site in Bing, whose index also feeds Microsoft Copilot and other AI assistants.&lt;/dd&gt;
  &lt;dt id="canonical-tag"&gt;Canonical tag&lt;/dt&gt;
  &lt;dd&gt;A link element that tells search engines which URL is the preferred version of a page when several URLs show the same or similar content. Google treats it as a strong hint.&lt;/dd&gt;
  &lt;dt id="cdn"&gt;CDN (content delivery network)&lt;/dt&gt;
  &lt;dd&gt;A network of servers around the world that stores copies of a site's files close to visitors, making pages faster and absorbing traffic spikes and attacks. Cloudflare is the best-known example.&lt;/dd&gt;
  &lt;dt id="citation-gap"&gt;Citation gap&lt;/dt&gt;
  &lt;dd&gt;A page AI assistants cite when answering your buyers' questions that names your competitors and leaves you out. Finding them is the job of a &lt;a href="https://gerriscorp.com/guides/citation-gap-analysis/"&gt;citation gap analysis&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="cls"&gt;CLS (Cumulative Layout Shift)&lt;/dt&gt;
  &lt;dd&gt;A Core Web Vital measuring how much visible content moves unexpectedly while a page loads. Good is 0.1 or less.&lt;/dd&gt;
  &lt;dt id="content-brief"&gt;Content brief&lt;/dt&gt;
  &lt;dd&gt;A one-page instruction that settles a page's question, answer, evidence, and limits before anyone writes. See &lt;a href="https://gerriscorp.com/guides/content-briefs/"&gt;writing a content brief&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="core-update"&gt;Core update&lt;/dt&gt;
  &lt;dd&gt;A broad recalibration of Google's ranking systems, announced a few times a year and logged on the Search Status Dashboard. Losses after one reflect being outranked, and Google says recovery can take months.&lt;/dd&gt;
  &lt;dt id="core-web-vitals"&gt;Core Web Vitals&lt;/dt&gt;
  &lt;dd&gt;Google's three page experience metrics: Largest Contentful Paint, Interaction to Next Paint, and Cumulative Layout Shift, measured from real Chrome users.&lt;/dd&gt;
  &lt;dt id="crawl-budget"&gt;Crawl budget&lt;/dt&gt;
  &lt;dd&gt;How many URLs a search engine is willing and able to crawl on a site in a given period. It matters mostly for large sites and sites with many low-value URLs.&lt;/dd&gt;
  &lt;dt id="crawled-currently-not-indexed"&gt;Crawled, currently not indexed&lt;/dt&gt;
  &lt;dd&gt;A Google Search Console status meaning Google fetched a page and chose not to index it. See the &lt;a href="https://gerriscorp.com/guides/crawled-not-indexed/"&gt;full guide&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="crawler-hints"&gt;Crawler Hints&lt;/dt&gt;
  &lt;dd&gt;A free Cloudflare feature that watches content change as it passes through Cloudflare's cache and notifies IndexNow engines such as Bing. Google doesn't take part.&lt;/dd&gt;
  &lt;dt id="discovered-currently-not-indexed"&gt;Discovered, currently not indexed&lt;/dt&gt;
  &lt;dd&gt;A Search Console status meaning Google knows a URL exists and hasn't crawled it yet, usually a sign of crawl priority problems.&lt;/dd&gt;
  &lt;dt id="dkim"&gt;DKIM&lt;/dt&gt;
  &lt;dd&gt;DomainKeys Identified Mail: a cryptographic signature on outgoing email, verified against a public key published in DNS.&lt;/dd&gt;
  &lt;dt id="dmarc"&gt;DMARC&lt;/dt&gt;
  &lt;dd&gt;A DNS policy telling receiving mail servers what to do with mail that fails SPF and DKIM alignment, and where to send reports. Policies are none, quarantine, and reject.&lt;/dd&gt;
  &lt;dt id="e-e-a-t"&gt;E-E-A-T&lt;/dt&gt;
  &lt;dd&gt;Experience, expertise, authoritativeness, and trustworthiness: the qualities Google's quality rater guidelines ask raters to assess, and a useful description of what makes content credible.&lt;/dd&gt;
  &lt;dt id="email-blocklist"&gt;Email blocklist&lt;/dt&gt;
  &lt;dd&gt;A list of IP addresses and domains that spam filters consult to reject or junk mail, also called a DNSBL. Getting off one starts with fixing the cause; see &lt;a href="https://gerriscorp.com/guides/email-blocklist-recovery/"&gt;blocklist recovery&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="geo"&gt;GEO (generative engine optimization)&lt;/dt&gt;
  &lt;dd&gt;Making a business and its content easy for AI systems to find, understand, and cite in generated answers.&lt;/dd&gt;
  &lt;dt id="git"&gt;Git&lt;/dt&gt;
  &lt;dd&gt;The version control system almost every software team uses. It records each change to a set of files, who made it, and why, so any version can be restored.&lt;/dd&gt;
  &lt;dt id="google-search-console"&gt;Google Search Console&lt;/dt&gt;
  &lt;dd&gt;Google's free service for monitoring how a site appears in Google Search: indexing, queries, clicks, Core Web Vitals, and errors.&lt;/dd&gt;
  &lt;dt id="hreflang"&gt;hreflang&lt;/dt&gt;
  &lt;dd&gt;An annotation that tells search engines which language or regional version of a page to show to which users.&lt;/dd&gt;
  &lt;dt id="hsts"&gt;HSTS (HTTP Strict Transport Security)&lt;/dt&gt;
  &lt;dd&gt;A response header that commits browsers to HTTPS for a site for a chosen length of time. Add preload only once every subdomain serves HTTPS, because preload is slow to reverse.&lt;/dd&gt;
  &lt;dt id="hype-cycle"&gt;Hype cycle&lt;/dt&gt;
  &lt;dd&gt;Gartner's curve for a new technology's reputation: a trigger, a peak of inflated expectations, a trough of disillusionment, a slope of enlightenment, and a plateau of productivity.&lt;/dd&gt;
  &lt;dt id="indexnow"&gt;IndexNow&lt;/dt&gt;
  &lt;dd&gt;An open protocol that lets a site notify Bing, Yandex, and other participating search engines immediately when pages are added, changed, or removed.&lt;/dd&gt;
  &lt;dt id="inp"&gt;INP (Interaction to Next Paint)&lt;/dt&gt;
  &lt;dd&gt;A Core Web Vital measuring how quickly a page responds to taps, clicks, and key presses. Good is 200 milliseconds or less. It replaced First Input Delay in March 2024.&lt;/dd&gt;
  &lt;dt id="json-ld"&gt;JSON-LD&lt;/dt&gt;
  &lt;dd&gt;A format for adding structured data to a page in a script block, recommended by Google for schema.org markup.&lt;/dd&gt;
  &lt;dt id="lcp"&gt;LCP (Largest Contentful Paint)&lt;/dt&gt;
  &lt;dd&gt;A Core Web Vital measuring how long the largest visible element, usually a hero image or headline, takes to appear. Good is 2.5 seconds or less.&lt;/dd&gt;
  &lt;dt id="llms-txt"&gt;llms.txt&lt;/dt&gt;
  &lt;dd&gt;A proposed convention, introduced in 2024, for a plain Markdown file at a site's root that summarizes the site and links to its key pages for language models. Adoption by AI companies is still emerging. This site publishes &lt;a href="https://gerriscorp.com/llms.txt"&gt;one&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="meta-description"&gt;Meta description&lt;/dt&gt;
  &lt;dd&gt;A short summary of a page in its HTML head, often shown under the title in search results. Usually 140 to 160 characters.&lt;/dd&gt;
  &lt;dt id="oai-searchbot"&gt;OAI-SearchBot&lt;/dt&gt;
  &lt;dd&gt;OpenAI's crawler for ChatGPT search. It's controlled separately from GPTBot, the training crawler, so a site can appear in ChatGPT answers while declining to be used for training.&lt;/dd&gt;
  &lt;dt id="orange-cloud"&gt;Orange cloud (proxied)&lt;/dt&gt;
  &lt;dd&gt;A Cloudflare DNS record whose traffic runs through Cloudflare's network, gaining caching, compression, and attack protection. Gray means Cloudflare answers DNS and steps aside.&lt;/dd&gt;
  &lt;dt id="pull-request"&gt;Pull request&lt;/dt&gt;
  &lt;dd&gt;A proposed set of changes to a repository, presented for review and comment before it's merged. Requiring one for every change gives an owner a complete history; see &lt;a href="https://gerriscorp.com/guides/hiring-developers/"&gt;hiring developers&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="readme"&gt;README&lt;/dt&gt;
  &lt;dd&gt;The introductory file at the top of a code repository. On a public GitHub repository it's a public page that AI assistants with web search read, so it should describe the live site; see &lt;a href="https://gerriscorp.com/guides/readme-ai-source/"&gt;your README is a source AI reads&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="repository"&gt;Repository&lt;/dt&gt;
  &lt;dd&gt;The folder of code and content Git tracks, along with its full history. Owning the repository, in an account the business controls, is owning the website.&lt;/dd&gt;
  &lt;dt id="robots-txt"&gt;robots.txt&lt;/dt&gt;
  &lt;dd&gt;A file at a site's root that tells crawlers which paths they may fetch. It controls crawling, not indexing; a blocked page can still appear in results if other sites link to it.&lt;/dd&gt;
  &lt;dt id="scaled-content-abuse"&gt;Scaled content abuse&lt;/dt&gt;
  &lt;dd&gt;The spam policy covering mass-produced pages made mainly to rank, by any method, human or machine. Google's 2026 spam updates focused on it.&lt;/dd&gt;
  &lt;dt id="schema-org"&gt;Schema.org&lt;/dt&gt;
  &lt;dd&gt;The shared vocabulary search engines use for structured data, describing organizations, people, products, services, articles, events, and much more.&lt;/dd&gt;
  &lt;dt id="server-side-rendering"&gt;Server-side rendering&lt;/dt&gt;
  &lt;dd&gt;Generating a page's full HTML on the server before sending it, so browsers and crawlers receive the content without running JavaScript. Static generation does the same at build time.&lt;/dd&gt;
  &lt;dt id="site-reputation-abuse"&gt;Site reputation abuse&lt;/dt&gt;
  &lt;dd&gt;Hosting third-party pages on an established site so they ride its reputation in search. Google introduced the policy in 2024 and updated it in August 2026.&lt;/dd&gt;
  &lt;dt id="soft-404"&gt;Soft 404&lt;/dt&gt;
  &lt;dd&gt;A page that tells visitors it doesn't exist or has nothing to show, while returning a 200 success status. Google treats it as an error.&lt;/dd&gt;
  &lt;dt id="spam-update"&gt;Spam update&lt;/dt&gt;
  &lt;dd&gt;An upgrade to the systems that detect violations of Google's spam policies. Unlike a core update, it targets specific behavior, and the sites it catches tend to fall fast.&lt;/dd&gt;
  &lt;dt id="speculation-rules"&gt;Speculation rules&lt;/dt&gt;
  &lt;dd&gt;A JSON block in a script tag that tells Chrome and Edge which links to prefetch or prerender before the reader clicks, so the next page opens at once. See &lt;a href="https://gerriscorp.com/guides/speculation-rules/"&gt;speculation rules&lt;/a&gt;.&lt;/dd&gt;
  &lt;dt id="spf"&gt;SPF&lt;/dt&gt;
  &lt;dd&gt;Sender Policy Framework: a DNS record listing the servers allowed to send email for a domain. It allows at most ten DNS lookups.&lt;/dd&gt;
  &lt;dt id="structured-data"&gt;Structured data&lt;/dt&gt;
  &lt;dd&gt;Machine-readable information about a page, usually schema.org in JSON-LD, describing what the page is about and how it relates to people, organizations, and other pages.&lt;/dd&gt;
  &lt;dt id="title-tag"&gt;Title tag&lt;/dt&gt;
  &lt;dd&gt;The HTML title of a page, shown in browser tabs and usually as the headline of a search result. Usually 50 to 60 characters.&lt;/dd&gt;
  &lt;dt id="ttfb"&gt;TTFB (time to first byte)&lt;/dt&gt;
  &lt;dd&gt;How long a browser waits for the first byte of a server's response. Slow TTFB delays everything else, and caching and CDNs reduce it.&lt;/dd&gt;
  &lt;dt id="xml-sitemap"&gt;XML sitemap&lt;/dt&gt;
  &lt;dd&gt;A file listing the URLs a site wants search engines to crawl, with optional last-modified dates.&lt;/dd&gt;
&lt;/dl&gt;

&lt;p class="related"&gt;Want these applied to your site? See &lt;a href="https://gerriscorp.com/services/"&gt;services&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>What Cloudflare's cache leaves exposed on a WordPress site</title>
    <link>https://gerriscorp.com/guides/wordpress-behind-cloudflare/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/wordpress-behind-cloudflare/</guid>
    <pubDate>Wed, 14 Oct 2026 12:00:00 +0000</pubDate>
    <description>Full-page caching keeps most visitors away from WordPress and MySQL. Here is the traffic that still reaches your server, and how to harden each path it takes.</description>
    <content:encoded>&lt;h1&gt;What Cloudflare's cache leaves exposed on a WordPress site&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-14"&gt;October 14, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Putting WordPress behind Cloudflare with full-page caching is the best single change I know for a slow WordPress site. Visitors get a saved copy of each page from a data center near them, and your server, PHP, and MySQL sit idle for most of the traffic. I think of it as a plate carrier: it covers the vital organs, where most of the hits land. The traffic it lets through goes straight to the parts of WordPress attackers aim at, so those parts need their own protection.&lt;/p&gt;

&lt;h2&gt;How the cache decides&lt;/h2&gt;
&lt;p&gt;Cloudflare doesn't cache HTML by default. You turn page caching on with Automatic Platform Optimization, which Cloudflare built for WordPress, or with a cache rule that tells Cloudflare to cache everything on chosen paths. APO comes with its own logic for what to skip. Cloudflare's documentation says it bypasses the cache for logged-in and administrator users so private content isn't cached, and that a request skips the cache when bypass cookies are present, naming logged-in, session, and WooCommerce cookies. With the Cloudflare WordPress plugin installed, the edge copy updates when you change content.&lt;/p&gt;
&lt;p&gt;A hand-written cache rule has none of that logic until you add it. A rule that caches everything and forgets the login cookie will serve one editor's view of the site to the public. If you write your own rules, add a bypass for requests carrying the WordPress login cookie, and test logged in and logged out before trusting it.&lt;/p&gt;

&lt;h2&gt;Traffic that still reaches your server&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Logins.&lt;/strong&gt; Every request to &lt;code&gt;wp-login.php&lt;/code&gt; and the dashboard runs PHP and queries the database. Password-guessing bots aim here first.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;XML-RPC.&lt;/strong&gt; &lt;code&gt;xmlrpc.php&lt;/code&gt; is an older remote-publishing interface that WordPress still ships. It accepts logins too, and bots use it for password guessing. Jetpack depends on it, so check before closing it.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Logged-in sessions.&lt;/strong&gt; Editors, members, and customers with a session cookie get uncached pages, by design.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Forms, comments, and search.&lt;/strong&gt; Every submission and search query goes to the origin.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Carts and checkout&lt;/strong&gt; on WooCommerce, which carry session cookies.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Anything requested by IP address&lt;/strong&gt;, if the server answers requests that skip Cloudflare entirely.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Hardening the paths that remain&lt;/h2&gt;
&lt;p&gt;WordPress's own hardening guide, in the Advanced Administration Handbook, covers most of this, and I follow it:&lt;/p&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Update core, plugins, and themes.&lt;/strong&gt; The handbook's first advice is to keep WordPress current, and it says the same of plugins.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Delete plugins you don't use.&lt;/strong&gt; The handbook says to remove an unused plugin from the system entirely. A deactivated plugin still sits on disk.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Use strong passwords and two-step login&lt;/strong&gt; for every account that can publish, and close accounts that belong to people who've left.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Turn off the dashboard file editor&lt;/strong&gt; with &lt;code&gt;define( 'DISALLOW_FILE_EDIT', true );&lt;/code&gt; in &lt;code&gt;wp-config.php&lt;/code&gt;, so a stolen administrator login can't rewrite your theme's code from the browser.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Protect the login.&lt;/strong&gt; The handbook suggests server-side password protection for the admin area as a second layer. At Cloudflare, a rate limiting rule on &lt;code&gt;wp-login.php&lt;/code&gt; stops bulk guessing before it reaches the server.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Close XML-RPC&lt;/strong&gt; when nothing needs it, with a Cloudflare rule that blocks the path or a plugin that disables it.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Back up the files and the database&lt;/strong&gt;, which the handbook names explicitly, and store the copies away from the server. Restore one to a test site once, so you know the backups work.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Accept only Cloudflare's traffic&lt;/strong&gt; at the origin, through your host's firewall or Cloudflare Tunnel, so nobody can go around the proxy.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;The maintenance never stops&lt;/h2&gt;
&lt;p&gt;Caching removes the speed problem and most of the load. It leaves the upkeep: updates every week, plugin audits, backup checks, and an eye on login attempts. That is the standing cost of running WordPress, and for stores, membership sites, and newsrooms it's worth paying.&lt;/p&gt;
&lt;p&gt;For a company site that changes a few times a year, it often isn't. A static site has no login page, no database, and no plugins, so the whole list above disappears. I covered that trade in &lt;a href="https://gerriscorp.com/services/wordpress-care/"&gt;WordPress care: caching, hardening, or a static move&lt;/a&gt;. For the Cloudflare settings that affect search engines and AI crawlers, see &lt;a href="https://gerriscorp.com/guides/cloudflare-seo/"&gt;Cloudflare settings that help or hurt SEO and AI crawlers&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.cloudflare.com/automatic-platform-optimization/about/"&gt;Cloudflare Docs: About Automatic Platform Optimization&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developer.wordpress.org/advanced-administration/security/hardening/"&gt;WordPress Advanced Administration Handbook: Hardening WordPress&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.cloudflare.com/cache/how-to/cache-rules/"&gt;Cloudflare Docs: Cache Rules&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.cloudflare.com/waf/rate-limiting-rules/"&gt;Cloudflare Docs: Rate limiting rules&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Speculation rules: make the next page open instantly in Chrome</title>
    <link>https://gerriscorp.com/guides/speculation-rules/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/speculation-rules/</guid>
    <pubDate>Wed, 14 Oct 2026 12:00:00 +0000</pubDate>
    <description>One script tag tells Chrome to fetch a page while the reader decides to click. How prefetch and prerender differ, which eagerness to pick, and what to exclude.</description>
    <content:encoded>&lt;h1&gt;Speculation rules: make the next page open instantly in Chrome&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-14"&gt;October 14, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Speculation rules are a short block of JSON in a script tag that tells Chrome and Edge which links to load before the reader clicks them. When the reader does click, the page is already in memory and opens at once. I added one rule to every site I run in October 2026. It costs nothing in browsers that don't support it, and on a static site it takes about five minutes.&lt;/p&gt;

&lt;h2&gt;Prefetch or prerender&lt;/h2&gt;
&lt;p&gt;A prefetch downloads the next page's HTML and holds it in memory without rendering it. A prerender goes further: Chrome's documentation says the page behaves as if it had been opened in an invisible background tab, then is shown when the reader arrives. Prerender is faster still and uses more memory and bandwidth, and it runs the page's scripts before anyone sees the page.&lt;/p&gt;
&lt;p&gt;I use prefetch. It removes the network wait, which is most of the delay on a static site, and it runs no scripts early, so analytics, consent tools, and anything else that should only run for a real visit behave exactly as before. Prerender suits sites whose pages are heavy to render and whose scripts are built to handle it. Chrome also caps how much it will speculate: its documentation allows up to 50 immediate prefetches and 10 immediate prerenders per page, so a rule that matches every link on a long page cannot flood the network.&lt;/p&gt;

&lt;h2&gt;Choose the eagerness&lt;/h2&gt;
&lt;p&gt;Eagerness sets how early Chrome acts on a rule. The four levels, from Chrome's documentation:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;immediate:&lt;/strong&gt; as soon as the rule is seen.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;eager:&lt;/strong&gt; on desktop, when the pointer rests on a link for 10 milliseconds; on mobile, shortly after a link enters the viewport.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;moderate:&lt;/strong&gt; on desktop, when the pointer rests on a link for 200 milliseconds, or on pointer down if that comes first; on mobile, using viewport heuristics after the reader stops scrolling.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;conservative:&lt;/strong&gt; on pointer or touch down, the default for rules that match links on the page.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Moderate is the setting I use. A 200-millisecond hover is a good signal of intent, and it fetches few pages the reader never opens.&lt;/p&gt;

&lt;h2&gt;The rule I use&lt;/h2&gt;
&lt;pre&gt;&lt;code&gt;&amp;lt;script type="speculationrules"&amp;gt;
{"prefetch": [{"where": {"and": [
  {"href_matches": "/*"},
  {"not": {"href_matches": ["/*.md", "/*.txt", "/*.xml"]}}
]}, "eagerness": "moderate"}]}
&amp;lt;/script&amp;gt;&lt;/code&gt;&lt;/pre&gt;
&lt;p&gt;It matches every same-site link and excludes the files no reader opens in a browser: Markdown copies, llms.txt, and XML feeds and sitemaps. Exclude anything with side effects as well. A logout link, an "add to cart" URL, or any GET request that changes state must never be prefetched. On a static site there are none; on an application, list them.&lt;/p&gt;

&lt;h2&gt;Check that it works&lt;/h2&gt;
&lt;p&gt;Open the site in Chrome, open DevTools, and go to the Application panel. Under background services, the speculative loads view lists the rules Chrome found on the page and the status of each speculation: not triggered, running, ready, or failed, with the reason for any failure. Hover over a link for a moment and watch its entry change. The Network panel shows the prefetch request itself.&lt;/p&gt;
&lt;p&gt;Chrome marks speculative requests with a Sec-Purpose request header, set to prefetch for a prefetch. That gives servers and log analysis a way to separate speculative fetches from real visits. A static host serves the page either way. A site that counts visits on the server side should exclude requests carrying that header, so prefetches never inflate the numbers.&lt;/p&gt;

&lt;h2&gt;Browser support&lt;/h2&gt;
&lt;p&gt;Chrome and Edge support speculation rules, with eagerness settings since version 121. Firefox does not support them, and Safari has them behind a flag. Unsupported browsers ignore the script tag, so the rule degrades to an ordinary link.&lt;/p&gt;

&lt;h2&gt;What it changes&lt;/h2&gt;
&lt;p&gt;Lighthouse measures a single page load, so speculation rules leave its scores where they were. What changes is the second page: the reader clicks and the page is already there. That is the part of speed a visitor notices most. In field data, the prefetched page's Largest Contentful Paint starts from a page already in memory, so real-user loading times improve once enough visits have been recorded.&lt;/p&gt;

&lt;h2&gt;Where I fit&lt;/h2&gt;
&lt;p&gt;I add speculation rules as part of a &lt;a href="https://gerriscorp.com/services/site-speed/"&gt;site speed&lt;/a&gt; engagement, along with the work Lighthouse does measure: render-blocking resources, image sizes, and third-party scripts. For the measurement side, start with my guide to &lt;a href="https://gerriscorp.com/guides/core-web-vitals/"&gt;Core Web Vitals&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developer.chrome.com/docs/web-platform/prerender-pages"&gt;Chrome for Developers: Prerender pages in Chrome for instant page navigations (updated January 23, 2026)&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developer.mozilla.org/en-US/docs/Web/API/Speculation_Rules_API"&gt;MDN: Speculation Rules API&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://web.dev/articles/lcp"&gt;web.dev: Largest Contentful Paint&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Your public GitHub README is a source AI search reads</title>
    <link>https://gerriscorp.com/guides/readme-ai-source/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/readme-ai-source/</guid>
    <pubDate>Wed, 14 Oct 2026 12:00:00 +0000</pubDate>
    <description>AI assistants with web search read public repositories and quote README files as fact. Keep the README current, point it to the live site, and keep secrets out.</description>
    <content:encoded>&lt;h1&gt;Your public GitHub README is a source AI search reads&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-14"&gt;October 14, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;If your website is built from a public GitHub repository, its README file is a public page, and AI assistants with web search read it. In October 2026 I asked Claude what gerriscorp.com was. Its answer quoted this site's own repository, and it described the site as a preview that hadn't launched, because the README still carried the setup notes I wrote before launch. The site had been live for nine days. The fix took ten minutes.&lt;/p&gt;

&lt;h2&gt;Why a README gets read&lt;/h2&gt;
&lt;p&gt;GitHub shows a repository's README on its front page, so the file is public web content like any other. When an assistant searches for a site or a company and the repository ranks for that name, the README is often the most direct description it can find: who built the site, what it is for, and what state it is in. Developers write READMEs for other developers, in the present tense, and then stop updating them. An assistant can't tell a launch checklist from a status report.&lt;/p&gt;
&lt;p&gt;Static sites on GitHub Pages, Netlify, and Vercel all tend to sit in public repositories. Agencies, open-source projects, and companies with developer documentation have the same exposure: every public repository that names the company is a page that can describe it.&lt;/p&gt;

&lt;h2&gt;What to put at the top&lt;/h2&gt;
&lt;p&gt;Open the README with two sentences for the reader who arrived from a search: what the repository builds, the live address, and the date it went live. Then say where to learn about the company itself. Mine now begins: "This repository builds https://gerriscorp.com, live since October 5, 2026. For what Gerris Corp does and who runs it, read the site; this file only explains how the site is built."&lt;/p&gt;
&lt;p&gt;Below that, write for developers as before. Rewrite any section that describes a past state as current: "the site starts as a noindexed preview" becomes a short "Live" section with the date.&lt;/p&gt;

&lt;h2&gt;Keep it current on the same schedule as the site&lt;/h2&gt;
&lt;p&gt;Treat the README as a page of the site. When the site's status, name, address, or ownership changes, the README changes in the same commit. A check in the build can enforce it: fail when the README mentions "preview" or "staging" while the site is set to live. My build scripts already check titles, descriptions, and links on every run, and the README is one more file to check.&lt;/p&gt;

&lt;h2&gt;Check every repository that names you&lt;/h2&gt;
&lt;p&gt;The live site's repository is rarely the only one. Old prototypes, a redesign that never shipped, a fork a contractor made, and a project from a previous company name can all sit in public accounts with READMEs describing a state of affairs that ended years ago. Search GitHub for the company name and the domain. An archived repository stays public and readable; archiving only freezes it. For each one that describes a past state, either add a first line that says so and links the current site, or make it private if nothing in it needs to stay public.&lt;/p&gt;

&lt;h2&gt;Keep confidential material out of it&lt;/h2&gt;
&lt;p&gt;A public README is public. Client names, internal hostnames, credentials, and notes about how a client relationship works belong in a private file. I keep a separate briefing file for each of my sites, listed in the repository's ignore file so it never reaches GitHub. If a public README already contains something private, edit it and treat the old version as exposed, since the history is public too. GitHub's documentation on repository visibility explains what changing a repository to private does and doesn't hide.&lt;/p&gt;

&lt;h2&gt;Check what the assistants say&lt;/h2&gt;
&lt;p&gt;Ask an assistant with web search to describe your company and your site, then read the sources it cites. If a repository appears among them, read its README the way a stranger would. Google's guidance for its AI features applies here as everywhere: the content the systems can find is the content they use. My guide to &lt;a href="https://gerriscorp.com/guides/citation-gap-analysis/"&gt;citation gap analysis&lt;/a&gt; covers checking sources systematically, and &lt;a href="https://gerriscorp.com/guides/repository-access/"&gt;giving your SEO repository access&lt;/a&gt; covers working in the repository itself.&lt;/p&gt;

&lt;h2&gt;Where I fit&lt;/h2&gt;
&lt;p&gt;I audit the public sources that describe a company, repositories included, and fix what they say. If your site is built from a public repository, &lt;a href="https://gerriscorp.com/contact/"&gt;send me the link&lt;/a&gt; and I'll tell you what an assistant would learn from it.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/customizing-your-repository/about-readmes"&gt;GitHub Docs: About READMEs&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/managing-repository-settings/setting-repository-visibility"&gt;GitHub Docs: Setting repository visibility&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/ai-features"&gt;Google Search Central: AI features and your website&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to write a content brief for pages AI search will quote</title>
    <link>https://gerriscorp.com/guides/content-briefs/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/content-briefs/</guid>
    <pubDate>Wed, 14 Oct 2026 12:00:00 +0000</pubDate>
    <description>A one page brief that fixes the question, answer, evidence, and limits before anyone writes, so the finished page answers first and holds up when quoted.</description>
    <content:encoded>&lt;h1&gt;How to write a content brief for pages AI search will quote&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-14"&gt;October 14, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A content brief is a one-page instruction that settles a page's question, its answer, its evidence, and its limits before anyone writes a sentence. For AI search the brief carries more weight than it used to, because answer engines lift single passages out of a page and repeat them without the context around them. A page built from a good brief answers in its first lines, supports every claim, and stays inside its own question. I learned the discipline as an executive editor running a staff of writers, and it is the same job now with a new kind of reader.&lt;/p&gt;

&lt;h2&gt;The question, word for word&lt;/h2&gt;
&lt;p&gt;Every brief starts with one question, written the way a buyer asks it. Take the wording from Search Console queries, sales call notes, or the questions in an AI visibility question set. "How do I fix Crawled, currently not indexed" is a brief. "Indexing issues" is a topic, and a page written from a topic answers nothing in particular.&lt;/p&gt;
&lt;p&gt;Under the question, name who asks it and what they will decide after reading. A developer deciding whether a template is at fault needs a different page from a marketing director deciding whether to hire help, even when they type the same words.&lt;/p&gt;

&lt;h2&gt;The answer, written by the expert first&lt;/h2&gt;
&lt;p&gt;The person who knows the subject writes the answer in two or three sentences before the writer starts. This is the passage an AI system is most likely to lift, so it has to stand alone: it names the thing, gives the answer, and includes the number or condition that makes it true. A writer can polish it. A writer working without it will produce a careful introduction that circles the answer for three paragraphs.&lt;/p&gt;
&lt;p&gt;Google's guidance on helpful content asks whether a page provides original information and substantial value beyond what other pages offer. The expert's answer is where that originality lives: the threshold from your own projects, the failure you have seen five times, the step everyone skips.&lt;/p&gt;

&lt;h2&gt;The evidence list&lt;/h2&gt;
&lt;p&gt;List every claim the page will make that a reader could doubt, and the source for each one: a vendor's documentation page, your own data with its date and sample, a public record. Writers open every source before citing it, and the brief says so. Answer engines cross-check, and a confident sentence with no support behind it is the sentence they drop or contradict.&lt;/p&gt;
&lt;p&gt;Mark which evidence is first-party. Data from your own client work, anonymized, is the material no competitor can copy, and it is what turns a page into a source other pages cite.&lt;/p&gt;

&lt;h2&gt;The sections, each one self-contained&lt;/h2&gt;
&lt;p&gt;Outline the sections as questions the reader will ask next, in the order they ask them. Each section opens with its own answer and makes sense if it is the only part quoted. Google says AI Overviews and AI Mode may run several related searches across subtopics to build a response, so a section that answers a sub-question cleanly can be the passage selected for that sub-search.&lt;/p&gt;

&lt;h2&gt;The limits&lt;/h2&gt;
&lt;p&gt;A brief says what the page does not cover and where that material lives instead, with the link. This is how a site avoids two pages answering the same question, and it keeps each page short enough to finish. When the limit points to a page that doesn't exist yet, the brief has found the next brief.&lt;/p&gt;

&lt;h2&gt;The links, title, and description&lt;/h2&gt;
&lt;p&gt;List the internal links the page must carry: up to the hub page for the subject, across to the pages a reader needs next, and from two or three existing pages that should now point to this one. Draft the title and meta description in the brief, inside the site's limits, because Google often uses the description for the snippet when it describes the page better than the page's own text. Structured data, where the page uses it, must match what the visible page says; Google's AI features guidance asks for exactly that.&lt;/p&gt;

&lt;h2&gt;The review before publishing&lt;/h2&gt;
&lt;p&gt;Before a page goes live, someone other than the writer checks it against the brief: the answer is in the first lines, every claim has its source, the limits held, and every link resolves. AI drafting tools speed up the writing and change none of this. Google's guidance on generative AI content judges the result by its accuracy, quality, and relevance, and so does every reader who decides whether to trust you.&lt;/p&gt;

&lt;h2&gt;Where I fit&lt;/h2&gt;
&lt;p&gt;I write briefs for clients' writers and agencies' content teams, supply the expert answer and the evidence list from the technical side, and review drafts before they publish. If your content team is producing pages that rank but never get quoted, &lt;a href="https://gerriscorp.com/contact/"&gt;send me two of them&lt;/a&gt; and I'll show you what the brief was missing. For the source side of the same problem, see my guide to &lt;a href="https://gerriscorp.com/guides/citation-gap-analysis/"&gt;citation gap analysis&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/fundamentals/creating-helpful-content"&gt;Google Search Central: Creating helpful, reliable, people first content&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/fundamentals/using-gen-ai-content"&gt;Google Search Central: Guidance on using generative AI content&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/ai-features"&gt;Google Search Central: AI features and your website&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/snippet"&gt;Google Search Central: Control your snippets in search results&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Citation gap analysis: finding the sources AI cites instead of you</title>
    <link>https://gerriscorp.com/guides/citation-gap-analysis/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/citation-gap-analysis/</guid>
    <pubDate>Wed, 14 Oct 2026 12:00:00 +0000</pubDate>
    <description>Collect the pages AI assistants cite when they name your competitors, sort them by type, and turn the ones that leave you out into a ranked outreach list.</description>
    <content:encoded>&lt;h1&gt;Citation gap analysis: finding the sources AI cites instead of you&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-14"&gt;October 14, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A citation gap is a page that AI assistants cite when they answer your buyers' questions, which names your competitors and leaves you out. Finding them is mechanical: ask the questions, collect every cited URL, sort the sources by type, and mark the ones where a competitor appears and you don't. The result is a ranked list of places to be, and it replaces guessing about which publications "AI trusts" with the actual pages it used.&lt;/p&gt;

&lt;h2&gt;Collect the sources, not only the answers&lt;/h2&gt;
&lt;p&gt;Start with a fixed set of buyer questions and run them with web search turned on. The answer text tells you who was named. The citations tell you why. OpenAI's API returns two lists for a web search answer: the URLs cited inline, and, on request, every source the model consulted, which is usually longer. Claude returns the cited URL with each passage it draws from a search result. Keep both lists for every answer, with the date and the engine.&lt;/p&gt;
&lt;p&gt;Bing Webmaster Tools now reports how often your own pages are cited in Copilot answers and lets you compare that citation share against other sites. That is the one place a search engine shows citation counts directly, and it covers only your side of the gap. The question set covers the rest.&lt;/p&gt;

&lt;h2&gt;Sort every source by type&lt;/h2&gt;
&lt;p&gt;A pile of URLs becomes useful once each one carries a label. I use seven:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;marketplace and directory listings (freelance platforms, agency directories, review sites)&lt;/li&gt;
  &lt;li&gt;"best of" roundups and ranked lists&lt;/li&gt;
  &lt;li&gt;practitioner guides written by third parties&lt;/li&gt;
  &lt;li&gt;vendor and tool company blogs&lt;/li&gt;
  &lt;li&gt;trade press and news&lt;/li&gt;
  &lt;li&gt;forums and communities&lt;/li&gt;
  &lt;li&gt;competitors' own sites&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;The labels decide the remedy. A listing gap is closed with a complete, current profile. A roundup gap is an editorial pitch. A practitioner-guide gap is closed by writing the better answer yourself and by giving the writers of those guides a reason to cite you. A competitor's own page that keeps getting cited is a content benchmark: read it and work out what it answers that your page doesn't.&lt;/p&gt;

&lt;h2&gt;Mark the gaps and rank them&lt;/h2&gt;
&lt;p&gt;For each source, record which names it mentions. A source that names two of your competitors and not you is a gap. Rank the gaps by how many questions cited the source, then by how reachable it is. A directory you can join this afternoon outranks a national publication that might cover you next year, even though the publication carries more weight.&lt;/p&gt;

&lt;h2&gt;What a real run turned up&lt;/h2&gt;
&lt;p&gt;In October 2026 I ran sixteen buyer questions about my own services through Claude with web search. On the hiring questions, the names came from marketplace listing pages, an agency directory, and three "best consultants" lists. On a question about recovering from a traffic drop, the answer named four well-known practitioners, and it found them through three recovery guides written by other people, none of them the practitioners' own pages. On the how-to questions, vendor blogs outnumbered practitioners' guides.&lt;/p&gt;
&lt;p&gt;My own gaps were plain. I was absent from the roundups that supplied the hiring names, and none of my guides was among the sources cited for the how-to questions. The answers that did name me rested almost entirely on pages I wrote, and the assistant said so. That sets the work: listings and roundups first, because they are reachable, and third-party coverage over the following months.&lt;/p&gt;

&lt;h2&gt;Closing a gap without buying your way in&lt;/h2&gt;
&lt;p&gt;Some roundups sell placement. A paid inclusion is an endorsement relationship, and the FTC's Endorsement Guides require a clear disclosure of any material connection between an advertiser and the person recommending it. An undisclosed paid placement is a legal problem before it is an SEO one, and a ranked list that discloses payment carries little weight with readers or with AI systems that read the disclosure. Pitch the editors on evidence instead: a specific result, a public case study, a quote they can use. I wrote separately about why &lt;a href="https://gerriscorp.com/guides/self-ranking-lists/"&gt;ranking yourself first in your own list&lt;/a&gt; stopped working.&lt;/p&gt;

&lt;h2&gt;Rerun it and watch the list move&lt;/h2&gt;
&lt;p&gt;Run the same questions monthly. A gap is closed when the source names you and the answers start to cite it beside your name. Sources drop out of the answers on their own as the assistants find newer pages, so pull the list fresh each month. Pair this with a log of identifiable referral visits, as described in my guide to &lt;a href="https://gerriscorp.com/guides/measure-ai-search/"&gt;measuring AI search&lt;/a&gt;, so the report shows outcomes as well as presence.&lt;/p&gt;

&lt;h2&gt;Where I fit&lt;/h2&gt;
&lt;p&gt;I run citation gap analyses for companies and as a specialist on agency teams: the question set, the source collection, the ranked gap list, and the pitches or pages that close each gap. If you want to know which sources are speaking for your competitors, &lt;a href="https://gerriscorp.com/contact/"&gt;tell me your market and three competitors&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/ai-features"&gt;Google Search Central: AI features and your website&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.openai.com/api/docs/guides/tools-web-search"&gt;OpenAI: Web search tool guide (citations and the full sources list)&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://blogs.bing.com/search/2026/6/New-AI-Visibility-Insights-in-Bing-Webmaster-Tools-Intents-Topics-Citation-Share-Compare/"&gt;Microsoft Bing: AI visibility insights in Bing Webmaster Tools&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.ecfr.gov/current/title-16/chapter-I/subchapter-B/part-255"&gt;FTC: Guides Concerning the Use of Endorsements and Testimonials in Advertising (16 CFR Part 255)&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to add SEO regression tests to your website's build</title>
    <link>https://gerriscorp.com/guides/seo-build-checks/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/seo-build-checks/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>Automated SEO checks that run on every build and pull request: titles, descriptions, canonicals, noindex, schema, sitemaps, headings, and links, with examples.</description>
    <content:encoded>&lt;h1&gt;How to add SEO regression tests to your website's build&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Most SEO damage on a coded website arrives in an ordinary pull request. Someone refactors a layout, and the canonical tag disappears from 4,000 pages. A content import runs, and every description is blank. Nobody notices for a month, because nobody was looking. A short set of automated checks, run on every build, looks every time. Here is what to test and how to wire it in.&lt;/p&gt;

&lt;h2&gt;Where the checks run&lt;/h2&gt;
&lt;p&gt;Continuous integration runs your tests automatically whenever code is pushed. On GitHub Actions, the results appear on the pull request itself, so a reviewer sees a failed check before anyone merges. SEO tests belong in the same place as the unit tests your developers already run: they're just tests whose subject is the built HTML.&lt;/p&gt;
&lt;p&gt;The simplest pattern is to build the site, then run a script over the output. A static or prerendered site produces a folder of HTML files to read directly. A server rendered site can be started in the CI job and fetched page by page from a list of representative URLs, one per template.&lt;/p&gt;

&lt;h2&gt;The checks worth having&lt;/h2&gt;
&lt;h3&gt;Titles and descriptions&lt;/h3&gt;
&lt;ul&gt;
  &lt;li&gt;Every page has exactly one title element and one meta description, and neither is empty.&lt;/li&gt;
  &lt;li&gt;No two pages share a title or a description. Google's documentation asks for distinct titles on every page and warns against repeated or boilerplate text, and it says the same of descriptions.&lt;/li&gt;
  &lt;li&gt;Lengths stay inside a range you choose. Google sets no length limit and truncates to fit the device, so the range is a house style. Mine is 50 to 60 characters for titles and 140 to 160 for descriptions, and the build enforces it.&lt;/li&gt;
&lt;/ul&gt;
&lt;h3&gt;Indexing signals&lt;/h3&gt;
&lt;ul&gt;
  &lt;li&gt;Every indexable page carries exactly one canonical tag, pointing at its own production URL.&lt;/li&gt;
  &lt;li&gt;No production page carries a noindex, unless it's on an explicit allow list.&lt;/li&gt;
  &lt;li&gt;robots.txt in the production build allows what you expect and lists the sitemap.&lt;/li&gt;
&lt;/ul&gt;
&lt;h3&gt;Sitemap agreement&lt;/h3&gt;
&lt;ul&gt;
  &lt;li&gt;Every URL in the sitemap exists in the build and returns 200.&lt;/li&gt;
  &lt;li&gt;Every indexable page in the build appears in the sitemap.&lt;/li&gt;
&lt;/ul&gt;
&lt;h3&gt;Structured data&lt;/h3&gt;
&lt;ul&gt;
  &lt;li&gt;Every JSON-LD block parses as valid JSON.&lt;/li&gt;
  &lt;li&gt;Each template emits the types it should, such as Product on product pages and Article on posts.&lt;/li&gt;
  &lt;li&gt;Key values match the visible page. Google's guidelines say structured data should represent the main content of the page and describe content the user can see, so a price or a name in the JSON-LD that differs from the page is a real fault.&lt;/li&gt;
&lt;/ul&gt;
&lt;h3&gt;Headings and markup&lt;/h3&gt;
&lt;ul&gt;
  &lt;li&gt;Exactly one H1 per page.&lt;/li&gt;
  &lt;li&gt;The HTML parses with balanced tags and no invalid nesting, which also heads off a class of React hydration errors.&lt;/li&gt;
&lt;/ul&gt;
&lt;h3&gt;Links&lt;/h3&gt;
&lt;ul&gt;
  &lt;li&gt;Every internal link points at a page that exists in the build.&lt;/li&gt;
  &lt;li&gt;No internal link points at a URL that redirects, so links go straight to the final address.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;A minimal example&lt;/h2&gt;
&lt;p&gt;A check doesn't need a framework. This Python sketch reads every built page and fails the build on a missing or duplicate title:&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;import pathlib, re, sys
seen, problems = {}, []
for page in pathlib.Path("out").rglob("*.html"):
    html = page.read_text(encoding="utf-8")
    titles = re.findall(r"&amp;lt;title&amp;gt;(.*?)&amp;lt;/title&amp;gt;", html, re.S)
    if len(titles) != 1:
        problems.append(f"{page}: {len(titles)} title elements")
        continue
    seen.setdefault(titles[0].strip(), []).append(str(page))
for title, pages in seen.items():
    if len(pages) &amp;gt; 1:
        problems.append(f"duplicate title {title!r}: {pages}")
print("\n".join(problems))
sys.exit(1 if problems else 0)&lt;/code&gt;&lt;/pre&gt;
&lt;p&gt;Add a step to the CI workflow that runs it after the build. A nonzero exit fails the check, and the message names the page and the rule, so the developer can fix it without asking anyone. For anything beyond a sketch, use a real HTML parser in place of regular expressions.&lt;/p&gt;

&lt;h2&gt;Checks that guard the writing, too&lt;/h2&gt;
&lt;p&gt;Build checks can protect content as well as markup. This site's build compares the text of every page with every other page and fails when two pages share more than 15% of their six-word phrases, or when a sentence of ten or more words appears on two pages. That keeps near-duplicate pages from creeping in as the site grows. A similar rule suits any site that publishes many pages from a shared template, such as location or category pages.&lt;/p&gt;

&lt;h2&gt;Rolling it out without blocking everyone&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Start in report only mode.&lt;/strong&gt; Run the checks and print the results without failing the build, and fix the existing backlog.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Turn on failure one rule at a time,&lt;/strong&gt; starting with the ones that cost the most when they break: canonicals, noindex, and robots.txt.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Keep messages specific.&lt;/strong&gt; "Missing canonical on /products/blue-widget/" gets fixed in a minute. "SEO check failed" gets the check switched off.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Make the rules a file in the repository,&lt;/strong&gt; reviewed like code, so changing a threshold is a decision someone signs off.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;What tests can't tell you&lt;/h2&gt;
&lt;p&gt;Build checks prove the HTML you ship meets your rules. They can't tell you how Google ranks it, whether a title persuades anyone to click, or what an AI assistant says about you. Pair them with a crawl of the live site after each release and with Search Console, and keep a person reading the pages.&lt;/p&gt;

&lt;p class="related"&gt;I write these checks into client repositories as part of &lt;a href="https://gerriscorp.com/services/seo-engineering/"&gt;SEO engineering&lt;/a&gt;. For the Next.js specifics they should cover, see the &lt;a href="https://gerriscorp.com/guides/nextjs-seo/"&gt;Next.js SEO checklist&lt;/a&gt;, and for safe repository setup, &lt;a href="https://gerriscorp.com/guides/repository-access/"&gt;giving an outside SEO access&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/actions/get-started/continuous-integration"&gt;GitHub Docs: About continuous integration&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/title-link"&gt;Google Search Central: Influencing your title links&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/snippet"&gt;Google Search Central: Control your snippets&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/structured-data/sd-policies"&gt;Google Search Central: General structured data guidelines&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to give an outside SEO safe access to your code repository</title>
    <link>https://gerriscorp.com/guides/repository-access/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/repository-access/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>A safe setup for letting an SEO consultant into your GitHub or GitLab repository: one repo, the Read role first, reviewed pull requests, and no secrets.</description>
    <content:encoded>&lt;h1&gt;How to give an outside SEO safe access to your code repository&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;On a site built with Next.js, React, or a headless CMS, the search problems live in the code, so the person fixing them needs to see it. Granting that access is a small, routine job when you set it up in the right order. Here is the order I ask clients to follow, and the reasons behind each step.&lt;/p&gt;

&lt;h2&gt;Why an SEO asks for the repository at all&lt;/h2&gt;
&lt;p&gt;A crawl tells me that 900 pages share a title. The repository tells me which component writes it, which data it reads, and whether a fix will reach every route that uses it. Without the code, I write a ticket describing the symptom and your developer goes looking for the cause. With it, the ticket names the file, or arrives as a pull request ready for review. Read access alone usually cuts the time from finding to fix by a wide margin.&lt;/p&gt;

&lt;h2&gt;Invite them to one repository, as an outside collaborator&lt;/h2&gt;
&lt;p&gt;On GitHub, a consultant belongs outside your organization's membership. Add them as an outside collaborator on the one repository that builds the website, rather than as an organization member who inherits base permissions across everything you own. Your other repositories, your internal tools, and your infrastructure code stay out of view. If the site is split across a frontend and a content repository, add them to both and nothing else.&lt;/p&gt;

&lt;h2&gt;Start with the Read role&lt;/h2&gt;
&lt;p&gt;GitHub's repository roles, from least access to most, are Read, Triage, Write, Maintain, and Admin. GitHub's own guidance is to pick the role that fits each person's function without giving them more access than they need.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Read&lt;/strong&gt; is GitHub's role for people who view or discuss a project without changing it. That covers the whole diagnostic phase.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Triage&lt;/strong&gt; adds managing issues and pull requests without write access, which suits a consultant who files tickets in your tracker.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Write&lt;/strong&gt; is for contributors who push to the project, which an SEO needs only once they start making changes themselves.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Maintain&lt;/strong&gt; and &lt;strong&gt;Admin&lt;/strong&gt; are for your own team. An outside SEO never needs either.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;GitLab has its own set: Guest, Reporter, Developer, Maintainer, and Owner, ordered from fewest permissions to most, plus specialized Planner and Security Manager roles. Reporter is the usual starting point for an outside reviewer there, and Developer is roughly the equivalent of GitHub's Write.&lt;/p&gt;

&lt;h2&gt;Grant Write only after main is protected&lt;/h2&gt;
&lt;p&gt;Before anyone outside your team can push, make sure nobody can push straight to your main branch. On GitHub, a ruleset or a branch protection rule controls who can push to a branch and can require reviews on pull requests before they merge. GitHub's documentation notes that when both a ruleset and a protection rule apply to the same branch, all the rules apply, and the most restrictive version of each wins.&lt;/p&gt;
&lt;p&gt;With that in place, Write access becomes safe: the consultant can create branches and open pull requests, and every change still waits for one of your people to approve it. Nothing they write reaches production without your review.&lt;/p&gt;

&lt;h2&gt;Use deploy previews to see each change before merge&lt;/h2&gt;
&lt;p&gt;Netlify builds a Deploy Preview for pull requests in a connected GitHub, GitLab, Bitbucket, or Azure DevOps repository, and Vercel creates a preview deployment for every push to a non-production branch and every pull request. Each preview gets its own URL. Your team can click through the change, and I can check what a crawler will see on it: titles, canonicals, structured data, status codes, and rendered content. Netlify adds an &lt;code&gt;X-Robots-Tag: noindex&lt;/code&gt; header to its previews, which asks search engines to leave them out of results. For anything sensitive, Netlify recommends password protection on top.&lt;/p&gt;

&lt;h2&gt;Keep secrets out of the repository&lt;/h2&gt;
&lt;p&gt;API keys, database passwords, and tokens belong in your host's environment variables, never in a committed file. Anyone you give read access can read every file and its history, so a key committed two years ago and deleted since is still visible to them. Two habits protect you:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Turn on push protection.&lt;/strong&gt; GitHub's push protection is a secret scanning feature that blocks pushes containing recognized credentials before they reach the repository, from the command line, the web interface, and the API.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Rotate anything that ever leaked.&lt;/strong&gt; If secret scanning finds an old key in the history, revoke it and issue a new one before you invite anyone in.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;An SEO needs no production credentials to do this work. If someone asks for them, ask why.&lt;/p&gt;

&lt;h2&gt;The CMS only option&lt;/h2&gt;
&lt;p&gt;Some teams can't open the codebase, for contractual or security reasons. That still leaves plenty of room. Give the consultant an editor role in the CMS for content, titles, and descriptions, Search Console access, and the preview URLs. They write tickets with the evidence and the acceptance criteria, and your developer makes the code changes. It's slower, and it works.&lt;/p&gt;

&lt;h2&gt;Remove access when the engagement ends&lt;/h2&gt;
&lt;p&gt;Put the end date in your own calendar on the day you grant access. When the work finishes, remove the outside collaborator from the repository, take them out of Search Console and the hosting dashboard, and revoke any deploy keys or tokens created for the engagement.&lt;/p&gt;

&lt;h2&gt;A checklist for the day you say yes&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;Invite the consultant as an outside collaborator on the website repository only.&lt;/li&gt;
  &lt;li&gt;Assign Read on GitHub, or Reporter on GitLab.&lt;/li&gt;
  &lt;li&gt;Confirm deploy previews are switched on for pull requests.&lt;/li&gt;
  &lt;li&gt;Check that push protection or secret scanning is enabled and the history is clean.&lt;/li&gt;
  &lt;li&gt;Add a ruleset or branch protection requiring a reviewed pull request on main.&lt;/li&gt;
  &lt;li&gt;Move to Write, or Developer, only once that rule is in place.&lt;/li&gt;
  &lt;li&gt;Set a reminder to remove every grant at the end date.&lt;/li&gt;
&lt;/ol&gt;

&lt;p class="related"&gt;This is the setup I ask for at the start of an &lt;a href="https://gerriscorp.com/services/seo-engineering/"&gt;SEO engineering&lt;/a&gt; engagement. For how I work once I'm in, see &lt;a href="https://gerriscorp.com/services/developers/"&gt;developer specs and QA&lt;/a&gt; and the guide to &lt;a href="https://gerriscorp.com/guides/ai-coding-assistants/"&gt;working safely with AI coding assistants&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/organizations/managing-user-access-to-your-organizations-repositories/managing-repository-roles/repository-roles-for-an-organization"&gt;GitHub Docs: Repository roles for an organization&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/repositories/configuring-branches-and-merges-in-your-repository/managing-rulesets/about-rulesets"&gt;GitHub Docs: About rulesets&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/code-security/secret-scanning/introduction/about-push-protection"&gt;GitHub Docs: About push protection&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.gitlab.com/user/permissions/"&gt;GitLab Docs: Roles and permissions&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.netlify.com/site-deploys/deploy-previews/"&gt;Netlify Docs: Deploy Previews&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://vercel.com/docs/deployments/preview-deployments"&gt;Vercel Docs: Preview deployments&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>A Next.js SEO checklist for the App Router</title>
    <link>https://gerriscorp.com/guides/nextjs-seo/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/nextjs-seo/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>A Next.js App Router SEO checklist: metadata, canonicals, title templates, streamed metadata, sitemap.ts, robots.ts, 308 redirects, real 404s, and schema.</description>
    <content:encoded>&lt;h1&gt;A Next.js SEO checklist for the App Router&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Next.js gives you everything a search engine needs, as code: metadata, sitemaps, robots rules, redirects, and status codes. Every one of them can also be left at a default that quietly hurts you. This is the list I work through on every Next.js site, in the order I check it.&lt;/p&gt;

&lt;h2&gt;1. Every route sets its own title and description&lt;/h2&gt;
&lt;p&gt;In the App Router, metadata comes from a &lt;code&gt;metadata&lt;/code&gt; object or a &lt;code&gt;generateMetadata&lt;/code&gt; function exported from a layout or page. Static pages can use the object. Pages built from data, such as products, posts, and locations, need &lt;code&gt;generateMetadata&lt;/code&gt; so each URL gets its own title and description from its own record.&lt;/p&gt;
&lt;p&gt;Both exports work only in Server Components, which keeps the tags in the server's HTML. If a page needs client-side features, keep &lt;code&gt;page.tsx&lt;/code&gt; as a Server Component and move the interactive parts into a child component.&lt;/p&gt;

&lt;h2&gt;2. The title template does what you meant&lt;/h2&gt;
&lt;p&gt;A &lt;code&gt;title.template&lt;/code&gt; in the root layout, such as &lt;code&gt;'%s | Brand'&lt;/code&gt;, adds a suffix to every child page's title, and &lt;code&gt;title.default&lt;/code&gt; fills in for pages that set none. That's convenient, and it's also how a site ends up with 2,000 pages titled with the brand name alone. Crawl the site and sort titles by frequency: any title used more than a few times is a page that fell through to the default.&lt;/p&gt;

&lt;h2&gt;3. Canonicals are set per route, with a metadataBase&lt;/h2&gt;
&lt;p&gt;Set &lt;code&gt;metadataBase&lt;/code&gt; once in the root layout, then give each route &lt;code&gt;alternates.canonical&lt;/code&gt;. Next.js combines the two into a full URL. Check the output on routes with query strings and on paginated lists, where a template canonical copied from the parent layout can point every page at page one.&lt;/p&gt;

&lt;h2&gt;4. You know what crawlers receive when metadata streams&lt;/h2&gt;
&lt;p&gt;When &lt;code&gt;generateMetadata&lt;/code&gt; waits on request time data, Next.js can stream the page first and append the metadata tags to the body once they resolve. It does that for crawlers it expects to run JavaScript, such as Googlebot, and blocks rendering to put the tags in the head for a list of HTML limited bots, including Bingbot, social preview bots, and several others. GPTBot, ClaudeBot, and PerplexityBot are absent from the default list in the Next.js source. If AI crawler visibility matters to you, fetch your key pages with those user agents and check where the title and canonical land. The &lt;code&gt;htmlLimitedBots&lt;/code&gt; option in the Next.js config lets you add them, at some cost in response time.&lt;/p&gt;

&lt;h2&gt;5. The sitemap is generated, complete, and honest&lt;/h2&gt;
&lt;p&gt;An &lt;code&gt;app/sitemap.ts&lt;/code&gt; file that returns an array of URLs produces &lt;code&gt;/sitemap.xml&lt;/code&gt;. Generate it from the same data source as the routes, include only canonical, indexable URLs that return 200, and set &lt;code&gt;lastModified&lt;/code&gt; from the content's real update date rather than the build time. Large sites can split into several sitemaps. Compare the sitemap with a crawl: every URL in one should be in the other.&lt;/p&gt;

&lt;h2&gt;6. robots.ts says what you mean, for each crawler&lt;/h2&gt;
&lt;p&gt;An &lt;code&gt;app/robots.ts&lt;/code&gt; file generates &lt;code&gt;/robots.txt&lt;/code&gt;, with rules per user agent and a sitemap line. Decide deliberately about AI crawlers, and make sure preview and staging deployments disallow everything while production allows what you intend. A single environment variable deciding which robots file ships is a common source of a production site that tells every crawler to stay out.&lt;/p&gt;

&lt;h2&gt;7. Redirects are permanent where you mean permanent&lt;/h2&gt;
&lt;p&gt;Redirects in &lt;code&gt;next.config.js&lt;/code&gt; take a &lt;code&gt;source&lt;/code&gt;, a &lt;code&gt;destination&lt;/code&gt;, and a &lt;code&gt;permanent&lt;/code&gt; flag. With &lt;code&gt;permanent: true&lt;/code&gt; Next.js returns a 308, the permanent status that tells browsers and search engines to remember the move. With &lt;code&gt;false&lt;/code&gt; it returns a 307, a temporary redirect that nothing caches. A migration map written with &lt;code&gt;permanent: false&lt;/code&gt; tells search engines to keep the old URLs. Check every rule's flag, and test the map against a crawl of the old site.&lt;/p&gt;

&lt;h2&gt;8. Missing pages return a real 404&lt;/h2&gt;
&lt;p&gt;Calling &lt;code&gt;notFound()&lt;/code&gt; renders the not-found page, returns a 404, and adds a noindex tag. One catch: if the check runs inside a Suspense boundary after the response has started streaming, the status is already sent, so the page keeps its 200 and relies on the noindex tag. Run the existence check before streaming begins on routes where a real 404 matters, such as products and posts that get deleted. Then crawl a list of deliberately broken URLs and confirm the status codes.&lt;/p&gt;

&lt;h2&gt;9. JSON-LD is rendered in the component&lt;/h2&gt;
&lt;p&gt;The Next.js recommendation is a plain &lt;code&gt;&amp;lt;script type="application/ld+json"&amp;gt;&lt;/code&gt; in the layout or page, built from the same data as the visible content. Escape the &lt;code&gt;&amp;lt;&lt;/code&gt; character in the JSON, as the Next.js guide shows, so content from a CMS can't inject script. Validate a sample from each template with the Rich Results Test or the Schema Markup Validator.&lt;/p&gt;

&lt;h2&gt;10. Navigation uses real links&lt;/h2&gt;
&lt;p&gt;The &lt;code&gt;Link&lt;/code&gt; component extends the HTML anchor element, so it produces an &lt;code&gt;&amp;lt;a href&amp;gt;&lt;/code&gt; that crawlers can follow. Navigation built from buttons with &lt;code&gt;router.push&lt;/code&gt; in a click handler produces no link at all. Search the codebase for click handlers that navigate, and crawl with JavaScript rendering off to see which pages are reachable from plain HTML.&lt;/p&gt;

&lt;h2&gt;11. The rendered page matches the server page&lt;/h2&gt;
&lt;p&gt;Finally, compare the raw HTML with the rendered DOM on each template, and run the pages in headless Chrome to collect console errors. Hydration mismatches, content that loads only after interaction, and client-side redirects all show up here. The &lt;a href="https://gerriscorp.com/guides/hydration-errors/"&gt;guide to React hydration errors&lt;/a&gt; covers the most common fault.&lt;/p&gt;

&lt;h2&gt;Make the checklist permanent&lt;/h2&gt;
&lt;p&gt;Every item above can become an automated check in the build, so a future change can't undo it. That's the subject of &lt;a href="https://gerriscorp.com/guides/seo-build-checks/"&gt;adding SEO tests to your website's build&lt;/a&gt;.&lt;/p&gt;

&lt;p class="related"&gt;I work through this list, and fix what it finds, as part of &lt;a href="https://gerriscorp.com/services/seo-engineering/"&gt;SEO engineering&lt;/a&gt; for Next.js and React sites. If you're bringing me into your repository, start with &lt;a href="https://gerriscorp.com/guides/repository-access/"&gt;how to give an outside SEO safe access&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://nextjs.org/docs/app/api-reference/functions/generate-metadata"&gt;Next.js: generateMetadata&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://nextjs.org/docs/app/api-reference/file-conventions/metadata/sitemap"&gt;Next.js: sitemap.xml&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://nextjs.org/docs/app/api-reference/file-conventions/metadata/robots"&gt;Next.js: robots.txt&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://nextjs.org/docs/app/api-reference/config/next-config-js/redirects"&gt;Next.js: redirects&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://nextjs.org/docs/app/api-reference/functions/not-found"&gt;Next.js: notFound&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://nextjs.org/docs/app/guides/json-ld"&gt;Next.js: JSON-LD&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://nextjs.org/docs/app/api-reference/components/link"&gt;Next.js: Link component&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://github.com/vercel/next.js/blob/canary/packages/next/src/shared/lib/router/utils/html-bots.ts"&gt;Next.js source: HTML limited bot list&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>The SEO tools for a website migration, and the job each one does</title>
    <link>https://gerriscorp.com/guides/migration-tools/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/migration-tools/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>The tools that carry a website migration safely: crawlers, Search Console, redirect testing, crawl comparison, and analytics, with the job each one does.</description>
    <content:encoded>&lt;h1&gt;The SEO tools for a website migration, and the job each one does&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;No single product migrates a website for you. A migration is a handful of jobs, and each has a tool that does it well: an inventory, a redirect map, a pre-launch comparison, a launch day test, and weeks of watching. Here is the kit I use, matched to the job, with the free options where they exist.&lt;/p&gt;

&lt;h2&gt;Job 1: list every URL that matters&lt;/h2&gt;
&lt;p&gt;The inventory is the foundation, and it comes from several tools at once, since no single source sees everything.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;A crawler.&lt;/strong&gt; Screaming Frog SEO Spider, Sitebulb, or a similar desktop crawler finds every URL linked from the site. Screaming Frog's free version stops at 500 URLs; a licence removes the limit.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Google Search Console.&lt;/strong&gt; The Performance report, exported by page over the last twelve to sixteen months, lists the URLs that actually earn impressions, including ones no longer linked from anywhere.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Analytics.&lt;/strong&gt; Landing pages from GA4 or whatever you run, for the URLs people arrive on from every channel.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;A backlink tool.&lt;/strong&gt; Ahrefs, Semrush, or Majestic for the URLs other sites link to, which carry value even if they earn no traffic.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The XML sitemaps&lt;/strong&gt; and, if you can get them, &lt;strong&gt;server logs.&lt;/strong&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Merge them into one spreadsheet with a row per URL and columns for clicks, links, and sessions. That's the list everything else is tested against.&lt;/p&gt;

&lt;h2&gt;Job 2: build and store the redirect map&lt;/h2&gt;
&lt;p&gt;The map pairs each old URL with its closest new equivalent. A spreadsheet is fine for drafting. The finished map belongs wherever the new site serves redirects: the server configuration, the host's redirect rules file on Netlify, Vercel, or Cloudflare, or the framework's own configuration. In Next.js, for example, redirects live in &lt;code&gt;next.config.js&lt;/code&gt;, and each rule's &lt;code&gt;permanent&lt;/code&gt; flag decides between a 308 and a 307. Keeping the map in version control means it gets reviewed, and nobody loses it.&lt;/p&gt;
&lt;p&gt;For large sites, pattern rules cover whole sections in one line, and exceptions handle the pages that don't fit the pattern. Test the patterns before launch, because one bad regex can redirect thousands of pages to the wrong place.&lt;/p&gt;

&lt;h2&gt;Job 3: compare staging with the live site&lt;/h2&gt;
&lt;p&gt;Before launch, crawl both the live site and the staging site with the same settings, then compare them. Screaming Frog's crawl comparison does this with URL Mapping, which uses a regex to pair the staging hostname's URLs with the live ones. You see, template by template, which titles, descriptions, headings, canonicals, structured data, and internal links changed. It needs a licence and database storage mode.&lt;/p&gt;
&lt;p&gt;Check the staging site's robots.txt and meta robots tags at the same time. Staging is supposed to be blocked from search. Production isn't, and the block coming across at launch is one of the most common migration disasters.&lt;/p&gt;

&lt;h2&gt;Job 4: test every redirect on launch day&lt;/h2&gt;
&lt;p&gt;Once the redirects are live, crawl the old URL list in Screaming Frog's list mode with Always Follow Redirects switched on, as Screaming Frog's migration guide describes. The All Redirects report shows each old URL, where it finally lands, and every hop in between, so you can spot 404s, chains, loops, and temporary redirects in minutes. Run it again a day later, after caches have cleared.&lt;/p&gt;
&lt;p&gt;For a quick spot check without a crawler, &lt;code&gt;curl -sI&lt;/code&gt; on the command line shows the status code and the Location header for any single URL.&lt;/p&gt;

&lt;h2&gt;Job 5: tell Google, then watch&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Search Console sitemaps.&lt;/strong&gt; Google's site move guidance is to submit the sitemaps for both the old and new URLs, then watch the indexed count move from one to the other. Warnings that the old URLs redirect are expected.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The Change of Address tool,&lt;/strong&gt; for moves from one domain or subdomain to another, used after the redirects are in place. It doesn't apply to HTTP to HTTPS, www changes, or path changes within one site.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;URL Inspection,&lt;/strong&gt; on a sample of important old and new URLs, to see Google's chosen canonical and indexing status.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The Performance report with a date comparison,&lt;/strong&gt; by page, to catch any section that loses more than the expected dip.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Bing Webmaster Tools,&lt;/strong&gt; the equivalent for Bing, which also supplies results to other search and AI products.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Server logs,&lt;/strong&gt; which Google also recommends checking during a move, to confirm Googlebot is crawling the new URLs and to catch error codes your crawler missed.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Job 6: prove the move held&lt;/h2&gt;
&lt;p&gt;A few weeks after launch, crawl the new site again and compare it with the pre-launch staging crawl, and rerun the old URL list. Anything that drifted, such as a redirect someone edited or a template change that dropped schema, shows up as a difference. If the numbers have fallen and stayed down, the &lt;a href="https://gerriscorp.com/guides/migration-recovery/"&gt;migration recovery guide&lt;/a&gt; covers what to do next.&lt;/p&gt;

&lt;h2&gt;What a tool can't do&lt;/h2&gt;
&lt;p&gt;Every tool above reports facts. None of them decides which new page is the right home for an old one, whether a thin page should be merged or kept, or which redirect rule is safe on a site with forty thousand URLs. That judgment is the migration work itself, and it's why the inventory and the map deserve the most time.&lt;/p&gt;

&lt;p class="related"&gt;I run migrations with this kit as &lt;a href="https://gerriscorp.com/services/migrations/"&gt;website migration SEO&lt;/a&gt;. For the full sequence, follow the &lt;a href="https://gerriscorp.com/guides/migration-checklist/"&gt;website migration checklist&lt;/a&gt;, and for redirect rules in depth, see &lt;a href="https://gerriscorp.com/guides/301-redirects/"&gt;301 redirects explained&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/site-move-with-url-changes"&gt;Google Search Central: Site moves with URL changes&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/webmasters/answer/9370220"&gt;Search Console Help: Change of Address tool&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.screamingfrog.co.uk/seo-spider/tutorials/how-to-use-the-seo-spider-in-a-site-migration/"&gt;Screaming Frog: How to use the SEO Spider in a site migration&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.screamingfrog.co.uk/seo-spider/tutorials/how-to-compare-crawls/"&gt;Screaming Frog: How to compare crawls&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://nextjs.org/docs/app/api-reference/config/next-config-js/redirects"&gt;Next.js: redirects&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Traffic fell after a site migration: how to find what broke and get it back</title>
    <link>https://gerriscorp.com/guides/migration-recovery/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/migration-recovery/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>A recovery plan for a website migration that lost search traffic: rebuild the old URL list, test every redirect, compare old and new pages, and fix in order.</description>
    <content:encoded>&lt;h1&gt;Traffic fell after a site migration: how to find what broke and get it back&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;The new site launched, and organic traffic dropped and stayed down. Some dip is normal while Google recrawls. A drop that's still there after a month has a cause, and it's almost always one of a short list. Here is the recovery routine I run, in the order that finds the money fastest.&lt;/p&gt;

&lt;h2&gt;First, separate normal turbulence from real loss&lt;/h2&gt;
&lt;p&gt;Google says to expect ranking fluctuations during any significant move, and that for a medium sized site it can take a few weeks or more before the new URLs gradually replace the old ones in results, longer for large sites. So the first question is shape. A dip that's recovering week over week is the move settling. A cliff that hasn't moved, or losses concentrated in one section, is a fault.&lt;/p&gt;
&lt;p&gt;In Search Console, compare the four weeks after launch with the four weeks before, and look at pages rather than totals. Sort by lost clicks. The top twenty URLs usually account for most of the loss, and they tell you where to dig.&lt;/p&gt;

&lt;h2&gt;Rebuild the list of old URLs&lt;/h2&gt;
&lt;p&gt;Every recovery starts with the URL inventory the migration should have started with. If nobody kept one, rebuild it from every source you can reach:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;The Search Console Performance report for the months before launch, exported by page.&lt;/li&gt;
  &lt;li&gt;Analytics landing pages for the same period.&lt;/li&gt;
  &lt;li&gt;The old XML sitemaps, from a backup or from the Wayback Machine.&lt;/li&gt;
  &lt;li&gt;A backlink tool's list of linked URLs on your domain.&lt;/li&gt;
  &lt;li&gt;Server logs, if the old host kept them.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Merge them and remove duplicates. This list, weighted by the traffic and links each URL carried, is what you'll test.&lt;/p&gt;

&lt;h2&gt;Test every old URL's redirect&lt;/h2&gt;
&lt;p&gt;Screaming Frog's own migration guide describes the method. Switch to list mode, turn on Always Follow Redirects under Config &amp;gt; Spider &amp;gt; Advanced, upload the old URLs, and crawl them. The All Redirects export then shows each old address, the final address it lands on, the type of each hop, and whether a chain or a loop is involved.&lt;/p&gt;
&lt;p&gt;Read the export for these failures, in this order:&lt;/p&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;404s and soft 404s.&lt;/strong&gt; Old URLs with no redirect at all. On a big site these are usually whole patterns, such as an old category path or a dropped file extension.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Redirects to the home page or a generic category.&lt;/strong&gt; Google treats a redirect to an unrelated page much like a missing page, and the old URL's value doesn't carry over.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Temporary redirects.&lt;/strong&gt; Google's documentation says a permanent redirect shows the new target in results, while a temporary one keeps showing the source. A migration built on 302s or 307s is telling Google to keep the old URLs.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Chains and loops.&lt;/strong&gt; Each extra hop slows crawling, and loops never resolve.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;JavaScript and meta refresh redirects.&lt;/strong&gt; Google ranks server-side redirects as the most reliable way to signal a move. Anything else should be replaced with one.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;Compare old pages with their new versions&lt;/h2&gt;
&lt;p&gt;A redirect that lands on the right URL still loses traffic if the new page is weaker. If you have a crawl from before launch, compare it with a crawl of the new site in Screaming Frog's comparison mode, using URL Mapping to pair old and new addresses. Without one, archived copies of the top pages will do. Check each pair for:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;Titles, descriptions, and H1s that were rewritten, shortened, or replaced with a template default.&lt;/li&gt;
  &lt;li&gt;Body copy cut down, split across tabs, or moved behind a script.&lt;/li&gt;
  &lt;li&gt;Structured data that existed before and doesn't now.&lt;/li&gt;
  &lt;li&gt;Internal links that used to point at the page and no longer do, which shows up as greater crawl depth.&lt;/li&gt;
  &lt;li&gt;Canonical tags pointing at staging, at the old domain, or at the wrong page.&lt;/li&gt;
  &lt;li&gt;A noindex or a robots.txt rule carried over from the staging site.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Check that Google can see the new site the way you think&lt;/h2&gt;
&lt;p&gt;Use URL Inspection on a sample of the top lost URLs and their new equivalents. Look at Google's chosen canonical, the indexing status, and the rendered page. Check the old sitemap and the new one in the Sitemaps report: Google's guidance is to submit both during a move, and to expect the indexed count on the old one to fall as the new one rises. If the old URLs still show as indexed weeks later, Google isn't following your redirects as you intended.&lt;/p&gt;
&lt;p&gt;For a move to a new domain, confirm the Change of Address request went in from the old property. Google's help is specific about when it applies: a move between domains or subdomains. HTTP to HTTPS, www to non-www, and path changes within one site don't use it.&lt;/p&gt;

&lt;h2&gt;Fix in order of value&lt;/h2&gt;
&lt;p&gt;Recovery work goes fastest when it follows the money. Fix the redirects for the twenty URLs that lost the most first, then the patterns that cover hundreds of URLs at once, then content and template gaps on the pages that matter, then the long tail. Keep every redirect in place: Google recommends at least a year, and longer is better for people following old links.&lt;/p&gt;
&lt;p&gt;After each round of fixes, run the list crawl again and confirm the failures are gone, then watch the affected pages in Search Console. Recovered pages usually come back over weeks, as Google recrawls them. I make no promise about full recovery, because rankings also move for reasons unrelated to the migration, but a site whose old URLs all resolve cleanly to equal or better pages has removed every cause it controls.&lt;/p&gt;

&lt;h2&gt;Keep the evidence&lt;/h2&gt;
&lt;p&gt;Write down what you found, what you fixed, and when, with the before-and-after crawl exports. If traffic doesn't come back in the expected window, that record is what lets the next person separate a migration fault from an algorithm update or a seasonal dip.&lt;/p&gt;

&lt;p class="related"&gt;I handle this as part of &lt;a href="https://gerriscorp.com/services/migrations/"&gt;website migration SEO&lt;/a&gt;, and for inherited projects that went badly, &lt;a href="https://gerriscorp.com/services/rescue/"&gt;project rescue&lt;/a&gt;. To prevent the next one, use the &lt;a href="https://gerriscorp.com/guides/migration-checklist/"&gt;website migration checklist&lt;/a&gt; and the &lt;a href="https://gerriscorp.com/guides/migration-tools/"&gt;tools I use for a website migration&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/site-move-with-url-changes"&gt;Google Search Central: Site moves with URL changes&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/301-redirects"&gt;Google Search Central: Redirects and Google Search&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/webmasters/answer/9370220"&gt;Search Console Help: Change of Address tool&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.screamingfrog.co.uk/seo-spider/tutorials/how-to-use-the-seo-spider-in-a-site-migration/"&gt;Screaming Frog: How to use the SEO Spider in a site migration&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.screamingfrog.co.uk/seo-spider/tutorials/how-to-compare-crawls/"&gt;Screaming Frog: How to compare crawls&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Hype or real shift? Eight tests for judging a new technology wave</title>
    <link>https://gerriscorp.com/guides/hype-or-shift/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/hype-or-shift/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>Eight tests for telling a lasting technology shift from a hype cycle, drawn from online communities, blogs, social media, crypto, and the AI wave of 2026.</description>
    <content:encoded>&lt;h1&gt;Hype or real shift? Eight tests for judging a new technology wave&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Every few years a technology arrives that everyone says changes everything. Sometimes it does. I've worked through enough of these waves to have a working method for telling the difference: online communities, blogs, social media, cryptocurrency, and now AI. The tests below are the ones I use with clients who have to decide what to spend on this year, while the noise is loudest.&lt;/p&gt;

&lt;h2&gt;Two ideas worth keeping in your pocket&lt;/h2&gt;
&lt;p&gt;Roy Amara, who led the Institute for the Future, gave his name to the most useful sentence about technology forecasting, now called Amara's law:&lt;/p&gt;
&lt;blockquote&gt;&lt;p&gt;We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.&lt;/p&gt;&lt;/blockquote&gt;
&lt;p&gt;Gartner's hype cycle draws the same shape as a curve: an innovation trigger, a peak of inflated expectations, a trough of disillusionment, a slope of enlightenment, and a plateau of productivity. The model has its critics, and plenty of technologies never climb out of the trough. As a reminder that loud and lasting are different measurements, both earn their place.&lt;/p&gt;

&lt;h2&gt;What I've watched up close&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Online communities.&lt;/strong&gt; I was on The WELL and ran MemeSpace long before "community manager" was a job. The tools changed completely. The thing people wanted, a place to talk with others who care about the same subject, never changed at all, and it's why Reddit threads now feed AI answers.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Blogs.&lt;/strong&gt; I ran blogs on Movable Type and helped develop blogger outreach in 2006. The format peaked and faded as a fashion, and the underlying shift, anyone can publish and be found, became permanent.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Social media.&lt;/strong&gt; I was on Facebook in 2006 and Twitter in 2007, and I've watched platforms rise, dominate, and lose their grip. The lasting change was in how news and reputations move.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Cryptocurrency.&lt;/strong&gt; I co-founded and edited a daily news site covering crypto and blockchain, based in Tallinn, and interviewed the people building it. Covering a market every day teaches you to separate press releases from working software.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;AI.&lt;/strong&gt; I use AI tools every working day and have built real software with them. I also catch them inventing things every week.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Eight tests&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Does it work when it's boring?&lt;/strong&gt; Demos show the best case. Ask what a normal Tuesday looks like for someone who has used it for six months.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Who pays, and for what?&lt;/strong&gt; Lasting shifts have customers paying for a result. Hype is often funded by investors paying for growth.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Does it survive the subsidy?&lt;/strong&gt; Many new tools are priced below cost to win users. Ask what the work costs at a price the vendor can sustain.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Do practitioners keep using it?&lt;/strong&gt; Watch the people who do the work after the novelty wears off. In Stack Overflow's 2025 survey, 84 percent of developers used or planned to use AI tools while more distrusted their accuracy than trusted it. That combination, heavy use with clear eyed doubt, is what a real tool looks like in its awkward years.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Are the objections fixed or renamed?&lt;/strong&gt; Real shifts answer their critics with better products. Hype answers them with new vocabulary.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Does it change a cost line?&lt;/strong&gt; The waves that lasted made something dramatically cheaper: publishing, distribution, communication. Find the line in your budget that it shrinks.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Can you leave?&lt;/strong&gt; Lock in tends to grow on the peak. Prefer tools that let you take your data with you.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;What does the skeptic in the room say?&lt;/strong&gt; Find the most knowledgeable doubter you can and listen to the strongest version of the case against.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;How to act while you're unsure&lt;/h2&gt;
&lt;p&gt;Be early and small. Try new tools on real work at low cost, keep what proves itself, and drop the rest without ceremony. Write down what you learn as you go, so the decision a year from now rests on your own evidence. Avoid both betting the company on the peak and waiting for the plateau, because by then your competitors have already learned what you're just starting to.&lt;/p&gt;

&lt;h2&gt;The gray beard in the room&lt;/h2&gt;
&lt;p&gt;Here's what I've learned about my own role. The rarest person in a meeting about new technology is someone who's genuinely excited by it and still asks for the receipts. I'm both, every time. Companies bring me in as the experienced voice for a young team, as the technical skeptic on a vendor call, as an interim or white-label executive who can speak for the technology, and as a guest on shows, panels, and interviews about AI and search. If you need someone to separate the real from the loud, start with my &lt;a href="https://gerriscorp.com/services/advisory/"&gt;advisory work&lt;/a&gt;, read &lt;a href="https://gerriscorp.com/guides/ai-vendor-pitch/"&gt;how I vet an AI vendor's pitch&lt;/a&gt;, or &lt;a href="https://gerriscorp.com/contact/"&gt;get in touch&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://en.wikipedia.org/wiki/Roy_Amara"&gt;Wikipedia: Roy Amara (Amara&amp;#x27;s law)&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://en.wikipedia.org/wiki/Gartner_hype_cycle"&gt;Wikipedia: Gartner hype cycle&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://survey.stackoverflow.co/2025/ai"&gt;Stack Overflow: 2025 Developer Survey, AI&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>React hydration errors and what they mean for SEO</title>
    <link>https://gerriscorp.com/guides/hydration-errors/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/hydration-errors/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>What React error 418 and other hydration mismatches are, why Google usually still indexes the page, what they cost in speed and analytics, and how to find them.</description>
    <content:encoded>&lt;h1&gt;React hydration errors and what they mean for SEO&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;If your browser console on a React or Next.js site shows "Minified React error #418," the page has a hydration error. It rarely stops Google from indexing the content, so it's easy to leave alone. It still costs you layout stability, responsiveness, and trustworthy analytics, and it usually points at a markup fault worth fixing at the source.&lt;/p&gt;

&lt;h2&gt;What hydration is&lt;/h2&gt;
&lt;p&gt;A server rendered React page arrives as finished HTML, so visitors and crawlers see content immediately. Then React loads in the browser and "attaches" itself to that HTML, wiring up buttons, menus, and forms. React calls this hydration: it turns the server's HTML snapshot into an interactive app.&lt;/p&gt;
&lt;p&gt;For that to work, the HTML React expects to find has to match what the server sent. React's documentation says it expects the rendered content to be identical, and that you should treat mismatches as bugs and fix them.&lt;/p&gt;

&lt;h2&gt;What error #418 means&lt;/h2&gt;
&lt;p&gt;Production builds of React strip long error messages to save bytes and print a number instead. React's error page decodes #418 as: hydration failed because the server rendered HTML didn't match the client, and as a result the tree will be regenerated on the client. In plain terms, React gave up on the server's HTML for that part of the page and built it again in the browser.&lt;/p&gt;

&lt;h2&gt;What it means for search&lt;/h2&gt;
&lt;p&gt;Google crawls the server's response first, and server rendering means the content is already in that HTML. Google's JavaScript SEO documentation still recommends server-side or pre-rendering for exactly that reason: it's faster for users and crawlers, and not every bot runs JavaScript. A hydration error happens after that HTML is delivered, so the words on the page usually still reach the index.&lt;/p&gt;
&lt;p&gt;The costs show up elsewhere:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Layout shift.&lt;/strong&gt; When React throws away server markup and re-renders, elements can jump, which counts against Cumulative Layout Shift.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Slower interaction.&lt;/strong&gt; Rebuilding part of the page on the main thread can delay the response to the visitor's first tap or click, which shows up in Interaction to Next Paint.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Broken widgets.&lt;/strong&gt; Components in the rebuilt section lose whatever state they had, so a menu, carousel, or form can reset or stop responding.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Messy analytics.&lt;/strong&gt; Tags that fire on load can fire twice, or against elements that have since been replaced.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Head tags are a separate concern. In Next.js, metadata set with &lt;code&gt;generateMetadata&lt;/code&gt; or a &lt;code&gt;metadata&lt;/code&gt; export is resolved on the server and sent as HTML, so titles and canonicals don't depend on hydration succeeding.&lt;/p&gt;

&lt;h2&gt;The usual causes&lt;/h2&gt;
&lt;p&gt;React's own list covers most of what I find on real sites:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Invalid nesting.&lt;/strong&gt; A &lt;code&gt;div&lt;/code&gt; inside a &lt;code&gt;p&lt;/code&gt;, or a link inside a link. The browser's HTML parser rearranges invalid nesting as it reads the page, so the DOM no longer matches the tree React rendered on the server.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Unbalanced HTML from a CMS.&lt;/strong&gt; Rich text injected with &lt;code&gt;dangerouslySetInnerHTML&lt;/code&gt; that contains a stray closing tag or an unclosed element. The browser repairs it on the way in, and React finds a different structure from the one it expected.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Dates and locales.&lt;/strong&gt; A date formatted on the server in one time zone or locale and on the client in another.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Random values.&lt;/strong&gt; IDs or keys generated with &lt;code&gt;Math.random()&lt;/code&gt; or &lt;code&gt;Date.now()&lt;/code&gt;, which differ between the two renders.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Browser checks during render.&lt;/strong&gt; A branch like &lt;code&gt;typeof window !== 'undefined'&lt;/code&gt; that renders one thing on the server and another in the browser.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Browser extensions.&lt;/strong&gt; Extensions that edit the page before React loads. These show up only for some visitors and aren't yours to fix, which is why it's worth reproducing in a clean browser profile first.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Finding them across a whole site&lt;/h2&gt;
&lt;p&gt;The browser console shows you one page at a time, and hydration errors often live on a handful of templates or on pages whose CMS content happens to contain bad markup. Spot checking misses them. I script headless Chrome with Playwright to load every URL in the sitemap or the crawl, collect each console error with the URL it came from, and group the results by template. A few hundred pages run in minutes.&lt;/p&gt;
&lt;p&gt;The production build tells you only that a mismatch happened. To see where, run the same page against a development build, where React warns about each mismatch with its full message instead of a number. Once the list is grouped, the cause for each template is usually obvious within a few minutes of reading.&lt;/p&gt;

&lt;h2&gt;Fixing them at the source&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Correct the nesting&lt;/strong&gt; in the component, using a &lt;code&gt;div&lt;/code&gt; or &lt;code&gt;span&lt;/code&gt; where a paragraph can't legally contain the child.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Repair the content&lt;/strong&gt; that carries unbalanced HTML, and sanitize rich text on the way into the CMS or at render time.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Make server and client agree&lt;/strong&gt; on dates by fixing the time zone and locale, and generate IDs with React's &lt;code&gt;useId&lt;/code&gt; rather than random numbers.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Move browser only logic&lt;/strong&gt; into an effect that runs after hydration.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Silence only the unavoidable.&lt;/strong&gt; React offers &lt;code&gt;suppressHydrationWarning&lt;/code&gt; for a single element whose text must differ, such as a timestamp. Use it on that element and nowhere else.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Then stop the problem coming back. A small check in the build that parses every rendered page and fails on unbalanced tags or invalid nesting catches the CMS content that would have caused the next one.&lt;/p&gt;

&lt;p class="related"&gt;I find and fix these as part of &lt;a href="https://gerriscorp.com/services/seo-engineering/"&gt;SEO engineering&lt;/a&gt; and &lt;a href="https://gerriscorp.com/services/crawler-visibility/"&gt;rendering diagnosis&lt;/a&gt;. For the wider rendering picture, see the &lt;a href="https://gerriscorp.com/guides/javascript-seo/"&gt;JavaScript SEO guide&lt;/a&gt; and &lt;a href="https://gerriscorp.com/guides/core-web-vitals/"&gt;Core Web Vitals&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://react.dev/reference/react-dom/client/hydrateRoot"&gt;React: hydrateRoot&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://react.dev/errors/418"&gt;React: Minified React error #418&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics"&gt;Google Search Central: Understand the JavaScript SEO basics&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://nextjs.org/docs/app/api-reference/functions/generate-metadata"&gt;Next.js: generateMetadata&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to hire and manage developers when you can't read code yourself</title>
    <link>https://gerriscorp.com/guides/hiring-developers/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/hiring-developers/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>A founder's method for hiring web developers: own the accounts first, interview for judgment, pay for a trial task, review pull requests, and manage AI use.</description>
    <content:encoded>&lt;h1&gt;How to hire and manage developers when you can't read code yourself&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Founders and marketing leads hire developers all the time without being able to judge the work. They pick the most confident interview, hope for the best, and find out six months later what they bought. There's a better way, and none of it requires learning to program. It requires owning the right accounts, asking questions that reveal judgment, and putting every change somewhere you can see it.&lt;/p&gt;

&lt;h2&gt;Own the accounts before you hire anyone&lt;/h2&gt;
&lt;p&gt;Create a GitHub organization in your company's name, and keep the domain registrar, hosting, DNS, email, analytics, and Search Console in accounts your business controls. GitHub's organizations exist for exactly this: shared ownership with per person permissions. Developers get invited in and can be removed cleanly. The most expensive situation I see in rescue work is a business whose site lives in a former contractor's personal account.&lt;/p&gt;

&lt;h2&gt;Write the problem down&lt;/h2&gt;
&lt;p&gt;A one page brief beats any job title. What the site does, what's broken or missing, what platform it runs on, what "done" would look like, and the deadline that actually matters. The brief attracts the right people, and it becomes the yardstick for everything that follows.&lt;/p&gt;

&lt;h2&gt;Interview for judgment&lt;/h2&gt;
&lt;p&gt;You can't grade code in an interview, and you don't need to. You can grade how someone thinks. Ask:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;"Walk me through something you built that's live now."&lt;/strong&gt; Open it together. Strong candidates explain choices and tradeoffs; weak ones list technologies.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;"Tell me about something that broke in production."&lt;/strong&gt; Everyone experienced has a story. Listen for how they found the cause and what they changed so it wouldn't repeat.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;"What would you need from me in week one?"&lt;/strong&gt; Good developers ask for access, the brief, and a decision maker. People who need nothing are planning to guess.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;"How do you use AI coding tools, and how do you check what they write?"&lt;/strong&gt; More on this below; the answer tells you a lot.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;"What part of this brief would you push back on?"&lt;/strong&gt; Someone who agrees with everything will also agree with every bad idea later.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Pay for a small trial task&lt;/h2&gt;
&lt;p&gt;Pick a real, contained piece of work: a page template, a bug, a performance fix. Write the acceptance criteria in plain English, pay for a few hours, and ask for the work as a pull request in your repository. You learn how they communicate, whether they meet the brief, and whether their work is readable to the next person. They learn whether they want the job. Either side can stop cleanly.&lt;/p&gt;

&lt;h2&gt;See every change through pull requests&lt;/h2&gt;
&lt;p&gt;A pull request is a proposed change, shown line by line, with a description and a place for comments, and nothing reaches the live site until someone approves it. Require one for every change, with a sentence in plain English explaining what and why. You won't read the code, but you'll read the descriptions, see the pace of work, and have a complete history if a developer leaves. If you have one technical person you trust, give them reviewer rights.&lt;/p&gt;

&lt;h2&gt;Manage the AI in the room&lt;/h2&gt;
&lt;p&gt;Nearly every developer now works with AI assistants. In Stack Overflow's 2025 Developer Survey, 84 percent of respondents said they use or plan to use AI tools, yet more distrusted the accuracy of those tools (46 percent) than trusted it (33 percent). The most common frustration, cited by 66 percent, was AI solutions that are "almost right, but not quite."&lt;/p&gt;
&lt;p&gt;That gap is your hiring signal. The developer you want uses AI for speed and checks everything it writes: runs it, tests it, and can explain any line in their pull request. The developer to avoid pastes output they can't explain. Ask candidates to walk you through a recent change and why each part is there. If they can't, nobody can.&lt;/p&gt;

&lt;h2&gt;Getting the best from junior developers&lt;/h2&gt;
&lt;p&gt;Junior developers are often quick, eager, and very good with new tools. They need what everyone needs, in larger doses: small, clearly written tickets, acceptance criteria they can test themselves, a reviewer who explains the "why" behind each requested change, and a running decision log so they don't relitigate settled questions. Paired with an experienced reviewer, a junior developer becomes productive fast. Left alone with a vague brief and an AI assistant, they produce confident code that someone will have to rewrite.&lt;/p&gt;

&lt;h2&gt;Red flags&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Work happens on their server or in their personal account.&lt;/li&gt;
  &lt;li&gt;Changes go live with no record of what changed.&lt;/li&gt;
  &lt;li&gt;Status updates are always "almost done."&lt;/li&gt;
  &lt;li&gt;Every problem belongs to the hosting company, the plugin, or the previous developer.&lt;/li&gt;
  &lt;li&gt;Requests for passwords sent by text message instead of invitations to your accounts.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Where I fit&lt;/h2&gt;
&lt;p&gt;I've spent decades on both sides of this table: writing Perl and PHP myself, then specifying, reviewing, and directing developers' work for clients, and now directing AI coding assistants the way a lead directs a team. I sit in on developer interviews as the technical interviewer, write the trial task and its acceptance criteria, review the pull requests, and mentor junior developers through their first months. When you need someone to own the whole picture part time, that's my &lt;a href="https://gerriscorp.com/services/technical-lead/"&gt;fractional technical lead&lt;/a&gt; work; for specs and review alone, see &lt;a href="https://gerriscorp.com/services/developers/"&gt;developer specs and QA&lt;/a&gt;. My guide to &lt;a href="https://gerriscorp.com/guides/repository-access/"&gt;giving an outside expert repository access&lt;/a&gt; covers the permissions side.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://survey.stackoverflow.co/2025/ai"&gt;Stack Overflow: 2025 Developer Survey, AI&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/organizations/collaborating-with-groups-in-organizations/about-organizations"&gt;GitHub Docs: About organizations&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/reviewing-changes-in-pull-requests/about-pull-request-reviews"&gt;GitHub Docs: About pull request reviews&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to fact check a confident AI answer about SEO before you act on it</title>
    <link>https://gerriscorp.com/guides/fact-check-ai-seo/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/fact-check-ai-seo/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>A claim by claim method for checking polished AI answers about SEO, using a real exchange about brand names in page titles and what the evidence supports.</description>
    <content:encoded>&lt;h1&gt;How to fact check a confident AI answer about SEO before you act on it&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;I asked an AI search assistant whether ending every page title with a pipe and the company name hurts a website. The answer came back polished, cited, and full of percentages. When I pushed my own opinion back at it, it declared that opinion “100% true” and invented psychology to support it. The recommendation itself was reasonable. A good share of the evidence was not. This is the routine I use to separate the two before anything reaches a client's site.&lt;/p&gt;

&lt;h2&gt;Turn the answer into a list of claims&lt;/h2&gt;
&lt;p&gt;Paste the answer into a document and give every sentence that asserts a fact its own line. In this exchange the list read: titles are cut off after about 60 characters; Google rewrites more than 75 percent of titles; Google strips pipes and brand names; Google displays the site name on its own; repeated suffixes cause duplication penalties and dilute relevance; searchers read a pipe as a sign of automated spam.&lt;/p&gt;
&lt;p&gt;A sound recommendation can rest on claims that are half wrong. Grading each line separately stops a good conclusion from carrying bad evidence into your report.&lt;/p&gt;

&lt;h2&gt;Grade every claim by where it comes from&lt;/h2&gt;
&lt;p&gt;Each line goes into one of four bins: the platform's own documentation, an independent study with a published method, a vendor test, or nothing traceable. Here is how this exchange graded out.&lt;/p&gt;
&lt;table&gt;
  &lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Claim&lt;/th&gt;&lt;th scope="col"&gt;What I found&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
  &lt;tbody&gt;
    &lt;tr&gt;&lt;td&gt;Titles are cut off at about 60 characters&lt;/td&gt;&lt;td&gt;Google's documentation sets no maximum length and says the displayed title is shortened to fit the device. Sixty characters is an industry habit, not a Google limit.&lt;/td&gt;&lt;/tr&gt;
    &lt;tr&gt;&lt;td&gt;Google rewrites over 75 percent of titles&lt;/td&gt;&lt;td&gt;I could not trace that figure to a primary dataset about pipes or brand names. The best documented study I found, by Zyppy, measured 61.6 percent of 80,959 titles on 2,370 sites, using desktop results from early 2022. Titles longer than 70 characters were rewritten 99.9 percent of the time.&lt;/td&gt;&lt;/tr&gt;
    &lt;tr&gt;&lt;td&gt;Google strips pipes&lt;/td&gt;&lt;td&gt;The same study found pipes removed or replaced in 41.0 percent of titles that used them, often swapped for a dash. Dashes were removed 19.7 percent of the time.&lt;/td&gt;&lt;/tr&gt;
    &lt;tr&gt;&lt;td&gt;Google strips brand names&lt;/td&gt;&lt;td&gt;The study names missing or superfluous brand names as a rewrite trigger and gives no percentage. Google says it may omit a site name that repeats the one already shown.&lt;/td&gt;&lt;/tr&gt;
    &lt;tr&gt;&lt;td&gt;Google shows the site name separately&lt;/td&gt;&lt;td&gt;Supported by Google's documentation.&lt;/td&gt;&lt;/tr&gt;
    &lt;tr&gt;&lt;td&gt;Repeated suffixes cause penalties or crawling problems&lt;/td&gt;&lt;td&gt;Nothing traceable. Google warns that boilerplate titles make results hard to tell apart, which is about usefulness to searchers, not a penalty.&lt;/td&gt;&lt;/tr&gt;
    &lt;tr&gt;&lt;td&gt;Searchers read a pipe as spam&lt;/td&gt;&lt;td&gt;Nothing traceable. I found no study measuring that reaction.&lt;/td&gt;&lt;/tr&gt;
  &lt;/tbody&gt;
&lt;/table&gt;

&lt;h2&gt;Follow every number back to its population&lt;/h2&gt;
&lt;p&gt;A statistic describes the set of things someone counted, at a particular time, in a particular way. Ask who measured it, when, on which device, and what exactly was counted. A rewrite rate for all titles says nothing specific about pipes. A desktop sample from early 2022 may not describe mobile results today. When an assistant cannot show where a number came from, or the source measures something else, drop the number and keep the point in plain words.&lt;/p&gt;

&lt;h2&gt;Distrust agreement that grows when you push&lt;/h2&gt;
&lt;p&gt;The clearest warning in this exchange came after I shared my own view. The follow-up adopted my wording, raised its certainty, and added mechanisms nobody had measured. An assistant that grows more confident because you sounded confident is reflecting you back at yourself. Open a fresh conversation and ask the opposite question: what is the case for keeping the company name in every title? If the answer flips that easily, the evidence was never doing the work.&lt;/p&gt;

&lt;h2&gt;Settle it with your own data&lt;/h2&gt;
&lt;p&gt;For title changes, the deciding evidence sits in your own Search Console. Change one group of pages, leave a comparable group untouched, and compare clicks and click-through rate across several weeks. A controlled result from your own site outranks any general statistic, accurate or invented. My guide to &lt;a href="https://gerriscorp.com/guides/traffic-drop/"&gt;diagnosing a Google traffic drop in Search Console&lt;/a&gt; covers lining changes up with dates, and the one on &lt;a href="https://gerriscorp.com/guides/measure-ai-search/"&gt;measuring AI citations, referrals, and leads&lt;/a&gt; applies the same discipline to AI search reporting.&lt;/p&gt;

&lt;h2&gt;What survived the check&lt;/h2&gt;
&lt;p&gt;After grading, the defensible advice fit in a few lines: write a specific title for each page, include the company name where it helps someone recognize or choose the result, keep titles concise, and measure the change. Dropping a repeated suffix remains a legitimate editorial call. It simply has to be presented as judgment backed by a test, never as escape from a penalty that does not exist.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/title-link"&gt;Google Search Central: Influencing your title links in search results&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://zyppy.com/seo/google-title-rewrites/"&gt;Zyppy: Google rewrites 61% of page title tags, a study of 80,959 titles from early 2022&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to get your email off a blocklist and back into the inbox</title>
    <link>https://gerriscorp.com/guides/email-blocklist-recovery/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/email-blocklist-recovery/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>A recovery plan for a small business whose email lands in spam or bounces: find the listing, fix the cause, request removal, and rebuild your reputation.</description>
    <content:encoded>&lt;h1&gt;How to get your email off a blocklist and back into the inbox&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Your invoices are bouncing, or your newsletter vanished into spam, and somebody says you're "blacklisted." Recovery is a procedure anyone can follow: find the actual problem, fix it, ask for removal through the front door, and then send mail people want for long enough that the filters change their minds. Here is the sequence.&lt;/p&gt;

&lt;h2&gt;First, the scam to avoid&lt;/h2&gt;
&lt;p&gt;Spamhaus, the most consequential blocklist operator, is blunt about it: there is never any charge or fee to remove a Spamhaus listing, any offer to remove one for a fee is a scam, and no third party can influence or speed up a removal. Anyone promising to make a listing disappear for a price is selling the thing you can do yourself for free, or nothing at all.&lt;/p&gt;

&lt;h2&gt;Step 1: find out what's actually blocked&lt;/h2&gt;
&lt;p&gt;"Blacklisted" covers several different problems, and the fix depends on which one you have.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Read the bounce.&lt;/strong&gt; A rejected message comes back with an error code and usually a sentence naming the reason and often the list. Keep a copy. That text is your most specific evidence.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Check the sending IP and the domain.&lt;/strong&gt; Spamhaus's IP and Domain Reputation Checker covers its lists, and multi list lookup tools such as MXToolbox check the rest at once. Note whether the listing is on your &lt;em&gt;domain&lt;/em&gt; or on the &lt;em&gt;IP address&lt;/em&gt; that sent the mail; on a shared email service, that IP belongs to the provider, and so does the fix.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Check the mailbox providers' own views.&lt;/strong&gt; Google Postmaster Tools shows Gmail's spam rate, reputation, authentication, and delivery errors for your domain. Microsoft's Smart Network Data Services shows Outlook.com's view of IP addresses you control. These aren't public lists, and they often matter more than the lists do.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Domain blocklists are a different animal&lt;/h2&gt;
&lt;p&gt;Most talk about blocklists is about IP addresses. A second family lists &lt;em&gt;domains&lt;/em&gt; that show up inside spam messages, and SURBL is the best known. A listing there can hurt any email that merely links to you, your own signature included, wherever a spam filter consults it. SURBL's ABUSE list covers abuse and spam sites.&lt;/p&gt;
&lt;p&gt;You can check these lists yourself with a DNS lookup, with two traps to know about. First, query the domain with the list's zone added, such as &lt;code&gt;example.com.multi.surbl.org&lt;/code&gt;: no answer means not listed, and an answer in 127.0.0.x means listed, with the last number telling you which list. SURBL documents 64 as its ABUSE list. Second, the big public resolvers are often refused: SURBL answers 127.0.0.1 when your access is blocked, and Spamhaus answers 127.255.255.254, so those replies tell you nothing about your domain. Always look up the list's published test entry alongside your own, such as &lt;code&gt;test.surbl.org&lt;/code&gt;; if the test entry doesn't come back listed, your result isn't real. The TXT record for a listed domain usually states which list it's on.&lt;/p&gt;

&lt;h2&gt;Step 2: fix the cause before you ask for anything&lt;/h2&gt;
&lt;p&gt;Every operator says the same thing in different words: the problem must be identified and corrected first. Spamhaus's documentation for its fastest list notes that it will let you remove an IP, then list it again at once if the problem is still detected. The usual causes:&lt;/p&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;A compromised account or website.&lt;/strong&gt; A hacked mailbox or a form on the website relaying spam. Change passwords, turn on two factor authentication, lock down the form with a CAPTCHA, and check for anything that sends mail you didn't write.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;A bad list.&lt;/strong&gt; Purchased, scraped, or years old, full of dead addresses and spam traps. Stop sending to it. Keep only people who asked to hear from you and have engaged recently.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Missing authentication.&lt;/strong&gt; Google and Yahoo both require SPF or DKIM at a minimum for everyone, plus valid forward and reverse DNS for sending IPs, and bulk senders need SPF, DKIM, and DMARC together, aligned with the From domain. Since May 2025 Outlook.com enforces the same for domains sending more than 5,000 messages a day.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;No easy way out.&lt;/strong&gt; Google requires marketing mail from bulk senders to support one click unsubscribe and to show a visible unsubscribe link. People who can't leave press "report spam" instead, and complaints are what sink you.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Marketing on the business domain.&lt;/strong&gt; When a newsletter's complaints land on the same domain as your invoices, the invoices suffer. Move bulk mail to a separate subdomain and sending service.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;Step 3: request removal, once, with the evidence&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Spamhaus:&lt;/strong&gt; start from its IP and Domain Reputation Checker and follow the removal link for the specific list. Self-service removals are limited, so use them only after the fix.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Barracuda:&lt;/strong&gt; its removal form asks for the IP, your contact details, and a reason. It says valid requests are usually handled within 12 hours, and that repeat requests are ignored.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Microsoft:&lt;/strong&gt; the Outlook.com postmaster site is the starting point, with troubleshooting for mail that Outlook.com blocks or sends to junk.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Gmail:&lt;/strong&gt; no list to be removed from. Google's guidelines say senders who don't follow its practices may not be considered for deliverability mitigations, so meeting every requirement comes first.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;If the listing is on your provider's IP,&lt;/strong&gt; contact the provider. They control that address and its removal.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Write each request the way you'd write an appeal: what happened, what you changed, and the date you changed it. One clear request beats five vague ones.&lt;/p&gt;

&lt;h2&gt;Step 4: rebuild slowly&lt;/h2&gt;
&lt;p&gt;Removal from a list only stops the blocking. Reputation with Gmail and Microsoft recovers from behavior over time. Send only to your most engaged recipients at first, at a steady, modest volume, and raise it gradually over several weeks while you watch Postmaster Tools. Google asks senders to keep the spam rate it reports below 0.1% and never to reach 0.3%; Yahoo's limit for all senders is the same 0.3%. A spike during the rebuild is the signal to slow down.&lt;/p&gt;

&lt;h2&gt;What money really buys, and when to start fresh&lt;/h2&gt;
&lt;p&gt;Big senders do recover faster, and money is part of the reason: dedicated sending IPs they control and warm up themselves, account managers at their email platforms, paid certification programs, and staff who watch these dashboards daily. All of it shortens the feedback loop; the filters still judge the mail itself. A small sender reaches the same result on the same rules, more slowly.&lt;/p&gt;
&lt;p&gt;Sometimes the practical answer is a fresh start: business mail stays on the main domain once it's clean, and marketing moves to a new subdomain on a reputable sending service, built up from scratch with a clean list. A new address can't carry a bad list or a hacked form forward, so fix those first or the new domain will end up in the same place.&lt;/p&gt;

&lt;h2&gt;Keep it from happening again&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Read your DMARC aggregate reports, or have a tool summarize them, so spoofing and misconfigured senders show up early.&lt;/li&gt;
  &lt;li&gt;Watch Postmaster Tools monthly, and sign up for feedback loops such as Microsoft's Junk Email Reporting Program and Yahoo's Complaint Feedback Loop.&lt;/li&gt;
  &lt;li&gt;Remove bounces and long silent addresses after every campaign.&lt;/li&gt;
  &lt;li&gt;Lock down every form on the site that can send email.&lt;/li&gt;
  &lt;li&gt;Give any domain that never sends mail a reject DMARC policy and an SPF record that allows nothing, so nobody can borrow its name. My &lt;a href="https://gerriscorp.com/guides/email-authentication/"&gt;email authentication guide&lt;/a&gt; covers SPF, DKIM, and DMARC.&lt;/li&gt;
&lt;/ul&gt;

&lt;p class="related"&gt;For how to diagnose missing mail before it becomes a listing, see the &lt;a href="https://gerriscorp.com/guides/email-deliverability/"&gt;email deliverability guide&lt;/a&gt;. I set up authentication and trace delivery problems as part of &lt;a href="https://gerriscorp.com/services/email-and-dns/"&gt;email authentication and DNS&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://www.spamhaus.org/faqs/domain-blocklist/"&gt;Spamhaus: Domain Blocklist FAQ&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.spamhaus.org/faqs/combined-spam-sources-css/"&gt;Spamhaus: Combined Spam Sources FAQ&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.surbl.org/lists"&gt;SURBL: Lists&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.surbl.org/guidelines"&gt;SURBL: Implementation guidelines&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/a/answer/81126"&gt;Google Workspace Admin Help: Email sender guidelines&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/a/answer/9981691"&gt;Google Workspace Admin Help: Postmaster Tools&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://sendersupport.olc.protection.outlook.com/pm/policies.aspx"&gt;Outlook.com Postmaster&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://sendersupport.olc.protection.outlook.com/snds/"&gt;Outlook.com Smart Network Data Services&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://senders.yahooinc.com/best-practices/"&gt;Yahoo Sender Hub: Best practices&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.barracudacentral.org/rbl/removal-request"&gt;Barracuda Central: Removal request&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to vet an AI vendor's pitch before you sign anything</title>
    <link>https://gerriscorp.com/guides/ai-vendor-pitch/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/ai-vendor-pitch/</guid>
    <pubDate>Tue, 13 Oct 2026 12:00:00 +0000</pubDate>
    <description>Twelve questions that expose AI washing in a vendor pitch, the SEC and FTC cases that define it, and why a technical skeptic belongs on the sales call.</description>
    <content:encoded>&lt;h1&gt;How to vet an AI vendor's pitch before you sign anything&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-13"&gt;October 13, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Every software sales deck now has AI on the first slide. Some of those products are remarkable. Some are a thin layer over a model you could rent yourself for a few dollars a month. A few have no meaningful AI at all. Regulators have a name for the last group, AI washing, and they've started bringing cases. These are the questions I ask when a client brings me onto a vendor call, and the answers that tell me which kind of product I'm looking at.&lt;/p&gt;

&lt;h2&gt;What AI washing looks like on the record&lt;/h2&gt;
&lt;p&gt;On March 18, 2024, the SEC announced its first AI washing cases, settled charges against two investment advisers, Delphia and Global Predictions, for claiming AI capabilities they didn't have. The penalties came to $400,000 combined. The SEC said Delphia had claimed to use client data in AI-driven investing when it wasn't using that data in its process.&lt;/p&gt;
&lt;p&gt;On September 25, 2024, the FTC announced Operation AI Comply, five actions against companies using AI claims to deceive customers. One was DoNotPay, marketed as "the world's first robot lawyer"; the FTC said the company never tested whether its output matched a lawyer's, and DoNotPay agreed to pay $193,000. FTC Chair Lina Khan put the principle in one line: "there is no AI exemption from the laws on the books."&lt;/p&gt;
&lt;p&gt;Those are the extreme cases. The everyday version is a vendor whose product works fine but whose pitch credits "proprietary AI" for what is a well-written prompt and a spreadsheet.&lt;/p&gt;

&lt;h2&gt;Twelve questions for the sales call&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Which model does this run on, and who hosts it?&lt;/strong&gt; A clear answer ("a hosted model from one of the big labs, called through their API") is fine. Evasion is a warning.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;What does your product add on top of that model?&lt;/strong&gt; Good answers name real work: your integrations, a curated dataset, a review workflow, evaluation tests.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Can you run it on my data, live, right now?&lt;/strong&gt; A recorded demo shows the best case. A live run on a sample of your own material shows the typical one.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;How often is it wrong, and how do you measure that?&lt;/strong&gt; Every AI system makes mistakes. Vendors with mature products have an error rate and a test set. Vendors without one say "it's very accurate."&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;What happens when it's wrong?&lt;/strong&gt; Who notices, who fixes it, and who carries the cost: you, the vendor, or your customer.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Where does my data go?&lt;/strong&gt; Which servers, which subprocessors, how long it's kept, and whether it trains anyone's model. Get the answer in the contract.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Who owns the output?&lt;/strong&gt; Content, code, analysis, and anything the system learns from your use.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;What does it cost at ten times my volume?&lt;/strong&gt; Per seat, per task, and per token pricing behave very differently once usage grows.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;How many humans are behind it?&lt;/strong&gt; Some "automated" services quietly route work to people. That can be a fine product; it should be an honest one.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;How do I leave?&lt;/strong&gt; Export formats, notice periods, and what happens to your data afterward.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Who else uses it for exactly this?&lt;/strong&gt; Ask for a reference customer with your use case, and call them.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;What would make you tell me not to buy this?&lt;/strong&gt; Honest vendors know where their product fits badly, and they'll tell you.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;Read the answers as evidence&lt;/h2&gt;
&lt;p&gt;I score each answer three ways: specific or vague, testable or unverifiable, and consistent with what the contract says. One vague answer is normal; sales teams don't know everything and should offer to follow up. A pattern of vague answers on accuracy, data, and cost is the signal. The product might still be good, and the vendor still hasn't shown it.&lt;/p&gt;
&lt;p&gt;For a structured view of risk once you're serious, NIST's AI Risk Management Framework gives you vocabulary for governance, testing, and monitoring that your legal and security teams will recognize.&lt;/p&gt;

&lt;h2&gt;Run a paid pilot with a scorecard&lt;/h2&gt;
&lt;p&gt;Before a full contract, run a short pilot on real work with success measures you set: accuracy on a sample you've already checked by hand, time saved per task, and the number of outputs a person had to correct. Write the measures down before the pilot starts, so nobody redefines success at the end. My guide to &lt;a href="https://gerriscorp.com/guides/fact-check-ai-seo/"&gt;fact checking confident AI answers&lt;/a&gt; describes the claim by claim habit I use to grade output.&lt;/p&gt;

&lt;h2&gt;Bring a technical skeptic to the call&lt;/h2&gt;
&lt;p&gt;Sales calls are designed for the buyer who doesn't know which questions to ask. Bringing one person who does changes the whole conversation. I sit in on vendor calls for clients as exactly that person: I've built on these models myself, so I can tell a real integration from a wrapper in a few questions, and I'm comfortable asking the awkward ones politely. Afterward you get a short written recommendation: what's real, what's missing, and what to ask for in the contract. That work falls under my &lt;a href="https://gerriscorp.com/services/advisory/"&gt;advisory and second opinions&lt;/a&gt;, and if the answer is to build something yourself, see &lt;a href="https://gerriscorp.com/services/ai-automation/"&gt;AI agents and automation&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://www.sec.gov/newsroom/press-releases/2024-36"&gt;SEC: SEC charges two investment advisers with making false and misleading statements about their use of artificial intelligence (March 18, 2024)&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.ftc.gov/news-events/news/press-releases/2024/09/ftc-announces-crackdown-deceptive-ai-claims-schemes"&gt;FTC: FTC announces crackdown on deceptive AI claims and schemes (September 25, 2024)&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.nist.gov/itl/ai-risk-management-framework"&gt;NIST: AI Risk Management Framework&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to measure AI search citations, referrals, and actual leads</title>
    <link>https://gerriscorp.com/guides/measure-ai-search/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/measure-ai-search/</guid>
    <pubDate>Mon, 12 Oct 2026 12:00:00 +0000</pubDate>
    <description>Build an honest AI search report that separates citations, referral visits, and leads using Bing Webmaster Tools, GA4, repeatable checks, and clear limits.</description>
    <content:encoded>&lt;h1&gt;How to measure AI search citations, referrals, and actual leads&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-12"&gt;October 12, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;An AI answer mentions your business. Someone sends a screenshot. Another person asks whether revenue has increased. Those are three different observations, and a useful report keeps them separate. I would build a small measurement routine around citations, identifiable visits, and business outcomes, with a note explaining what each source can and cannot see.&lt;/p&gt;

&lt;h2&gt;Define the outcome before choosing a dashboard&lt;/h2&gt;
&lt;p&gt;A citation is a reference to a page in an answer. A mention names the business and might provide no link. A referral is a visit with an identifiable source. A lead is a business event, such as an inquiry or completed booking. None of those measures is interchangeable with the others.&lt;/p&gt;
&lt;p&gt;Choose the question the report needs to answer. If the goal is correcting inaccurate descriptions, an answer log may be the best instrument. If the goal is qualified inquiries, analytics and the inquiry record matter more. If the goal is finding which articles become sources, provider citation reporting deserves attention.&lt;/p&gt;
&lt;p&gt;I would avoid combining these into an invented visibility score. A score can hide a rise in mentions alongside a fall in meaningful visits. Show the underlying observations and let the reader understand the tradeoffs.&lt;/p&gt;

&lt;h2&gt;Read Bing's citation reporting within its boundaries&lt;/h2&gt;
&lt;p&gt;Bing Webmaster Tools provides an AI Performance report for Microsoft Copilot, Bing, and selected partner experiences. Microsoft's June 2026 update added preview capabilities for intent, topic, citation share, and period comparisons. These are provider observations, not a census of every AI product.&lt;/p&gt;
&lt;p&gt;Review the cited pages and the retrieval queries associated with them. A grounding query reflects the system's search for supporting material; do not present it as a verbatim customer question. Topic and intent groupings can suggest editorial questions worth investigating, but their labels are not a substitute for reading the actual pages.&lt;/p&gt;
&lt;p&gt;Microsoft defines citation share around the citations for a particular grounding query. It is not traffic share or a content quality score. When reporting a change, retain the dates and scope of the comparison. A movement may reflect demand, system changes, or other sources as well as your own revisions.&lt;/p&gt;

&lt;h2&gt;Build a separate view of identifiable visits&lt;/h2&gt;
&lt;p&gt;In GA4, examine session source and medium alongside the landing page. Start with the referring domains that actually appear in your own data rather than copying an enormous list from a marketing article. Check whether visits from an assistant are being grouped with referrals or another channel.&lt;/p&gt;
&lt;p&gt;Google Analytics lets a property use custom channel groups while leaving its default channel definitions intact. A custom group can make recognized AI referrals easier to review. Document the matching rules and check for overlap with existing categories before using the group in a recurring report.&lt;/p&gt;
&lt;p&gt;Source information is incomplete. Apps, browser privacy settings, and links that do not pass a referrer can obscure where a visit began. A visit without source information is not automatically an AI visit. Report the identifiable portion as such, and resist estimating the rest from a guess.&lt;/p&gt;
&lt;p&gt;For links you control in a campaign, consistent tagging can help attribution. You cannot assume that an assistant will preserve tags or that every visitor will follow the cited link. Keep citation totals and session totals in separate columns.&lt;/p&gt;

&lt;h2&gt;Keep Google's reporting distinct&lt;/h2&gt;
&lt;p&gt;Google documents traffic from its AI search features within the Search Console Performance report's Web search type. That aggregate is not a dedicated citation inventory. Do not label a change in all Google impressions as a measured change in AI citations.&lt;/p&gt;
&lt;p&gt;Use page and query comparisons to locate changes in search performance, then investigate them. A title revision, a changed canonical, seasonality, or an indexing issue may matter. A chart moving after a publishing change establishes timing; it does not prove the change caused the movement.&lt;/p&gt;

&lt;h2&gt;Create a repeatable answer log&lt;/h2&gt;
&lt;p&gt;A manual check is useful when its conditions are saved. Choose a short list of genuine customer questions and preserve the exact wording. Record the product, mode, date, relevant account or location context, answer summary, cited addresses, and factual inaccuracies.&lt;/p&gt;
&lt;p&gt;Repeat the same questions on a schedule, with a consistent procedure. If you change the prompt halfway through, label that as a new test. Keep unexpected or unfavorable answers in the record rather than retaining only flattering screenshots.&lt;/p&gt;
&lt;p&gt;A practical observation log might contain these fields:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Question identifier:&lt;/strong&gt; a stable label linked to the saved wording.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Experience tested:&lt;/strong&gt; the provider and the relevant search mode.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Observed answer:&lt;/strong&gt; the claim being checked and whether it is accurate.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Source addresses:&lt;/strong&gt; the pages actually linked in that response.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Follow-up:&lt;/strong&gt; the correction, page improvement, or further test required.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;This is a sample of answers under recorded conditions. It cannot establish your universal position across all users, prompts, and products. The purpose is to detect patterns and factual problems you can act on.&lt;/p&gt;

&lt;h2&gt;Connect the observations to inquiries&lt;/h2&gt;
&lt;p&gt;Define a small set of meaningful business events. A contact submission may count; a button click alone may not. Test that the event fires when the action succeeds, and check whether duplicate events inflate the count. Compare the analytics record with actual received inquiries before presenting a conversion total.&lt;/p&gt;
&lt;p&gt;An optional “How did you hear about this business?” field can add context, but self reported answers have their own limits. A customer may recall an assistant while the final visit came through a branded search. Preserve both observations without forcing them into one source label.&lt;/p&gt;
&lt;p&gt;A hypothetical monthly report could list cited pages, identifiable assistant referrals, qualified inquiries from those visits, and unresolved attribution questions. Keep small counts visible. One additional lead can produce a dramatic percentage change when the previous total was tiny.&lt;/p&gt;

&lt;h2&gt;Review the evidence before commissioning more content&lt;/h2&gt;
&lt;p&gt;Ask which pages are repeatedly useful, which questions remain unanswered, and whether the cited explanation is accurate. Improve a specific weakness and record the release date. Compare an appropriate later period, while acknowledging other changes that occurred during it.&lt;/p&gt;
&lt;p&gt;The report should end with a decision: repair a factual error, improve a relevant article, investigate tracking, or leave a useful page alone. For the publication side of that process, use &lt;a href="https://gerriscorp.com/guides/indexnow-ai-search/"&gt;the IndexNow and AI search release workflow&lt;/a&gt;. If the tracking itself needs work, see &lt;a href="https://gerriscorp.com/services/analytics/"&gt;analytics setup and repair&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://blogs.bing.com/search/2026/6/New-AI-Visibility-Insights-in-Bing-Webmaster-Tools-Intents-Topics-Citation-Share-Compare/"&gt;Microsoft Bing AI visibility reporting update&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/ai-features"&gt;Google Search Central AI features and reporting&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/analytics/answer/9756891"&gt;Google Analytics default channel groups&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>IndexNow, Bing, and a practical AI search workflow for fall 2026</title>
    <link>https://gerriscorp.com/guides/indexnow-ai-search/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/indexnow-ai-search/</guid>
    <pubDate>Mon, 12 Oct 2026 12:00:00 +0000</pubDate>
    <description>Build a reliable publishing workflow for AI search with IndexNow, Bing Webmaster Tools, tested deployments, full-text feeds, and measurable checkpoints.</description>
    <content:encoded>&lt;h1&gt;IndexNow, Bing, and a practical AI search workflow for fall 2026&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-12"&gt;October 12, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A website can publish excellent advice and still have a poor release process. The repository changes, the build finishes, a notification runs, and nobody checks which version reached the public domain. My priority for AI search this fall is making that chain observable. Give each release a clear destination, verify the content there, and keep evidence of what happened afterward.&lt;/p&gt;

&lt;h2&gt;Assign a separate job to each part of the system&lt;/h2&gt;
&lt;p&gt;A sitemap describes the site's intended public inventory. Internal links give visitors and crawlers routes through it. An RSS feed distributes articles to subscribers and other publishing systems. IndexNow sends change notifications to participating engines. None of these substitutes for a page that explains something accurately.&lt;/p&gt;
&lt;p&gt;The distinctions matter when a stakeholder asks whether a guide has been “submitted to AI.” There is no single submission step that covers every assistant. Search products can retrieve information through different providers and their own crawlers. IndexNow can support freshness in participating indexes, but it is not a pipeline into every language model.&lt;/p&gt;
&lt;p&gt;I would name the checkpoints in a release record: committed, built, deployed, publicly verified, notified, crawled, and indexed. A citation is another observation, not a synonym for any of those states. That vocabulary prevents a green build badge from becoming an unsupported visibility claim.&lt;/p&gt;

&lt;h2&gt;Prepare the page before setting up notifications&lt;/h2&gt;
&lt;p&gt;Choose a permanent address and connect it to the right guide category. Confirm the title describes the question, the introduction supplies an answer, and the author's identity is clear. Check the visible content against the structured data. If a comparison uses numbers, publish the assumptions and the date of the comparison.&lt;/p&gt;
&lt;p&gt;Then request the public address directly. It should return the expected page without requiring an editor session. Inspect the canonical and indexing controls, and test important images and downloadable resources. Compare the raw HTML with the rendered page when JavaScript is involved. My &lt;a href="https://gerriscorp.com/guides/javascript-seo/"&gt;JavaScript visibility guide&lt;/a&gt; provides a focused diagnostic sequence.&lt;/p&gt;
&lt;p&gt;Google's AI search documentation keeps the eligibility requirements within ordinary Search: the page must be indexed and allowed to appear with a snippet. Extra AI specific markup is not required. I would spend the first hour correcting access or content problems before adding another machine readable file.&lt;/p&gt;

&lt;h2&gt;Make IndexNow part of the release process&lt;/h2&gt;
&lt;p&gt;The protocol verifies control of a host through a key file. Publish that file where the submitted key location points, and confirm its contents from the public domain. Submit the final URLs on that host, rather than preview addresses or a development domain. Preserve the verification file through template changes and migrations.&lt;/p&gt;
&lt;p&gt;Send notifications for meaningful additions, revisions, or removals. The documentation permits batches of up to 10,000 URLs, but a normal editorial release rarely needs that many. A deployment that changes only a decorative detail does not justify repeatedly announcing the entire archive.&lt;/p&gt;
&lt;p&gt;For a static site, the sequence I would implement is:&lt;/p&gt;
&lt;ol&gt;
  &lt;li&gt;Build the source and validate the generated HTML, sitemap, and feeds.&lt;/li&gt;
  &lt;li&gt;Publish the approved commit to the production host.&lt;/li&gt;
  &lt;li&gt;Wait until the changed public pages match the intended release.&lt;/li&gt;
  &lt;li&gt;Submit the affected URL inventory to IndexNow.&lt;/li&gt;
  &lt;li&gt;Save the response and review engine reporting later.&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;A successful API response records acceptance, not a completed crawl. A response awaiting key verification also needs follow-up. Record the HTTP status alongside the host, time, and submitted count. If a request fails, inspect the cause before retrying; do not create an endless loop that hides a broken verification file.&lt;/p&gt;

&lt;h2&gt;Use Bing Webmaster Tools as a diagnostic instrument&lt;/h2&gt;
&lt;p&gt;Verify the business domain in Bing Webmaster Tools and review its sitemap and URL inspection information. Look at the exact production address when investigating a missing article. Keep old and new URLs together during a migration so a redirect can be evaluated against its intended destination.&lt;/p&gt;
&lt;p&gt;A submission report and an index report answer different questions. If a URL was notified but remains unavailable in search, examine the content, access rules, canonical choice, and engine feedback. Repeated submissions do not fix an incorrect canonical or a page that needs authentication.&lt;/p&gt;
&lt;p&gt;I would also retain the existing Google Search Console routine. IndexNow's participant list does not currently include Google. Use the site's Google reports to assess Google discovery and indexing rather than infer success from a Bing notification.&lt;/p&gt;

&lt;h2&gt;Give the Guides a feed that another site can actually use&lt;/h2&gt;
&lt;p&gt;A summary feed is enough for a headline widget. Republishing a Guide needs more: article content, a stable identifier, a source link, reliable dates, and absolute links inside the imported body. A relative image or navigation address that works on the originating website may point somewhere else after import.&lt;/p&gt;
&lt;p&gt;The &lt;a href="https://gerriscorp.com/rss.xml"&gt;Gerris Guides RSS feed&lt;/a&gt; carries full article content, and the &lt;a href="https://gerriscorp.com/feed.xml"&gt;Atom version&lt;/a&gt; provides an alternative. An importer should preserve the original source URL and recognize the same identifier on subsequent runs. Test one article in Plone before enabling automatic ingestion for the whole archive.&lt;/p&gt;
&lt;p&gt;Decide whether the destination displays excerpts or republishes complete articles. A full copy deserves visible attribution and a deliberate canonical policy. A feed cannot control the metadata that a separate CMS emits. Check the resulting Plone page, including its links and images, rather than assume feed parsing completed the SEO work.&lt;/p&gt;

&lt;h2&gt;Measure the business result independently&lt;/h2&gt;
&lt;p&gt;For each release, retain a short note of the question it answers, the production URL, the publication date, and the intended reader. Review search visibility and referred traffic against that record. If you test assistant answers, keep the question wording and the linked sources so another person can inspect the observation.&lt;/p&gt;
&lt;p&gt;The strongest operational outcome is a process that can distinguish publication failure from indexing delay and editorial weakness. Those require different remedies. Start with one useful Guide, establish the checkpoints, then repeat the process for the next article. For broader release planning, the &lt;a href="https://gerriscorp.com/guides/migration-checklist/"&gt;website migration checklist&lt;/a&gt; shows how inventories and verification fit together.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://www.indexnow.org/documentation"&gt;IndexNow protocol documentation&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.indexnow.org/faq"&gt;IndexNow frequently asked questions&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/ai-features"&gt;Google guidance for AI search features&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>SEO 101: a first principles primer for any website</title>
    <link>https://gerriscorp.com/guides/seo-101/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/seo-101/</guid>
    <pubDate>Fri, 09 Oct 2026 12:00:00 +0000</pubDate>
    <description>SEO from first principles for absolute beginners: how search works, people-first content, E-E-A-T, titles, links, alt text, speed, and accessibility.</description>
    <content:encoded>&lt;h1&gt;SEO 101: a first principles primer for any website&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-09"&gt;October 9, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Search engine optimization sounds like a dark art. It's mostly common sense, applied carefully. A search engine has one job: send each searcher to the page that best answers the question. Everything in SEO follows from helping it do that job, and from never trying to fool it. This primer works for any website on any platform, from a hand-coded page to a store with ten thousand products.&lt;/p&gt;

&lt;h2&gt;Principle 1: understand the machine&lt;/h2&gt;
&lt;p&gt;Search works in three stages.&lt;/p&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Crawling.&lt;/strong&gt; Programs called crawlers follow links from page to page and download what they find.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Indexing.&lt;/strong&gt; The engine reads each page, figures out what it's about, and files it in an enormous catalog.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Serving.&lt;/strong&gt; When someone searches, the engine pulls the most useful pages from that catalog and ranks them.&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;AI answers in Google, Bing, and ChatGPT search draw on the same kind of catalog. A page that can't be crawled can't be indexed, and a page that isn't indexed can't be ranked or quoted by anyone. That is why the technical basics come first.&lt;/p&gt;

&lt;h2&gt;Principle 2: write for people first&lt;/h2&gt;
&lt;p&gt;Google's own guidance asks one question of every page: was it made to help a person, or to attract search traffic? Write the page you'd want to find. Answer the question fully, in plain language, early on the page. Include what only you know: your prices, your process, your photos, your results. A page that would read the same with any company's name swapped in gives a searcher nothing new.&lt;/p&gt;

&lt;h2&gt;Principle 3: earn trust with E-E-A-T&lt;/h2&gt;
&lt;p&gt;Google's search quality raters judge pages for experience, expertise, authoritativeness, and trustworthiness, and Google describes trust as the most important of the four. You show them on the page itself:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Experience:&lt;/strong&gt; evidence you've actually done the thing, such as real photos, real numbers, and lessons learned the hard way.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Expertise:&lt;/strong&gt; a named author with a short bio and the credentials that matter for the topic.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Authoritativeness:&lt;/strong&gt; other respected people and sites mentioning and linking to you.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Trust:&lt;/strong&gt; accurate content, a real address and contact details, clear policies, HTTPS, and sources for your claims.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Principle 4: label every page clearly&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Title tag:&lt;/strong&gt; the clickable headline in search results. Make it unique, specific, and about 50 to 60 characters, with the page's main subject near the front.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Meta description:&lt;/strong&gt; the summary under the headline. A clear 140 to 160 character description earns the click, even though it plays no direct part in ranking.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;One H1:&lt;/strong&gt; a single main heading that says what the page is about, followed by H2 and H3 subheadings in order, like an outline.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Readable URLs:&lt;/strong&gt; &lt;code&gt;/services/roof-repair/&lt;/code&gt; tells people and crawlers more than &lt;code&gt;/page?id=4817&lt;/code&gt;.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Principle 5: link with words that mean something&lt;/h2&gt;
&lt;p&gt;Internal links, from one page of your site to another, are how crawlers find your pages and how they learn what each page covers. The clickable words matter. "Read more," "click here," and "learn more" tell nobody anything, and a screen reader user who pulls up a list of a page's links hears the same useless phrase ten times. Make the link text describe the destination, the way "see my &lt;a href="https://gerriscorp.com/guides/core-web-vitals/"&gt;guide to Core Web Vitals&lt;/a&gt;" names exactly where the click leads. Link from strong pages to the pages you most want found, and make sure every important page is reachable through ordinary links, without relying on search boxes or scripts.&lt;/p&gt;

&lt;h2&gt;Principle 6: describe every meaningful image&lt;/h2&gt;
&lt;p&gt;Alt text is the written description of an image, read aloud by screen readers and read by search engines. Describe what the picture shows in a short sentence, and include any words that appear in it. Skip "image of" and "photo of"; screen readers already announce that it's an image. Decorative images, such as background flourishes, get an empty alt attribute so screen readers skip them. Keyword lists in alt text help nobody and look like spam.&lt;/p&gt;

&lt;h2&gt;Principle 7: make it fast and stable&lt;/h2&gt;
&lt;p&gt;Google's Core Web Vitals measure three things: how fast the main content appears (Largest Contentful Paint, good under 2.5 seconds), how quickly the page responds to a tap or click (Interaction to Next Paint, good under 200 milliseconds), and whether the layout jumps around while loading (Cumulative Layout Shift, good under 0.1). Compress and correctly size images, load fewer scripts, and give images width and height so the page doesn't shift. Test any page free in PageSpeed Insights. For the details, see my &lt;a href="https://gerriscorp.com/guides/core-web-vitals/"&gt;Core Web Vitals guide&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;Principle 8: build it so everyone can use it&lt;/h2&gt;
&lt;p&gt;Accessibility and SEO want the same things: real text, logical headings, descriptive links, alt text, captions, enough color contrast, and a site that works with a keyboard alone. The Web Content Accessibility Guidelines (WCAG) are the standard, and in the United States the ADA has been applied to websites in court. Every accessibility fix makes the site easier to crawl too. See &lt;a href="https://gerriscorp.com/guides/ada-website-compliance/"&gt;ADA website compliance&lt;/a&gt; and &lt;a href="https://gerriscorp.com/guides/lynx-qa/"&gt;testing a site in a text browser&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;My standing recommendation: put the site behind Cloudflare&lt;/h2&gt;
&lt;p&gt;For almost every site, I recommend running DNS and caching through &lt;a href="https://www.cloudflare.com/"&gt;Cloudflare&lt;/a&gt;. Even its free plan speeds up pages, handles HTTPS certificates, and absorbs attacks, and it sells domains at cost. Hosted builders such as Shopify, Squarespace, and Wix are the exception; with those, keep Cloudflare to DNS only. My &lt;a href="https://gerriscorp.com/guides/cloudflare-seo/"&gt;Cloudflare guide&lt;/a&gt; covers which plan to choose, what to budget, and the settings that help or hurt search.&lt;/p&gt;

&lt;h2&gt;Principle 9: open the door for crawlers&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Keep &lt;a href="https://gerriscorp.com/guides/robots-txt/"&gt;robots.txt&lt;/a&gt; from blocking pages you want found.&lt;/li&gt;
  &lt;li&gt;Publish an XML sitemap listing your important pages, and submit it in Google Search Console. Then open Bing Webmaster Tools and use its import option to copy every verified site and sitemap over from Search Console in a few clicks. Both tools are free, and I check both regularly to make sure the sitemaps are read without errors.&lt;/li&gt;
  &lt;li&gt;Turn on IndexNow, through Cloudflare, your SEO plugin, or your platform, so Bing and other participating engines hear about new and changed pages right away.&lt;/li&gt;
  &lt;li&gt;Use HTTPS everywhere, and pick one version of the domain, with or without www, and redirect the other to it.&lt;/li&gt;
  &lt;li&gt;Add &lt;a href="https://gerriscorp.com/guides/structured-data/"&gt;structured data&lt;/a&gt; that matches what the page visibly says.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Principle 10: stay white hat&lt;/h2&gt;
&lt;p&gt;Every shortcut on this list has hurt real businesses, some permanently: buying links, keyword stuffing, hidden text, pages written only to rank, mass AI content nobody reviewed, fake reviews, and fake authors. The search engines publish their spam policies openly. Read them once, and you'll recognize the pitches in your inbox for what they are.&lt;/p&gt;

&lt;h2&gt;Your first week&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;Verify the site in Google Search Console and Bing Webmaster Tools, and submit the sitemap.&lt;/li&gt;
  &lt;li&gt;List your ten most important pages, and give each a unique title, description, and H1.&lt;/li&gt;
  &lt;li&gt;Replace every "read more" and "click here" with descriptive link text.&lt;/li&gt;
  &lt;li&gt;Add alt text to every meaningful image.&lt;/li&gt;
  &lt;li&gt;Run your home page and one inner page through PageSpeed Insights, and fix the largest problem it names.&lt;/li&gt;
  &lt;li&gt;Ask three happy customers for reviews on your Google Business Profile.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;Further reading&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/fundamentals/seo-starter-guide"&gt;Google's SEO Starter Guide&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/essentials"&gt;Google Search Essentials&lt;/a&gt;, including the spam policies&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/fundamentals/creating-helpful-content"&gt;Creating helpful, reliable, people-first content&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://static.googleusercontent.com/media/guidelines.raterhub.com/en//searchqualityevaluatorguidelines.pdf"&gt;Google's Search Quality Rater Guidelines&lt;/a&gt; (PDF), where E-E-A-T is defined&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/title-link"&gt;How Google writes title links&lt;/a&gt; and &lt;a href="https://developers.google.com/search/docs/appearance/snippet"&gt;how it writes snippets&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.bing.com/webmasters/help/webmaster-guidelines-30fba23a"&gt;Bing Webmaster Guidelines&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://web.dev/articles/vitals"&gt;Web Vitals on web.dev&lt;/a&gt; and &lt;a href="https://pagespeed.web.dev/"&gt;PageSpeed Insights&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.w3.org/WAI/WCAG22/quickref/"&gt;WCAG 2.2 quick reference&lt;/a&gt; and &lt;a href="https://www.w3.org/WAI/WCAG22/Understanding/link-purpose-in-context.html"&gt;the W3C's explanation of link purpose&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://webaim.org/techniques/alttext/"&gt;WebAIM on alternative text&lt;/a&gt;, the &lt;a href="https://www.w3.org/WAI/tutorials/images/decision-tree/"&gt;W3C alt text decision tree&lt;/a&gt;, and the &lt;a href="https://wave.webaim.org/"&gt;WAVE accessibility checker&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://schema.org/docs/gs.html"&gt;Getting started with schema.org&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://moz.com/beginners-guide-to-seo"&gt;Moz's Beginner's Guide to SEO&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Want a second pair of eyes on your first week's list? &lt;a href="https://gerriscorp.com/contact/"&gt;Send me the site&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/fundamentals/seo-starter-guide"&gt;Google Search Central: SEO Starter Guide&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/fundamentals/how-search-works"&gt;Google Search Central: In-depth guide to how Google Search works&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/fundamentals/creating-helpful-content"&gt;Google Search Central: Creating helpful, reliable, people-first content&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/links-crawlable"&gt;Google Search Central: Link best practices&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://web.dev/articles/vitals"&gt;web.dev: Web Vitals&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.w3.org/WAI/WCAG22/quickref/"&gt;W3C: WCAG 2.2 quick reference&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Ranking yourself first in a "best of" list now backfires in AI answers</title>
    <link>https://gerriscorp.com/guides/self-ranking-lists/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/self-ranking-lists/</guid>
    <pubDate>Fri, 09 Oct 2026 12:00:00 +0000</pubDate>
    <description>How self-ranked best agency lists came to steer AI answers, why Google, Anthropic, and the data now push back, and what earns a real AI recommendation instead.</description>
    <content:encoded>&lt;h1&gt;Ranking yourself first in a "best of" list now backfires in AI answers&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-09"&gt;October 9, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Ask ChatGPT, Gemini, or Google's AI Mode for the best AI search agency and you get a confident shortlist. Look at where that shortlist comes from and you find something funny: most of the sources are "best agency" lists written by agencies that put themselves first. For about two years that trick worked. In 2026 the search engines, the AI companies, and the data all turned against it at once.&lt;/p&gt;

&lt;h2&gt;How the trick works&lt;/h2&gt;
&lt;p&gt;AI assistants assemble recommendations from pages that already compare options, and a list titled "best GEO agencies in 2026" is exactly that kind of page. So agencies publish one, rank themselves first, and fill the rest with respectable competitors so the list looks fair. A few are candid about it. One list opens with its author admitting the self-ranking in a single word: "Guilty." Others score themselves highest on a scoring system they designed. The pattern is easy to spot in the search results for any "best agency" query this month.&lt;/p&gt;
&lt;p&gt;One agency studied the effect from the inside. Tracking 48 AI answers to four "which agency" prompts in late summer 2026, it found that almost every frequently cited source was an agency's own list ranking itself first, and that a single agency's site supplied the source for 14 of the 48 answers. Its conclusion, in its own words: the agencies AI recommends are, to a large degree, the agencies that published the lists AI reads. That agency then published its own scored list, and kept itself out of the ranking.&lt;/p&gt;

&lt;h2&gt;Why it stopped working&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Google changed the rules in May.&lt;/strong&gt; Its spam policies now open by including attempts to manipulate generative AI responses in Search. Lists built mainly to plant a recommendation sit squarely in that territory, alongside hidden instructions and fake sources.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The citations are already falling.&lt;/strong&gt; Lily Ray tracked 100 "best B2B software" queries in Google's AI Overviews. Compared with June, September's answers cited 38 percent fewer self-promotional "best of" lists, and the brand that published the list was left out of the recommendation 83 percent of the time, up from 69 percent. One researcher and one sample, but the direction is clear.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The AI companies are closing the door too.&lt;/strong&gt; Anthropic's updated usage policy, released October 8 and effective November 12, forbids using Claude to manipulate the sources search engines and AI systems draw answers from by seeding content that misrepresents its origin, authorship, or independence. Its example: networks of sites posing as unaffiliated sources that corroborate the same claims.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;When the engine, the referee, and the tool maker all move in the same direction, the tactic has a short future.&lt;/p&gt;

&lt;h2&gt;What earns a real recommendation&lt;/h2&gt;
&lt;p&gt;Assistants are trying to do what a careful buyer does: find independent evidence that several unrelated people reached the same conclusion. Give them that evidence.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Reviews on platforms you don't control:&lt;/strong&gt; Google Business Profile, Clutch, G2, and the directories of your industry.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Lists written by someone else:&lt;/strong&gt; trade press, analysts, associations, and customers' own write-ups.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Case studies with checkable detail:&lt;/strong&gt; a named client where permitted, dates, before-and-after figures, and the method.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Honest comparison pages:&lt;/strong&gt; if you compare yourself with competitors, say so plainly, state who each option suits, and include the ones that beat you on something.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;A consistent entity:&lt;/strong&gt; the same name, description, people, and profiles everywhere, so every system recognizes one business.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;If you already published a self-ranked list&lt;/h2&gt;
&lt;p&gt;Disclose the authorship at the top in a sentence a reader can't miss, publish the scoring method, and score yourself by the same rules as everyone else, or step out of the ranking entirely. Better still, turn it into a buyer's checklist that helps people choose well, with your own offer described once, at the end, as one option.&lt;/p&gt;

&lt;h2&gt;How to vet anyone selling AI visibility&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;Show me an AI answer that names one of your clients, with the date.&lt;/li&gt;
  &lt;li&gt;Which tool measures it, and can I see the raw prompts?&lt;/li&gt;
  &lt;li&gt;What do you do off my site, and do any of those placements hide who paid for them?&lt;/li&gt;
  &lt;li&gt;Do you publish "best of" lists that rank your own firm, or place clients in them?&lt;/li&gt;
  &lt;li&gt;Which of your tactics would you stop doing if Google or Anthropic wrote about them tomorrow?&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;Every answer to the last question tells you how long the results will last. For the wider picture, see &lt;a href="https://gerriscorp.com/guides/search-in-2026/"&gt;the state of search in 2026&lt;/a&gt; and &lt;a href="https://gerriscorp.com/guides/ai-assistants/"&gt;how AI assistants decide what to say about your business&lt;/a&gt;, or &lt;a href="https://gerriscorp.com/contact/"&gt;ask me what the assistants say about yours&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://lilyraynyc.substack.com/p/prediction-the-next-massive-google"&gt;Lily Ray: Prediction, the next massive Google update is around the corner&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/essentials/spam-policies"&gt;Google Search Central: Spam policies for Google web search&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.searchenginejournal.com/claude-rules-fake-sources-ai-answers/592456/"&gt;Search Engine Journal: Claude&amp;#x27;s new rules target fake sources built to sway AI answers&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.tobeagency.co/learn/12-best-ai-seo-agencies-in-2026-scored-on-what-they-can-prove"&gt;Tobe Agency: 12 best AI SEO, AEO and GEO agencies in 2026, scored&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.yesoptimist.com/best-geo-agencies"&gt;Optimist: The 7 best GEO agencies driving real revenue from AI in 2026&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>The state of search in 2026: every Google update so far, and the AI shift behind them</title>
    <link>https://gerriscorp.com/guides/search-in-2026/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/search-in-2026/</guid>
    <pubDate>Fri, 09 Oct 2026 12:00:00 +0000</pubDate>
    <description>Every 2026 Google core and spam update with dates, the new spam rule against manipulating AI answers, what Bing and ChatGPT now report, and what to do now.</description>
    <content:encoded>&lt;h1&gt;The state of search in 2026: every Google update so far, and the AI shift behind them&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-09"&gt;October 9, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;2026 has been the busiest year for Google's ranking systems in a long time: two core updates, four spam updates, and a rule change that brings AI Overviews and AI Mode under the spam policies. Here is what happened, in order, from Google's own records, along with what Bing and OpenAI now let site owners see, and what all of it means for anyone who publishes on the web.&lt;/p&gt;

&lt;h2&gt;The 2026 updates, from Google's dashboard&lt;/h2&gt;
&lt;p&gt;Google logs every confirmed ranking update on its Search Status Dashboard. Through early October, 2026 has seven:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;February Discover update:&lt;/strong&gt; began February 5, ran about 22 days. It changed what appears in the Discover feed, separately from Search.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;March spam update:&lt;/strong&gt; began March 24, finished in under a day.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;March core update:&lt;/strong&gt; began March 27, ran about 12 days.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;May core update:&lt;/strong&gt; began May 21, ran about 12 days, finishing June 2.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;June spam update:&lt;/strong&gt; began June 24, ran about two days.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;August spam update:&lt;/strong&gt; began August 18, ran under three days.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;September spam update:&lt;/strong&gt; began September 24 and ran nearly two weeks, far longer than the year's earlier spam updates.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Core updates reassess how every page ranks against everything else, and pages can rise or fall. Spam updates improve the automated systems that detect policy violations, and the sites they hit tend to drop hard and fast.&lt;/p&gt;

&lt;h2&gt;The big rule change: AI answers are now inside the spam policies&lt;/h2&gt;
&lt;p&gt;In May, Google revised the opening of its spam policies. Spam now explicitly includes "attempting to manipulate generative AI responses in Google Search," which covers AI Overviews and AI Mode. The June spam update was widely reported as the first to enforce it. In practice, the tactics now in scope include:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;Hidden text or instructions aimed at AI systems: white-on-white copy, zero-size fonts, or prompts tucked into markup that try to make an assistant favor a brand.&lt;/li&gt;
  &lt;li&gt;Structured data that claims things the visible page doesn't say.&lt;/li&gt;
  &lt;li&gt;Networks of near-identical "best of" pages built to seed the answers an AI assembles.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Security researchers have shown for two years that hidden page text can steer AI search tools. Search engines are now treating those tricks the way they treated keyword stuffing in 2005.&lt;/p&gt;

&lt;h2&gt;Scaled content abuse is the year's main target&lt;/h2&gt;
&lt;p&gt;Google's policy defines scaled content abuse by intent and outcome: many pages made primarily to manipulate rankings, written by automation, by people, or by both. Automation alone is never the violation; the purpose is. Reporting on the August and September updates centered on sites that mass-published AI-written pages with no review and nothing original in them. Two older policies stayed in the news alongside it:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Site reputation abuse:&lt;/strong&gt; renting a trusted site's subfolder or subdomain to third-party content, coupons, gambling, or affiliate pages, to borrow its rankings.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Expired domain abuse:&lt;/strong&gt; buying a dead domain with history and filling it with low-value pages to exploit its old reputation.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;What Google says about optimizing for AI Overviews&lt;/h2&gt;
&lt;p&gt;Google's own documentation is blunt: there are no additional requirements to appear in AI Overviews or AI Mode, and no special optimizations, files, or schema are needed. A page has to be indexed and eligible to show a snippet. The recommended practices are the ordinary ones: allow crawling, link pages internally, keep important content in text, and make structured data match the visible page. Traffic from both features is counted in Search Console's Performance report under the Web search type. Every vendor selling a secret AI Overview formula is selling something Google says doesn't exist.&lt;/p&gt;

&lt;h2&gt;Bing and ChatGPT now show their work&lt;/h2&gt;
&lt;p&gt;Bing Webmaster Tools added an AI Performance report, in public preview since February. It counts how often Copilot, Bing's AI summaries, and some partner integrations cite your pages, and it lists the grounding queries that led to those citations. It counts citations only, no clicks, and it is a rare first-party view of how an AI search product uses a site.&lt;/p&gt;
&lt;p&gt;OpenAI documents its crawlers separately. &lt;code&gt;OAI-SearchBot&lt;/code&gt; decides whether a site can appear in ChatGPT search; &lt;code&gt;GPTBot&lt;/code&gt; collects training data. Each is controlled independently in robots.txt, so a site can welcome search and decline training. A site that blocks OAI-SearchBot, often by accident through a firewall or CDN bot setting, simply drops out of ChatGPT's answers.&lt;/p&gt;

&lt;h2&gt;A measurement change that still confuses reports&lt;/h2&gt;
&lt;p&gt;In September 2025, Google stopped honoring the &lt;code&gt;num=100&lt;/code&gt; parameter that let rank trackers load a hundred results at once. Desktop impressions in Search Console fell for most sites and average position jumped, because bots had been inflating the counts. Any year-over-year report that spans September 2025 needs that footnote, or it will show a decline that never happened to real visitors.&lt;/p&gt;

&lt;h2&gt;What wins in this environment&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Evidence of real work.&lt;/strong&gt; Original data, photos, tests, case details, and first-hand experience: material an assistant can't produce by summarizing other pages.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Fewer, better pages.&lt;/strong&gt; Consolidate thin and overlapping pages into ones that fully answer a question.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Clean access.&lt;/strong&gt; Check robots.txt, CDN, and firewall settings for every crawler you want: Googlebot, Bingbot, OAI-SearchBot, and the rest.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Honest markup.&lt;/strong&gt; Structured data that matches the page, and nothing hidden.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;A clear entity.&lt;/strong&gt; Consistent names, addresses, and profiles across the web, so every engine and assistant knows exactly who you are.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;For the crawler side, see &lt;a href="https://gerriscorp.com/guides/ai-assistants/"&gt;how AI assistants decide what to say about your business&lt;/a&gt; and &lt;a href="https://gerriscorp.com/guides/robots-txt/"&gt;my robots.txt guide&lt;/a&gt;. If a 2026 update has already hit you, start with &lt;a href="https://gerriscorp.com/guides/traffic-drop/"&gt;how to diagnose a traffic drop&lt;/a&gt;, or &lt;a href="https://gerriscorp.com/contact/"&gt;send me the dates and the domain&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://status.search.google.com/products/rGHU1u87FJnkP6W2GwMi/history"&gt;Google Search Status Dashboard: Ranking incident history&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/essentials/spam-policies"&gt;Google Search Central: Spam policies for Google web search&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/ai-features"&gt;Google Search Central: AI features and your website&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.searchenginejournal.com/seo-pulse-google-spam-update-rolls-out-ai-manipulation-in-scope/580565/"&gt;Search Engine Journal: Google spam update rolls out, AI manipulation in scope&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.bing.com/webmasters/"&gt;Bing Webmaster Tools&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.openai.com/docs/bots"&gt;OpenAI: Overview of OpenAI crawlers&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Cloudflare settings that help or hurt SEO and AI crawlers</title>
    <link>https://gerriscorp.com/guides/cloudflare-seo/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/cloudflare-seo/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>The Cloudflare settings that decide whether Googlebot, Bingbot, and AI crawlers can reach your site, plus the caching, HTTPS, and redirect choices that help.</description>
    <content:encoded>&lt;h1&gt;Cloudflare settings that help or hurt SEO and AI crawlers&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt; · Updated &lt;time datetime="2026-10-09"&gt;October 9, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Cloudflare sits in front of millions of websites, and a handful of its switches decide whether search engines and AI assistants can read yours. I set it up for clients often, and I audit it every time a site's crawling looks strange. These are the settings I check first.&lt;/p&gt;

&lt;h2&gt;Why I recommend it, and what to budget&lt;/h2&gt;
&lt;p&gt;For nearly every client site, I recommend moving DNS to Cloudflare and caching through it. My own experience and every AI assistant I've asked agree on that, and I pay for Cloudflare Pro on &lt;a href="https://chrisabraham.com/"&gt;chrisabraham.com&lt;/a&gt;. Cloudflare pays me nothing for the recommendation, though given how often I make it, I've joked that it should.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Registration and DNS only:&lt;/strong&gt; if Cloudflare just holds the domain and answers DNS while traffic goes straight to your host, the free plan is all you need. Cloudflare Registrar charges the registry's wholesale price, with no markup on renewals.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Proxied on the free plan:&lt;/strong&gt; route the site through Cloudflare and even the free plan caches images, scripts, and stylesheets close to visitors, compresses them, provides certificates, and soaks up attacks.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Paid plans:&lt;/strong&gt; Pro and above add features such as Polish, which compresses and converts images on the fly. When a proxied site benefits from those, I tell clients to budget about $50 a month for Cloudflare's paid services, starting with Pro.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Hosted builders:&lt;/strong&gt; Shopify, Squarespace, and Wix run their own CDN and certificates. Keep their records on DNS only and follow each platform's connection instructions.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Proxied or DNS only&lt;/h2&gt;
&lt;p&gt;Every DNS record in Cloudflare has a cloud icon. Orange means proxied: visitors and crawlers reach Cloudflare's network, which applies caching, security, and rules before passing the request to your server. Grey means DNS only: Cloudflare answers the DNS lookup and then steps aside. Only proxied records get Cloudflare's speed and protection, and only proxied records can be blocked by its security features. When a crawling problem appears, the cloud color is the first thing to confirm.&lt;/p&gt;
&lt;p&gt;Mail records always stay grey. MX targets and the hostnames used by mail services must resolve to the real servers.&lt;/p&gt;

&lt;h2&gt;AI crawler controls&lt;/h2&gt;
&lt;p&gt;In July 2025 Cloudflare began blocking known AI crawlers by default on newly added domains, and it offers a one-click switch to block them on existing ones. The current controls live under AI Crawl Control, where each crawler can be allowed or blocked by name, and Cloudflare can also manage robots.txt for you with its own content signals added.&lt;/p&gt;
&lt;p&gt;This matters because the crawlers are different jobs. GPTBot and ClaudeBot gather training data. OAI-SearchBot, Claude-SearchBot, and PerplexityBot build the indexes that assistants search when they answer questions live. Blocking the search crawlers keeps you out of those answers entirely. Decide crawler by crawler, then confirm the setting matches the decision. The &lt;a href="https://gerriscorp.com/guides/ai-assistants/"&gt;AI assistants guide&lt;/a&gt; explains which crawler feeds which product.&lt;/p&gt;

&lt;h2&gt;Bot protection and challenges&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Bot Fight Mode&lt;/strong&gt; on the free plan challenges traffic that looks automated. Cloudflare exempts its list of verified bots, which includes Googlebot and Bingbot, and the WAF can't override this mode, so a misfire is hard to work around.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Super Bot Fight Mode&lt;/strong&gt; on paid plans has a separate setting for verified bots. Set it to allow.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Under Attack Mode&lt;/strong&gt; puts a challenge page in front of every visitor, and crawlers can't pass it. Use it during an actual attack and turn it off when the attack ends.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Custom WAF rules&lt;/strong&gt; that block by country, by user agent, or by request rate catch crawlers too. Googlebot crawls mostly from the United States, so a rule that challenges American traffic hits it directly.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;The proof is in Security, then Events. Filter by user agent for Googlebot or Bingbot and look for blocks or managed challenges. In Search Console, the Crawl stats report under Settings shows host status and response codes from Google's side, and a URL Inspection live test shows exactly what Googlebot received.&lt;/p&gt;

&lt;h2&gt;Caching HTML&lt;/h2&gt;
&lt;p&gt;By default Cloudflare caches static files such as images, CSS, and JavaScript, and passes HTML requests through to your server. For a site whose pages are the same for every visitor, a Cache Rule that marks HTML as eligible for cache can cut time to first byte to a few dozen milliseconds worldwide. Exclude anything personal: carts, checkouts, account pages, admin paths, and any request carrying a login cookie. Set an edge cache time you can live with, and purge the cache on deploy.&lt;/p&gt;

&lt;h2&gt;HTTPS settings&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;SSL mode Full (strict).&lt;/strong&gt; Cloudflare connects to your server over HTTPS and checks its certificate. Flexible mode connects over plain HTTP, and when the server also redirects to HTTPS, the result is an endless redirect loop.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Always Use HTTPS.&lt;/strong&gt; Sends every HTTP request to HTTPS with a 301 at the edge, so the server needs no rule of its own.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;HSTS.&lt;/strong&gt; Tells browsers to refuse plain HTTP for a period you choose. Start with a short max age, and add preload only when every subdomain serves HTTPS, because preload is slow to undo.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Features that change your HTML&lt;/h2&gt;
&lt;p&gt;Several Cloudflare features rewrite pages on the way out, and each deserves a deliberate decision.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Rocket Loader&lt;/strong&gt; defers JavaScript by rewriting script tags. It can speed up simple sites and break complex ones, including analytics and consent banners. Test with it off before blaming anything else.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Email Address Obfuscation&lt;/strong&gt; replaces email addresses in the HTML with a script. Visitors see the address, and crawlers that read raw HTML see a placeholder.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Auto Minify&lt;/strong&gt; was retired in 2024. If a build still depends on it, move minification into the build.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Redirects at the edge&lt;/h2&gt;
&lt;p&gt;Redirect Rules handle patterns, such as moving a whole folder or forcing the www or bare domain. Bulk Redirects handle long lists of one-to-one moves from a file. Both answer at the edge with a true 301, which makes Cloudflare a good home for a migration's redirect map. Keep all the redirects in one layer; rules split between Cloudflare and the origin server are where chains and loops come from. The &lt;a href="https://gerriscorp.com/guides/301-redirects/"&gt;redirects guide&lt;/a&gt; covers mapping and testing.&lt;/p&gt;

&lt;h2&gt;Crawler Hints&lt;/h2&gt;
&lt;p&gt;Crawler Hints notifies search engines that support IndexNow, including Bing and Yandex, when your content changes, based on what Cloudflare sees passing through its cache. It's a free switch under Caching and worth turning on for any site that changes regularly. Google doesn't use IndexNow, so it does nothing for Google.&lt;/p&gt;

&lt;section class="faq"&gt;
&lt;h2&gt;Questions about Cloudflare and search&lt;/h2&gt;
&lt;h3&gt;Does Cloudflare's shared IP address hurt rankings?&lt;/h3&gt;
&lt;p&gt;No. Millions of sites share Cloudflare's addresses, and Google judges sites by their content and links, not by who else is on the IP.&lt;/p&gt;
&lt;h3&gt;Should a small site use Cloudflare at all?&lt;/h3&gt;
&lt;p&gt;Yes, at least for DNS. Cloudflare's free DNS is fast and well documented, and proxying can be added record by record later. This site uses Cloudflare for DNS only, because GitHub Pages already provides HTTPS and a CDN.&lt;/p&gt;
&lt;/section&gt;

&lt;p class="related"&gt;I configure and audit Cloudflare as part of &lt;a href="https://gerriscorp.com/services/site-speed/"&gt;site speed and Cloudflare work&lt;/a&gt; and &lt;a href="https://gerriscorp.com/services/crawler-visibility/"&gt;crawler visibility audits&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.cloudflare.com/cache/concepts/default-cache-behavior/"&gt;Cloudflare Docs: Default cache behavior&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.cloudflare.com/images/polish/"&gt;Cloudflare Docs: Polish&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.cloudflare.com/cache/advanced-configuration/crawler-hints/"&gt;Cloudflare Docs: Crawler Hints&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.cloudflare.com/bots/concepts/bot/verified-bots/"&gt;Cloudflare Docs: Verified bots&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.cloudflare.com/cache/"&gt;Cloudflare Docs: Cache&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/robots/intro"&gt;Google Search Central: Introduction to robots.txt&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>The year-end SEO check: one afternoon that protects next year</title>
    <link>https://gerriscorp.com/guides/year-end-seo-check/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/year-end-seo-check/</guid>
    <pubDate>Wed, 07 Oct 2026 12:00:00 +0000</pubDate>
    <description>A one-afternoon year-end SEO check for small and midsize sites: indexing, sitemaps, analytics retention, domains and certificates, listings, and AI answers.</description>
    <content:encoded>&lt;h1&gt;The year-end SEO check: one afternoon that protects next year&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-07"&gt;October 7, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;December is quiet for a lot of websites, which makes it the right time to find the problems that pile up unnoticed. None of these checks takes long, and together they catch the small failures that otherwise surface in March as a mystery traffic drop. Block an afternoon, work down the list, and write down what you find.&lt;/p&gt;

&lt;h2&gt;1. Indexing: is Google seeing what you think it is?&lt;/h2&gt;
&lt;p&gt;Open Search Console's Page indexing report and compare the number of indexed pages with the number of pages you actually want in search. A sudden rise in excluded pages, or important pages listed as crawled but not indexed, deserves a closer look. Spot check five of your most valuable pages with URL Inspection to confirm they're indexed and that Google chose the canonical you intended.&lt;/p&gt;

&lt;h2&gt;2. Sitemaps: current and clean&lt;/h2&gt;
&lt;p&gt;In the Sitemaps report, confirm each sitemap was read recently and without errors. A sitemap should list only live pages you want indexed. If it still includes redirected, deleted, or noindexed URLs from this year's changes, clean it up.&lt;/p&gt;

&lt;h2&gt;3. Analytics: keep the history you'll want&lt;/h2&gt;
&lt;p&gt;Google Analytics 4 keeps detailed event data for only two months by default, and you can extend it to fourteen months in the property's data retention settings. If yours still says two, change it now, before you need a year-over-year comparison you can't make. While you're there, confirm that key events like form submissions and calls still fire after this year's site changes.&lt;/p&gt;

&lt;h2&gt;4. Domains and certificates&lt;/h2&gt;
&lt;p&gt;List every domain the business owns, including old ones that only redirect. Check renewal dates and turn on auto renew for the ones you'll keep. A lapsed domain can be bought by someone else and pointed anywhere. Confirm HTTPS certificates renew automatically, and that redirects from old domains still land on the right pages.&lt;/p&gt;

&lt;h2&gt;5. Redirects from this year's changes&lt;/h2&gt;
&lt;p&gt;If you redesigned, renamed, or retired pages this year, crawl your old URLs and confirm each one returns a single redirect to a relevant page, with no chains and no redirects to the home page by default.&lt;/p&gt;

&lt;h2&gt;6. Speed and stability&lt;/h2&gt;
&lt;p&gt;Look at the Core Web Vitals report for pages that slipped from good to needs improvement. The cause is often something added during the year: a chat widget, a new tracking script, or oversized images on a new template.&lt;/p&gt;

&lt;h2&gt;7. Listings and holiday hours&lt;/h2&gt;
&lt;p&gt;Check your Business Profile for accurate hours, categories, photos, and contact details. Set special hours for the holidays still ahead, and confirm Bing Places and Apple Maps match.&lt;/p&gt;

&lt;h2&gt;8. What AI assistants say about you&lt;/h2&gt;
&lt;p&gt;Ask ChatGPT, Gemini, Claude, Perplexity, and Google's AI Overviews the questions a customer would ask about your business. Note what's wrong and where each answer seems to come from. Year end is a good moment to set this baseline and correct the sources, so you can compare next December.&lt;/p&gt;

&lt;h2&gt;9. Structured data&lt;/h2&gt;
&lt;p&gt;Run your key page types through a schema validator. Prices, hours, addresses, and people who left the company are the facts that go stale in structured data first.&lt;/p&gt;

&lt;h2&gt;10. Access and ownership&lt;/h2&gt;
&lt;p&gt;List who has access to Search Console, analytics, the Business Profile, the domain registrar, DNS, and hosting. Remove former staff and agencies that no longer work with you, and make sure at least two people inside the business hold owner access to each. Lost access is one of the most expensive problems I'm asked to untangle, and it's easiest to prevent while everyone is still reachable.&lt;/p&gt;

&lt;h2&gt;11. Write it down&lt;/h2&gt;
&lt;p&gt;End with one page of notes: what you checked, what you fixed, and what needs a developer or a budget next year. That page becomes the start of next year's plan and the baseline for next December. Keep it with last year's, and the trend over several Decembers tells you more than any single report.&lt;/p&gt;

&lt;h2&gt;Save the big launches for January&lt;/h2&gt;
&lt;p&gt;A redesign or migration launched the week before a holiday shutdown breaks with nobody around to notice for two weeks. December is for findings; January is for shipping the bigger changes.&lt;/p&gt;

&lt;p&gt;Prefer to hand this off? I run this check as a fixed-price review and leave you with the written list. &lt;a href="https://gerriscorp.com/contact/"&gt;Tell me about your site&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/webmasters/answer/7440203"&gt;Search Console Help: Page indexing report&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/webmasters/answer/7451001"&gt;Search Console Help: Sitemaps report&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/analytics/answer/7667196"&gt;Analytics Help: Data retention&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/business/answer/6303076"&gt;Google Business Profile Help: How to set special hours&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Deleting hundreds of location pages: why core rankings fall, and the fix</title>
    <link>https://gerriscorp.com/guides/removed-location-pages/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/removed-location-pages/</guid>
    <pubDate>Wed, 07 Oct 2026 12:00:00 +0000</pubDate>
    <description>What happens when a site prunes hundreds of city or area code landing pages: why core pages can drop, how to decide each URL's fate, and how to recover.</description>
    <content:encoded>&lt;h1&gt;Deleting hundreds of location pages: why core rankings fall, and the fix&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-07"&gt;October 7, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;It's one of the most common stories I hear from service businesses. Years ago someone built a landing page for every city, county, or area code the company serves. Later, someone decided they were thin, or risky, or just embarrassing, and deleted them. Months afterward the important pages, the ones nobody touched, are ranking lower. The deletion may be the cause, a coincidence, or one of several causes, and the only way to know is to look.&lt;/p&gt;

&lt;h2&gt;Why the pages existed, and why removing them can be right&lt;/h2&gt;
&lt;p&gt;Hundreds of nearly identical pages that differ only by a place name are close to what Google's spam policies call doorway abuse: pages made to rank for many similar searches that funnel visitors to the same place. Pruning them is often the correct long-term decision. The damage comes from how the pages are removed.&lt;/p&gt;

&lt;h2&gt;Five ways a cleanup hurts the pages you kept&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Lost internal links.&lt;/strong&gt; Every location page usually linked to the core service pages. Delete five hundred pages and you delete thousands of internal links pointing at the pages you care about.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Lost inbound links.&lt;/strong&gt; Some of those pages earned links from local directories, partners, or press. If they now return errors, that value stops flowing.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Redirects to nowhere useful.&lt;/strong&gt; Sending every deleted page to the home page looks tidy, but Google can treat irrelevant redirects much like missing pages.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Relevance that moved with them.&lt;/strong&gt; A few of those pages may have quietly ranked for valuable searches the core pages never covered.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;A coincidence in the timeline.&lt;/strong&gt; Large removals often happen alongside redesigns, template changes, or algorithm updates, and the losses get blamed on whichever change people remember.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;Step one: rebuild the inventory&lt;/h2&gt;
&lt;p&gt;Before deciding anything, list every removed URL. If the old list is gone, rebuild it from Search Console's historical performance data, old XML sitemaps, archived crawls, server logs, and analytics landing page reports. For each URL record its traffic and impressions before removal, any external links pointing to it, and what it returns today: a redirect, a 404, a 410, or a soft 404.&lt;/p&gt;

&lt;h2&gt;Step two: line up the dates&lt;/h2&gt;
&lt;p&gt;Put the removal date, any redesign or migration, and known Google updates on one timeline next to the core pages' clicks and positions. If the decline started before the removal, the deletion didn't start it. If it began within weeks after, it's a strong suspect. Treat any unusual query patterns as a separate track until the data connects them.&lt;/p&gt;

&lt;h2&gt;Step three: decide each URL's fate&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Restore&lt;/strong&gt; the handful that earned real traffic or links and can be made genuinely useful, with local details only that location can offer.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Redirect&lt;/strong&gt; pages with links or traffic to the closest relevant page, such as a regional service page, using a 301.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Let go&lt;/strong&gt; of the rest with a 404 or 410. Google treats both as signals to drop the URL, and that's fine for pages that had no value.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Every URL gets its own decision, recorded in the inventory.&lt;/p&gt;

&lt;h2&gt;Step four: rebuild the paths&lt;/h2&gt;
&lt;p&gt;Replace the internal links the old pages provided. A service area page that lists the regions you cover, linked from the main navigation or footer, and service pages that mention where you work, put prominence back on the pages that matter without recreating hundreds of thin ones.&lt;/p&gt;

&lt;h2&gt;Step five: measure the right things&lt;/h2&gt;
&lt;p&gt;Watch the core pages' impressions, clicks, and positions for the searches they're meant to win, along with leads. Expect weeks, not days, before Google recrawls and reassesses. Keep a short change log so each improvement can be tied to the change that caused it.&lt;/p&gt;

&lt;h2&gt;What a useful location page looks like&lt;/h2&gt;
&lt;p&gt;If some locations deserve a page, give each one facts no other page has: the team or crew that serves it, real jobs completed there, local regulations or permits, travel times, and reviews from customers in that area. A page that would read the same with a different town name swapped in is the kind that caused the cleanup in the first place.&lt;/p&gt;

&lt;h2&gt;If you haven't deleted them yet&lt;/h2&gt;
&lt;p&gt;Do the inventory first. Decide each page's fate with the data in hand, set up the redirects and internal links before anything is removed, and remove in stages so you can see the effect of each batch.&lt;/p&gt;

&lt;p&gt;Already deleted and watching rankings slide? This is the kind of forensic work I do. See &lt;a href="https://gerriscorp.com/guides/traffic-drop/"&gt;how to diagnose a traffic drop&lt;/a&gt;, or &lt;a href="https://gerriscorp.com/contact/"&gt;tell me what happened and when&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/essentials/spam-policies"&gt;Google Search Central: Spam policies for Google web search&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/http-network-errors"&gt;Google Search Central: How HTTP status codes affect Google&amp;#x27;s crawlers&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/301-redirects"&gt;Google Search Central: Redirects and Google Search&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to diagnose a Google traffic drop in Search Console</title>
    <link>https://gerriscorp.com/guides/traffic-drop/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/traffic-drop/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>A step-by-step method for finding why organic traffic fell: confirm the drop is real, isolate pages and queries, check updates and indexing, and name the cause.</description>
    <content:encoded>&lt;h1&gt;How to diagnose a Google traffic drop in Search Console&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Organic traffic fell and everyone wants an answer by Friday. The answer is almost always findable, and it comes from narrowing the drop down until only one explanation fits. This is the order I work in.&lt;/p&gt;

&lt;h2&gt;1. Confirm the drop is real&lt;/h2&gt;
&lt;p&gt;Start by ruling out measurement. Analytics can fall while search traffic is fine.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;Compare Google Analytics with Search Console clicks over the same dates. If Search Console is flat and analytics fell, the problem is tracking: a broken tag, a new consent banner, a site redesign that dropped the snippet from some templates.&lt;/li&gt;
  &lt;li&gt;Check that the date ranges match in length and in days of the week. A Monday-to-Sunday week compared with a Thursday-to-Wednesday week shows a false drop on a business site.&lt;/li&gt;
  &lt;li&gt;Compare with the same period last year. Many businesses fall every summer or every January, and that's seasonality.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;One reporting change to know about: in September 2025 Google stopped supporting a parameter that rank-tracking tools used to load a hundred results at a time. Desktop impressions fell sharply for many sites and average position appeared to improve, with no change in clicks. A drop in impressions alone from that month is usually that.&lt;/p&gt;

&lt;h2&gt;2. Read clicks, impressions, CTR, and position together&lt;/h2&gt;
&lt;table&gt;
&lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Pattern&lt;/th&gt;&lt;th scope="col"&gt;Likely meaning&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;&lt;td&gt;Impressions and clicks both fell, position fell&lt;/td&gt;&lt;td&gt;Rankings dropped: an algorithm update, a technical problem, or stronger competition&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Impressions steady, clicks and CTR fell&lt;/td&gt;&lt;td&gt;The results page changed: an AI Overview, more ads, a new feature above you, or a rewritten title&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Impressions fell, position steady&lt;/td&gt;&lt;td&gt;Fewer people are searching: seasonality, a news cycle ending, or a product falling out of fashion&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Everything fell to near zero&lt;/td&gt;&lt;td&gt;The site is blocked, deindexed, or under a manual action&lt;/td&gt;&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;

&lt;h2&gt;3. Find where the loss is&lt;/h2&gt;
&lt;p&gt;Use the Compare option in the Performance report, then sort each tab by the change in clicks.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Pages.&lt;/strong&gt; Is the loss spread across the site, or concentrated in one section or template? A drop in one template points to something technical about that template.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Queries.&lt;/strong&gt; Separate branded searches from everything else with a regular expression filter on the brand name. Branded loss means fewer people are looking for the business; unbranded loss means rankings.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Countries and devices.&lt;/strong&gt; A drop on mobile alone suggests a mobile rendering or speed problem. A drop in one country suggests hreflang, a local competitor, or a regional update.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Search appearance and search type.&lt;/strong&gt; Lost rich results, or lost image and video traffic, show up here.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;4. Line it up with dates&lt;/h2&gt;
&lt;p&gt;Find the first day of the decline as precisely as the daily data allows, then check what happened that week.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Google's Search Status Dashboard&lt;/strong&gt; lists confirmed core updates, spam updates, and incidents with start and end dates.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Your own changes.&lt;/strong&gt; Deploys, plugin and theme updates, CMS upgrades, migrations, CDN or security changes, new consent tools, and changes to robots.txt or redirects. Ask the developers for the deploy log.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Outside events.&lt;/strong&gt; A competitor's launch, a press story, a product recall.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;5. Check the technical reports&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Security and Manual actions.&lt;/strong&gt; One look, and either there's a notice or there isn't.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Page indexing.&lt;/strong&gt; A rise in "Crawled, currently not indexed," "Excluded by noindex," "Blocked by robots.txt," or "Duplicate, Google chose different canonical" that starts on the same date is a strong lead.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Crawl stats.&lt;/strong&gt; Under Settings: host status problems, a spike in server errors, slower response times, or a sudden change in crawl volume.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;URL Inspection.&lt;/strong&gt; Run a live test on a page that lost traffic and read the rendered HTML. Content missing from Google's render is a common, invisible cause. The &lt;a href="https://gerriscorp.com/guides/javascript-seo/"&gt;JavaScript SEO guide&lt;/a&gt; has the tests.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;6. Separate the problems&lt;/h2&gt;
&lt;p&gt;Drops often have more than one cause at once: a core update lands the same month a redesign broke a template, or a backlink spam spike shows up beside a redirect error. Treat each candidate as its own hypothesis and test it against the data separately. When the evidence for each one points to different pages or different dates, they're different problems, and each gets its own fix.&lt;/p&gt;

&lt;h2&gt;7. Write it down&lt;/h2&gt;
&lt;p&gt;The deliverable is a short document: the size and start date of the drop, where it is concentrated, the cause, the evidence for it, the explanations that were ruled out and why, and the fix with tickets the developers can act on. Then mark the fix's ship date and watch the same report, filtered the same way, over the following weeks.&lt;/p&gt;

&lt;section class="faq"&gt;
&lt;h2&gt;Questions about traffic drops&lt;/h2&gt;
&lt;h3&gt;How fast does traffic recover after a fix?&lt;/h3&gt;
&lt;p&gt;A technical fix such as an unblocked template often recovers within weeks, as Google recrawls. Losses from a core update usually move only when Google runs a later update, after the content has improved.&lt;/p&gt;
&lt;h3&gt;How far back does Search Console go?&lt;/h3&gt;
&lt;p&gt;Sixteen months. For longer history, set up the bulk export to BigQuery now, because it only collects data from the day it starts.&lt;/p&gt;
&lt;/section&gt;

&lt;p class="related"&gt;Finding the cause of lost traffic is the core of &lt;a href="https://gerriscorp.com/services/technical-seo/"&gt;technical SEO forensics&lt;/a&gt;. See also the &lt;a href="https://gerriscorp.com/case-studies/indexing-recovery/"&gt;indexing recovery case study&lt;/a&gt;, and if the drop followed a relaunch, &lt;a href="https://gerriscorp.com/guides/migration-recovery/"&gt;recovering traffic after a site migration&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/monitor-debug/debugging-search-traffic-drops"&gt;Google Search Central: Debugging drops in Google Search traffic&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://status.search.google.com/"&gt;Google Search Status Dashboard&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/webmasters/answer/7440203"&gt;Search Console Help: Page indexing report&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Structured data in 2026: which schema markup is worth adding</title>
    <link>https://gerriscorp.com/guides/structured-data/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/structured-data/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>Which schema.org markup still earns rich results in Google, which types were retired, how JSON LD helps search engines and AI understand entities, and testing.</description>
    <content:encoded>&lt;h1&gt;Structured data in 2026: which schema markup is worth adding&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Structured data is a machine-readable description of what a page is about: this is a product with this price, this is a business at this address, this article was written by this person. It can earn enhanced listings in Google, and it helps every system that reads the page, including AI assistants, understand the facts without guessing.&lt;/p&gt;

&lt;h2&gt;The format&lt;/h2&gt;
&lt;p&gt;Use JSON-LD: a block of JSON in a &lt;code&gt;script&lt;/code&gt; tag with the type &lt;code&gt;application/ld+json&lt;/code&gt;, using the vocabulary from schema.org. Google supports Microdata and RDFa too, and recommends JSON-LD because it sits apart from the visible HTML and is easy to generate from a template. A minimal example for a business:&lt;/p&gt;
&lt;pre&gt;&amp;lt;script type="application/ld+json"&amp;gt;
{
  "@context": "https://schema.org",
  "@type": "ProfessionalService",
  "@id": "https://example.com/#business",
  "name": "Example Consulting",
  "url": "https://example.com/",
  "telephone": "+1 555 0100",
  "address": {"@type": "PostalAddress", "addressLocality": "Arlington",
              "addressRegion": "VA", "postalCode": "22204", "addressCountry": "US"}
}
&amp;lt;/script&amp;gt;&lt;/pre&gt;

&lt;h2&gt;What still earns rich results&lt;/h2&gt;
&lt;p&gt;Google has narrowed the list of types that change how a result looks. In 2023 it limited FAQ rich results to well-known government and health sites and dropped How-to rich results. In 2025 it retired several more, including Claim Review, Estimated Salary, and Vehicle Listing, and stopped showing breadcrumb trails on mobile results. The types that still produce visible features include:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Product&lt;/strong&gt;, with offers, price, availability, shipping, and return policy, for product snippets and merchant listings.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Review snippets and aggregate ratings&lt;/strong&gt; on products, recipes, books, software, and other eligible types. Reviews of your own business on your own site are excluded.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Organization and LocalBusiness&lt;/strong&gt;, for the logo and business details Google uses in knowledge panels.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Article&lt;/strong&gt;, for headline, image, dates, and author.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Event, JobPosting, Recipe, VideoObject, and Course list&lt;/strong&gt;, each tied to its own search feature.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;BreadcrumbList&lt;/strong&gt;, still shown on desktop and still useful for understanding site structure.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Google's search gallery in its developer documentation is the current list, and it changes, so check it before building anything around a feature.&lt;/p&gt;

&lt;h2&gt;Markup without a rich result still helps&lt;/h2&gt;
&lt;p&gt;A retired rich result doesn't make the markup useless. FAQPage markup no longer produces dropdowns for most sites, and it still states plainly which question each answer belongs to. Search engines use structured data to understand entities: who the business is, who wrote the page, which person is which, and how they relate. Microsoft's Bing team has said that schema markup helps its language models understand content. When an assistant has to decide whether two mentions refer to the same business, consistent structured data across the web removes doubt.&lt;/p&gt;

&lt;h2&gt;Build a connected graph&lt;/h2&gt;
&lt;p&gt;The strongest pattern is one graph per page, where each thing has an &lt;code&gt;@id&lt;/code&gt; and refers to the others by it. The WebSite is published by the Organization. The Organization is founded by the Person. The Article is written by the Person and is about the Service. Every page on this site carries a graph like that, built from one template, so the facts never contradict each other.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;sameAs&lt;/strong&gt; links an entity to its profiles elsewhere: a Wikipedia article, Wikidata, a professional profile, a Google Business Profile. Those links are how systems confirm identity.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Stable identifiers.&lt;/strong&gt; Use the same &lt;code&gt;@id&lt;/code&gt; for the same entity on every page.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The same facts everywhere.&lt;/strong&gt; Name, address, phone, founding date, and founder should match the visible page, the business profile, and the directories.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Rules that keep markup safe&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Mark up only what visitors can see.&lt;/strong&gt; Reviews, prices, and FAQs in the markup must appear on the page. Hidden or invented markup can earn a manual action for spammy structured data, which removes rich results for the whole site.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Keep it current.&lt;/strong&gt; A price or an event date in the markup that disagrees with the page is a quality problem. Generate the markup from the same data as the page.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Server-rendered.&lt;/strong&gt; Google can read JSON-LD injected by JavaScript, and many other crawlers can't. Put it in the HTML.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;How to validate it&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Google's Rich Results Test&lt;/strong&gt; shows which rich results a page is eligible for and flags missing required properties.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The Schema Markup Validator&lt;/strong&gt; at validator.schema.org checks any schema.org markup against the full vocabulary, including types Google ignores.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Search Console&lt;/strong&gt; lists each detected rich result type under Enhancements, with valid and invalid counts across the site and alerts when a template breaks.&lt;/li&gt;
&lt;/ul&gt;

&lt;section class="faq"&gt;
&lt;h2&gt;Questions about structured data&lt;/h2&gt;
&lt;h3&gt;Is structured data a ranking factor?&lt;/h3&gt;
&lt;p&gt;Google says it isn't a direct ranking factor. It affects how results look and how well Google understands the page, and both of those affect clicks.&lt;/p&gt;
&lt;h3&gt;Can a plugin handle it?&lt;/h3&gt;
&lt;p&gt;Yoast, Rank Math, and Shopify themes add sensible defaults for common types. For a connected graph, custom business types, or multiple plugins that each add their own markup, it usually needs a hand-built template.&lt;/p&gt;
&lt;/section&gt;

&lt;p class="related"&gt;I build schema and entity markup as part of &lt;a href="https://gerriscorp.com/services/on-page-seo/"&gt;on-page SEO&lt;/a&gt; and &lt;a href="https://gerriscorp.com/services/ai-search/"&gt;AI search visibility&lt;/a&gt; work.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data"&gt;Google Search Central: Introduction to structured data&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/structured-data/local-business"&gt;Google Search Central: Local business structured data&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How robots.txt works, and the mistakes that cost traffic</title>
    <link>https://gerriscorp.com/guides/robots-txt/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/robots-txt/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>How robots.txt rules are matched, why a Disallow rule still lets pages be indexed, how to handle AI crawlers, and how to test robots.txt changes before launch.</description>
    <content:encoded>&lt;h1&gt;How robots.txt works, and the mistakes that cost traffic&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;robots.txt is a plain text file at the root of a site that tells crawlers which paths they may fetch. It's small, it's public, and a single wrong line in it can take a site out of Google. Here's how it actually behaves.&lt;/p&gt;

&lt;h2&gt;Where it lives and what it covers&lt;/h2&gt;
&lt;p&gt;The file must sit at the root of the host, such as &lt;code&gt;https://example.com/robots.txt&lt;/code&gt;. It applies only to that exact protocol and host: &lt;code&gt;www.example.com&lt;/code&gt;, &lt;code&gt;shop.example.com&lt;/code&gt;, and the HTTP version each need their own. Google reads up to 500 kibibytes of it and generally caches it for up to a day, so changes take effect within about 24 hours.&lt;/p&gt;

&lt;h2&gt;The syntax&lt;/h2&gt;
&lt;pre&gt;User-agent: *
Disallow: /cart/
Disallow: /search
Allow: /search/help

User-agent: GPTBot
Disallow: /

Sitemap: https://example.com/sitemap.xml&lt;/pre&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;User-agent&lt;/strong&gt; starts a group of rules for a named crawler. &lt;code&gt;*&lt;/code&gt; means every crawler that has no group of its own. A crawler obeys only the most specific group that names it and ignores the rest, so a bot with its own group skips everything under &lt;code&gt;*&lt;/code&gt;.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Disallow&lt;/strong&gt; and &lt;strong&gt;Allow&lt;/strong&gt; take a path prefix. &lt;code&gt;Disallow: /search&lt;/code&gt; blocks &lt;code&gt;/search&lt;/code&gt;, &lt;code&gt;/search?q=shoes&lt;/code&gt;, and &lt;code&gt;/searching&lt;/code&gt;.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Wildcards.&lt;/strong&gt; &lt;code&gt;*&lt;/code&gt; matches any characters, and &lt;code&gt;$&lt;/code&gt; anchors the end. &lt;code&gt;Disallow: /*.pdf$&lt;/code&gt; blocks every URL ending in .pdf.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Conflicts.&lt;/strong&gt; When an Allow and a Disallow both match, Google uses the longer, more specific rule, and Allow wins a tie.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Sitemap&lt;/strong&gt; lines can appear anywhere and apply to every crawler.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Google ignores &lt;code&gt;Crawl-delay&lt;/code&gt;; Bing honors it. Google also stopped supporting &lt;code&gt;Noindex&lt;/code&gt; lines in robots.txt in 2019.&lt;/p&gt;

&lt;h2&gt;Blocking crawling is different from blocking indexing&lt;/h2&gt;
&lt;p&gt;This is the most expensive misunderstanding. Disallow stops a crawler from fetching a page. It doesn't stop the URL from being indexed. If other pages link to a blocked URL, Google can index the address alone, with no content, and Search Console reports it as "Indexed, though blocked by robots.txt."&lt;/p&gt;
&lt;p&gt;To keep a page out of results, let it be crawled and give it a &lt;code&gt;noindex&lt;/code&gt; robots meta tag or an &lt;code&gt;X-Robots-Tag: noindex&lt;/code&gt; header. Google has to fetch the page to see that instruction, so a page that's both disallowed and noindexed stays in the index, because the noindex is never read. For content that must stay private, use a login. robots.txt is public, and listing secret paths in it advertises them.&lt;/p&gt;

&lt;h2&gt;Status codes matter&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;200&lt;/strong&gt;: Google follows the rules in the file.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;404 or another 4xx&lt;/strong&gt;: Google treats the site as having no restrictions and crawls everything.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;500 or another 5xx, or a timeout&lt;/strong&gt;: Google treats the whole site as disallowed and pauses crawling. If the error lasts, Google falls back to its last good copy, and after about a month without one it treats the site as unrestricted.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;A server or firewall that errors only on robots.txt can quietly stop all crawling, which is why it's one of the first URLs I check on a site that lost traffic.&lt;/p&gt;

&lt;h2&gt;AI crawlers&lt;/h2&gt;
&lt;p&gt;Each AI company publishes the user agents it uses, and each honors robots.txt. They split into crawlers that gather training data, such as GPTBot, ClaudeBot, and CCBot, and crawlers that fetch pages for live answers, such as OAI-SearchBot, Claude-SearchBot, and PerplexityBot. Google-Extended and Applebot-Extended are tokens rather than separate crawlers: they control whether content fetched by Googlebot and Applebot may be used for AI models, and blocking them leaves ordinary search crawling alone. The &lt;a href="https://gerriscorp.com/guides/ai-assistants/"&gt;AI assistants guide&lt;/a&gt; lists which crawler feeds which product.&lt;/p&gt;
&lt;p&gt;This site lets every AI crawler read everything, and keeps the Markdown copies of its pages out of traditional search engines, so they never compete with the HTML pages. Because a crawler follows only its own group, naming the search engines in one group and the AI crawlers in another gives each its own rules:&lt;/p&gt;
&lt;pre&gt;User-agent: Googlebot
User-agent: Bingbot
Allow: /
Disallow: /*.md$

User-agent: GPTBot
User-agent: ClaudeBot
User-agent: PerplexityBot
Allow: /&lt;/pre&gt;

&lt;h2&gt;Mistakes I find most often&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;The staging file goes live.&lt;/strong&gt; &lt;code&gt;Disallow: /&lt;/code&gt; protects a staging site, gets copied to production at launch, and blocks everything. Check robots.txt on launch day before anything else.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Blocking CSS and JavaScript.&lt;/strong&gt; Google renders pages, and a disallowed script folder can leave it looking at a broken layout with missing content.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Blocking parameters that carry real pages.&lt;/strong&gt; A broad rule like &lt;code&gt;Disallow: /*?&lt;/code&gt; also blocks paginated categories and filtered pages that rank.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;A bot's own group dropping the general rules.&lt;/strong&gt; Adding a group for one crawler without repeating the shared rules gives that crawler free run of everything else.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Testing changes&lt;/h2&gt;
&lt;p&gt;Search Console's robots.txt report shows the version Google last fetched, when it fetched it, and any parsing errors. Before changing a live file, test the proposed rules against a list of important URLs with a parser that follows Google's rules, such as Google's open-source robots.txt library or a crawler like Screaming Frog with a custom robots.txt. Then confirm the change in Search Console after Google picks it up.&lt;/p&gt;

&lt;p class="related"&gt;robots.txt is part of every &lt;a href="https://gerriscorp.com/services/crawler-visibility/"&gt;crawler visibility audit&lt;/a&gt; and every &lt;a href="https://gerriscorp.com/services/technical-seo/"&gt;technical SEO&lt;/a&gt; engagement I run.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/robots/intro"&gt;Google Search Central: Introduction to robots.txt&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://platform.openai.com/docs/bots"&gt;OpenAI: Overview of OpenAI crawlers&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Lynx for website QA: text browser checks before every launch</title>
    <link>https://gerriscorp.com/guides/lynx-qa/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/lynx-qa/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>How to use the Lynx text browser to check content, links, and noindex tags across a whole site before launch, plus online options for teams without a terminal.</description>
    <content:encoded>&lt;h1&gt;Lynx for website QA: text browser checks before every launch&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Lynx has been around since the early nineties, and it's still one of the most useful tools in my launch kit. Because it ignores styling and never runs JavaScript, it strips a page down to what the server actually sends. Scripted from the command line, it can check every page of a site in minutes, which makes it a natural fit for redesigns, migrations, and release checklists.&lt;/p&gt;

&lt;h2&gt;Getting Lynx&lt;/h2&gt;
&lt;p&gt;I run Lynx on a Linux server over SSH, from an old ThinkPad, and that's the simplest setup. On Debian or Ubuntu:&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;sudo apt update &amp;amp;&amp;amp; sudo apt install lynx&lt;/code&gt;&lt;/pre&gt;
&lt;p&gt;Other systems: &lt;code&gt;brew install lynx&lt;/code&gt; on macOS with Homebrew, &lt;code&gt;sudo dnf install lynx&lt;/code&gt; on Fedora, and on Windows, the same apt command inside the Windows Subsystem for Linux. Once it's installed, &lt;code&gt;lynx --version&lt;/code&gt; confirms it works.&lt;/p&gt;

&lt;h2&gt;Commands worth memorizing&lt;/h2&gt;
&lt;table&gt;
&lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Command&lt;/th&gt;&lt;th scope="col"&gt;What it gives you&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;&lt;td&gt;&lt;code&gt;lynx URL&lt;/code&gt;&lt;/td&gt;&lt;td&gt;An interactive text browser. Arrows move, Enter follows a link, &lt;kbd&gt;q&lt;/kbd&gt; quits, and &lt;kbd&gt;\&lt;/kbd&gt; toggles the HTML source.&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;&lt;code&gt;lynx -dump URL&lt;/code&gt;&lt;/td&gt;&lt;td&gt;The rendered text, followed by a numbered list of every link.&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;&lt;code&gt;lynx -dump -nolist URL&lt;/code&gt;&lt;/td&gt;&lt;td&gt;Text only, no link list. Best for comparing two versions of a page.&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;&lt;code&gt;lynx -dump -listonly URL&lt;/code&gt;&lt;/td&gt;&lt;td&gt;Only the links, numbered, with their full addresses.&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;&lt;code&gt;lynx -source URL&lt;/code&gt;&lt;/td&gt;&lt;td&gt;The raw HTML, handy for checking head tags.&lt;/td&gt;&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;

&lt;h2&gt;Five checks I script before a launch&lt;/h2&gt;
&lt;h3&gt;1. Every page in the sitemap has real text&lt;/h3&gt;
&lt;p&gt;Pull the URLs from the sitemap and count the words Lynx sees on each. A product page that comes back with forty words when it should have four hundred is a template hiding its content behind scripts.&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;curl -s https://example.com/sitemap.xml | grep -o '&amp;lt;loc&amp;gt;[^&amp;lt;]*' | sed 's/&amp;lt;loc&amp;gt;//' |
while read u; do echo "$(lynx -dump -nolist "$u" | wc -w) $u"; done | sort -n | head&lt;/code&gt;&lt;/pre&gt;
&lt;p&gt;The thinnest pages float to the top of the list.&lt;/p&gt;
&lt;h3&gt;2. Staging and production say the same thing&lt;/h3&gt;
&lt;p&gt;After a migration, compare the old page with the new one as text. Missing paragraphs, dropped FAQs, and lost disclaimers show up immediately, without any styling noise.&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;diff &amp;lt;(lynx -dump -nolist https://old.example.com/pricing/) \
     &amp;lt;(lynx -dump -nolist https://www.example.com/pricing/)&lt;/code&gt;&lt;/pre&gt;
&lt;h3&gt;3. No stray noindex survived from staging&lt;/h3&gt;
&lt;p&gt;The most expensive launch mistake I know is a robots noindex tag carried over from a staging site. One loop over the sitemap catches it:&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;... | while read u; do lynx -source "$u" | grep -qi 'noindex' &amp;amp;&amp;amp; echo "NOINDEX $u"; done&lt;/code&gt;&lt;/pre&gt;
&lt;h3&gt;4. Links are real, and they all resolve&lt;/h3&gt;
&lt;p&gt;The link list shows only genuine anchor links, so navigation built from click handlers simply won't appear. Feed the list to curl to catch broken targets and redirect chains:&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;lynx -dump -listonly https://example.com/ | awk '/https?:/{print $2}' | sort -u |
xargs -n1 curl -s -o /dev/null -w '%{http_code} %{url_effective}\n' | grep -v '^200'&lt;/code&gt;&lt;/pre&gt;
&lt;h3&gt;5. The first screen is about the page&lt;/h3&gt;
&lt;p&gt;Read the top twenty lines of each key template's dump. If a consent banner, a mega menu, and a promotional strip come before the first heading, assistive technology and crawlers wade through all of it too. Reordering the markup, or adding a skip link, fixes the problem for both.&lt;/p&gt;

&lt;h2&gt;Reading the results with a team&lt;/h2&gt;
&lt;p&gt;Text dumps are easy to share, and they make strong evidence in a developer ticket: here's the page as a crawler receives it, and here's the paragraph that's missing. Nobody has to install anything to read a text file, and the comparison is hard to argue with.&lt;/p&gt;

&lt;h2&gt;When a terminal isn't an option&lt;/h2&gt;
&lt;p&gt;Plenty of teams can't or won't install command-line tools. These get most of the way there:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;a href="https://www.delorie.com/web/lynxview.html"&gt;Lynx Viewer&lt;/a&gt;, a free web service that renders a public URL much as Lynx would.&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.seo-browser.com/"&gt;SEO Browser&lt;/a&gt;, which places the browser's version of a page beside a crawler's and highlights what JavaScript added.&lt;/li&gt;
  &lt;li&gt;The &lt;a href="https://search.google.com/test/rich-results"&gt;Rich Results Test&lt;/a&gt; and Search Console's URL Inspection, which both show the HTML Google rendered.&lt;/li&gt;
  &lt;li&gt;The &lt;a href="https://chrispederick.com/work/web-developer/"&gt;Web Developer extension&lt;/a&gt; for Chrome and Firefox, which turns off CSS and JavaScript with a click.&lt;/li&gt;
  &lt;li&gt;Accessibility checkers and screen readers: &lt;a href="https://wave.webaim.org/"&gt;WAVE&lt;/a&gt; offers a no-styles view, and &lt;a href="https://www.nvaccess.org/"&gt;NVDA&lt;/a&gt; or Apple's VoiceOver read a page in the same top-to-bottom order Lynx displays. An accessibility auditor will usually run some version of this test anyway.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;&lt;strong&gt;Here's the short version: if a page works in Lynx, you're way more than halfway there.&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;Add it to the checklist&lt;/h2&gt;
&lt;p&gt;I add a Lynx pass to every launch: sitemap word counts, a noindex sweep, a link check, and a text diff of the most important templates. It takes about ten minutes and catches problems a styled browser hides. For the full launch sequence, see &lt;a href="https://gerriscorp.com/guides/migration-checklist/"&gt;the website migration SEO checklist&lt;/a&gt;, and for the deeper rendering questions, &lt;a href="https://gerriscorp.com/guides/javascript-seo/"&gt;how to tell if JavaScript is hiding your content&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://lynx.invisible-island.net/"&gt;Lynx project home&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://packages.debian.org/stable/lynx"&gt;Debian package: lynx&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://formulae.brew.sh/formula/lynx"&gt;Homebrew formula: lynx&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://learn.microsoft.com/en-us/windows/wsl/install"&gt;Microsoft Learn: Install WSL&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.w3.org/WAI/test-evaluate/preliminary/"&gt;W3C: Easy checks, a first review of web accessibility&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>hreflang for English-language sites serving the US, UK, Canada, and Australia</title>
    <link>https://gerriscorp.com/guides/hreflang/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/hreflang/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>When an English language business needs hreflang for the US, UK, Canada, Australia, Ireland, and New Zealand, how to write the tags, and errors that void it.</description>
    <content:encoded>&lt;h1&gt;hreflang for English-language sites serving the US, UK, Canada, and Australia&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A business that sells in several English-speaking countries often ends up with near-identical pages for each one. hreflang tells Google which version belongs to which audience, so a shopper in Leeds sees prices in pounds and a shopper in Toronto sees Canadian dollars. Written wrong, it does nothing at all.&lt;/p&gt;

&lt;h2&gt;Do you need it?&lt;/h2&gt;
&lt;p&gt;hreflang is for sites with separate versions of a page for different languages or regions. One English site that serves everyone the same content needs none. The tags start paying off when the versions differ in ways that matter to the visitor:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;Prices and currency&lt;/li&gt;
  &lt;li&gt;Shipping, taxes, returns, and delivery times&lt;/li&gt;
  &lt;li&gt;Legal terms, regulated claims, and contact details&lt;/li&gt;
  &lt;li&gt;Spelling and vocabulary: color and colour, sneakers and trainers, apartment and flat&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Without hreflang, Google sees several pages saying almost the same thing, may pick one as canonical for every country, and shows the American page in Australia. With it, Google treats the pages as a set and swaps in the right member for each searcher.&lt;/p&gt;

&lt;h2&gt;The codes&lt;/h2&gt;
&lt;p&gt;Each value is an ISO 639-1 language code, optionally followed by an ISO 3166-1 country code:&lt;/p&gt;
&lt;table&gt;
&lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Audience&lt;/th&gt;&lt;th scope="col"&gt;Code&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;&lt;td&gt;United States&lt;/td&gt;&lt;td&gt;&lt;code&gt;en-US&lt;/code&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;United Kingdom&lt;/td&gt;&lt;td&gt;&lt;code&gt;en-GB&lt;/code&gt; (never en-UK, which isn't a valid code)&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Canada, English&lt;/td&gt;&lt;td&gt;&lt;code&gt;en-CA&lt;/code&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Canada, French&lt;/td&gt;&lt;td&gt;&lt;code&gt;fr-CA&lt;/code&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Australia&lt;/td&gt;&lt;td&gt;&lt;code&gt;en-AU&lt;/code&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;New Zealand&lt;/td&gt;&lt;td&gt;&lt;code&gt;en-NZ&lt;/code&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Ireland&lt;/td&gt;&lt;td&gt;&lt;code&gt;en-IE&lt;/code&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;English speakers anywhere else&lt;/td&gt;&lt;td&gt;&lt;code&gt;en&lt;/code&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Fallback for everyone unmatched&lt;/td&gt;&lt;td&gt;&lt;code&gt;x-default&lt;/code&gt;&lt;/td&gt;&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;The language is required. A country on its own, such as &lt;code&gt;gb&lt;/code&gt;, is read as a language code, and since there's no language with that code, the annotation is ignored.&lt;/p&gt;

&lt;h2&gt;Writing the tags&lt;/h2&gt;
&lt;p&gt;Every version in the set lists every version, including itself:&lt;/p&gt;
&lt;pre&gt;&amp;lt;link rel="alternate" hreflang="en-US" href="https://example.com/us/boots/"&amp;gt;
&amp;lt;link rel="alternate" hreflang="en-GB" href="https://example.com/uk/boots/"&amp;gt;
&amp;lt;link rel="alternate" hreflang="en-AU" href="https://example.com/au/boots/"&amp;gt;
&amp;lt;link rel="alternate" hreflang="x-default" href="https://example.com/boots/"&amp;gt;&lt;/pre&gt;
&lt;p&gt;The same block goes on the US, UK, and Australian pages. The annotations can also live in the XML sitemap, which is easier to maintain on large sites, or in HTTP headers for files such as PDFs. Pick one method. Mixing them invites the two copies to drift apart, and a set that disagrees with itself is ignored.&lt;/p&gt;

&lt;h2&gt;The rules that void it&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Missing return links.&lt;/strong&gt; If the UK page names the US page and the US page doesn't name the UK page back, Google ignores the pair. This is the most common failure.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Pointing at non-canonical URLs.&lt;/strong&gt; Every URL in the set should return a 200 and declare itself as canonical. A UK page whose canonical points at the US page tells Google to drop the UK page, and hreflang can't override that.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Relative or mixed URLs.&lt;/strong&gt; Use full absolute URLs, with the same protocol and host style used everywhere else.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Tags only in JavaScript.&lt;/strong&gt; Put them in the server's HTML or in the sitemap.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;How to structure the versions&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Subfolders&lt;/strong&gt; such as &lt;code&gt;/uk/&lt;/code&gt; and &lt;code&gt;/au/&lt;/code&gt; on one domain share all its authority and are the simplest to run. This is my default recommendation.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Country domains&lt;/strong&gt; such as &lt;code&gt;.co.uk&lt;/code&gt; and &lt;code&gt;.com.au&lt;/code&gt; send the strongest local signal and build trust with local buyers, and each domain has to earn its own links.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Subdomains&lt;/strong&gt; such as &lt;code&gt;uk.example.com&lt;/code&gt; work, with more setup and no real advantage over subfolders.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Don't redirect by location&lt;/h2&gt;
&lt;p&gt;Redirecting visitors automatically by IP address hides versions from Google, because Googlebot crawls mostly from the United States and would only ever see the American pages. Let every URL load for everyone, and offer a banner that suggests the local version. hreflang handles search; the banner handles direct visitors.&lt;/p&gt;

&lt;h2&gt;Bing&lt;/h2&gt;
&lt;p&gt;Bing gives hreflang less weight and looks more at the &lt;code&gt;content-language&lt;/code&gt; meta tag and the page's &lt;code&gt;lang&lt;/code&gt; attribute. Set &lt;code&gt;&amp;lt;html lang="en-GB"&amp;gt;&lt;/code&gt; on UK pages, and so on, so both engines get a clear signal.&lt;/p&gt;

&lt;h2&gt;Checking it&lt;/h2&gt;
&lt;p&gt;Search Console retired its International Targeting report in 2022, so checking falls to crawlers. Screaming Frog and Sitebulb both validate hreflang sets, flag missing return links and non-canonical targets, and export the errors by page. After that, search from each country with a VPN or with Google's country settings and confirm the right page appears. Spot check a few sets by hand as well: open the source of each version and confirm the block of annotations is identical on every page in the set.&lt;/p&gt;

&lt;p class="related"&gt;I sort out international setups as part of &lt;a href="https://gerriscorp.com/services/technical-seo/"&gt;technical SEO&lt;/a&gt; and &lt;a href="https://gerriscorp.com/services/ecommerce/"&gt;ecommerce SEO&lt;/a&gt; engagements, for clients across the English-speaking world.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/specialty/international/localized-versions"&gt;Google Search Central: Tell Google about localized versions of your page&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/specialty/international/managing-multi-regional-sites"&gt;Google Search Central: Managing multi-regional and multilingual sites&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Google Consent Mode v2 for GA4: a practical setup guide</title>
    <link>https://gerriscorp.com/guides/ga4-consent-mode/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/ga4-consent-mode/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>What Consent Mode v2 controls in GA4 and Google Ads, how basic and advanced mode differ, how to set regional defaults, and how to check that it really works.</description>
    <content:encoded>&lt;h1&gt;Google Consent Mode v2 for GA4: a practical setup guide&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Consent Mode is how Google's tags learn whether a visitor agreed to cookies, and how they behave when the answer is no. Set up well, it keeps analytics honest and the site compliant. Set up badly, it either drops most of your data or quietly ignores the visitor's choice.&lt;/p&gt;

&lt;h2&gt;The four signals&lt;/h2&gt;
&lt;table&gt;
&lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Signal&lt;/th&gt;&lt;th scope="col"&gt;Controls&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;&lt;td&gt;&lt;code&gt;analytics_storage&lt;/code&gt;&lt;/td&gt;&lt;td&gt;Cookies for analytics, such as GA4's visitor identifier&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;&lt;code&gt;ad_storage&lt;/code&gt;&lt;/td&gt;&lt;td&gt;Cookies for advertising, such as conversion and remarketing cookies&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;&lt;code&gt;ad_user_data&lt;/code&gt;&lt;/td&gt;&lt;td&gt;Whether visitor data may be sent to Google for advertising&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;&lt;code&gt;ad_personalization&lt;/code&gt;&lt;/td&gt;&lt;td&gt;Whether that data may be used for personalized ads and remarketing&lt;/td&gt;&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;Version 2 added the last two in late 2023. Since March 2024 Google has required them for advertisers who want to use audience and conversion features with traffic from the European Economic Area, under the EU's Digital Markets Act.&lt;/p&gt;

&lt;h2&gt;Basic and advanced mode&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Basic mode.&lt;/strong&gt; Google's tags don't load at all until the visitor consents. Nothing is collected from people who decline or ignore the banner.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Advanced mode.&lt;/strong&gt; The tags load right away with consent denied and send cookieless pings: no identifiers, no stored cookies. When the visitor accepts, the tags switch to full measurement.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Advanced mode lets GA4 estimate the behavior of visitors who declined, through behavioral modeling, once a property has enough traffic. Google's documented minimum is roughly a thousand daily events from visitors with analytics denied, and a thousand daily users with analytics granted, sustained for a week or more. Smaller sites never reach it, and for them the choice between modes makes little difference to the reports.&lt;/p&gt;

&lt;h2&gt;Setting the defaults&lt;/h2&gt;
&lt;p&gt;The default state must be set before any Google tag loads. With the gtag snippet, that means a &lt;code&gt;consent default&lt;/code&gt; command above the config line. Defaults can differ by region, using ISO country codes:&lt;/p&gt;
&lt;pre&gt;gtag('consent', 'default', {
  ad_storage: 'denied', ad_user_data: 'denied',
  ad_personalization: 'denied', analytics_storage: 'denied',
  region: ['AT','BE','BG','CH','CY','CZ','DE','DK','EE','ES','FI','FR','GB',
           'GR','HR','HU','IE','IS','IT','LI','LT','LU','LV','MT','NL','NO',
           'PL','PT','RO','SE','SI','SK']
});
gtag('consent', 'default', {
  ad_storage: 'denied', ad_user_data: 'denied',
  ad_personalization: 'denied', analytics_storage: 'granted'
});&lt;/pre&gt;
&lt;p&gt;That's the pattern this site uses: advertising signals denied everywhere, because the site runs no ads, and analytics denied by default in the EEA, the UK, and Switzerland, where the law requires prior consent, and granted elsewhere. The most specific matching region wins. When a visitor accepts or declines, the banner sends a &lt;code&gt;consent update&lt;/code&gt; command with the new values.&lt;/p&gt;
&lt;p&gt;In Google Tag Manager, set defaults with a tag on the Consent Initialization trigger, which fires before every other trigger. Most consent platforms certified by Google ship a Tag Manager template that does this for you.&lt;/p&gt;

&lt;h2&gt;Choosing a consent platform&lt;/h2&gt;
&lt;p&gt;Publishers who show Google ads to visitors in the EEA, the UK, or Switzerland must use a consent platform from Google's certified list. For everyone else it's still the safer choice, because certified platforms send Consent Mode signals correctly by default. Check that the platform supports regional rules, so visitors in places that don't require a banner aren't shown one.&lt;/p&gt;

&lt;h2&gt;Checking that it works&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Tag Assistant.&lt;/strong&gt; Connect it to the site and open the Consent tab for each event. It shows the default state, each update, and which tags fired under which state.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The network panel.&lt;/strong&gt; Every GA4 request carries a &lt;code&gt;gcs&lt;/code&gt; parameter. &lt;code&gt;G100&lt;/code&gt; means advertising and analytics storage are both denied; &lt;code&gt;G111&lt;/code&gt; means both granted; &lt;code&gt;G101&lt;/code&gt; means advertising denied and analytics granted.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The cookie list.&lt;/strong&gt; In a fresh private window from a region set to denied, load a page without touching the banner. No &lt;code&gt;_ga&lt;/code&gt; cookie should exist until you accept.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;GA4 admin.&lt;/strong&gt; Under Data collection, the consent settings screen reports whether analytics and advertising consent signals are being received.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Mistakes I find most often&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;The default is set after the tag loads.&lt;/strong&gt; The first hit goes out with no consent state, often with cookies set. Order matters.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Two sources of truth.&lt;/strong&gt; A banner plugin and a hand-coded snippet both set defaults, with different values.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The banner blocks the page.&lt;/strong&gt; A consent script that loads synchronously in the head, or a banner that pushes the layout, hurts Core Web Vitals. Load it asynchronously and show it as an overlay.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Hard-coded tags outside the consent setup.&lt;/strong&gt; An old Universal Analytics snippet, a Facebook pixel, or a heatmap script still firing regardless of the visitor's choice.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;What to expect in the reports&lt;/h2&gt;
&lt;p&gt;After a consent banner goes live in Europe, GA4's European sessions usually fall, sometimes sharply, because declined visitors are no longer counted as users. That's the system working. Search Console doesn't use cookies, so its click counts are unaffected, and comparing the two separates a consent effect from a real traffic loss.&lt;/p&gt;

&lt;p class="related"&gt;Analytics setup and repair is part of &lt;a href="https://gerriscorp.com/services/analytics/"&gt;GA4 and analytics work&lt;/a&gt;. For a consent problem that hid half a site's traffic, see the &lt;a href="https://gerriscorp.com/case-studies/consent-blocking/"&gt;consent blocking case study&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/tag-platform/security/guides/consent"&gt;Google Tag Platform: Set up consent mode&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Faceted navigation SEO: filters, parameters, and crawl budget</title>
    <link>https://gerriscorp.com/guides/faceted-navigation/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/faceted-navigation/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>How product filters create millions of crawlable URLs, when crawl budget becomes a real problem, and how to decide which filter pages deserve to be indexed.</description>
    <content:encoded>&lt;h1&gt;Faceted navigation SEO: filters, parameters, and crawl budget&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Filters are how shoppers narrow a catalog: size, color, brand, price, rating. Every combination can become its own URL, and a store with a few thousand products can produce millions of them. Google spends its time crawling those combinations instead of the pages that sell.&lt;/p&gt;

&lt;h2&gt;How the problem grows&lt;/h2&gt;
&lt;p&gt;A category with ten colors, eight sizes, twelve brands, and five price bands has 4,800 single combinations before sort order, page number, and the order the parameters appear in. Multiply by every category and the site has far more filter URLs than products. Most of them show the same handful of items, many show none, and almost nobody searches for them.&lt;/p&gt;
&lt;p&gt;Google's documentation names faceted navigation as the most common cause of crawling problems it sees, and its December 2024 guidance on the subject is the clearest statement of what it wants.&lt;/p&gt;

&lt;h2&gt;When crawl budget actually matters&lt;/h2&gt;
&lt;p&gt;Google says crawl budget is a concern mainly for sites with more than a million unique pages that change weekly, or more than ten thousand pages that change daily. Smaller sites rarely need to think about it. The signs that a site has crossed the line:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;New products take weeks to appear in Google.&lt;/li&gt;
  &lt;li&gt;Search Console's Page indexing report shows a large and growing count of "Discovered, currently not indexed."&lt;/li&gt;
  &lt;li&gt;Crawl stats show most of Googlebot's requests going to parameter URLs.&lt;/li&gt;
  &lt;li&gt;Server logs show Googlebot fetching filter combinations far more often than product pages.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Server logs are the definitive source. Search Console shows a sample; the logs show every request Googlebot made and where its time went.&lt;/p&gt;

&lt;h2&gt;Decide which filters deserve pages&lt;/h2&gt;
&lt;p&gt;Some filtered views match real searches. "Women's trail running shoes," "red sofa," and "Makita cordless drills" are all a category plus one filter, and each can be a strong landing page. Use keyword research to find the combinations with search demand, usually one filter deep and occasionally two. Give each one a clean, permanent URL such as &lt;code&gt;/shoes/womens/trail/&lt;/code&gt;, a unique title and heading, some introductory text, and links from the category navigation. Everything else stays a convenience for shoppers and stays out of the crawl.&lt;/p&gt;

&lt;h2&gt;Keep the rest away from crawlers&lt;/h2&gt;
&lt;p&gt;Google's guidance offers two approaches for filter URLs that shouldn't be indexed.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Block them in robots.txt.&lt;/strong&gt; Disallow the parameters that create combinations, such as &lt;code&gt;Disallow: /*?*color=&lt;/code&gt;. This stops crawling immediately and saves the most budget. The &lt;a href="https://gerriscorp.com/guides/robots-txt/"&gt;robots.txt guide&lt;/a&gt; covers the syntax.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Use URL fragments.&lt;/strong&gt; Filters applied after a &lt;code&gt;#&lt;/code&gt;, handled by JavaScript, never create new crawlable URLs, because Google ignores fragments.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Other tools are weaker for this job:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Canonical tags&lt;/strong&gt; pointing filter pages at the category consolidate signals over time, and Google still has to crawl every URL to read them, so they save no crawl budget.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;noindex&lt;/strong&gt; keeps pages out of results and also has to be crawled to work. Over time Google crawls noindexed pages less often.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;nofollow on filter links&lt;/strong&gt; discourages crawling and is a hint Google may disregard.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Search Console's URL Parameters tool, which used to handle this, was retired in 2022.&lt;/p&gt;

&lt;h2&gt;If filter pages must be crawlable&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Use the standard &lt;code&gt;&amp;amp;&lt;/code&gt; separator and the same parameter order every time, so one combination has one URL.&lt;/li&gt;
  &lt;li&gt;Return a 404 for combinations with no products, and for nonsense values. A "no results" page with a 200 status becomes a soft 404 and wastes crawls.&lt;/li&gt;
  &lt;li&gt;Keep sort order and items-per-page out of the URL, or block them, since they reshuffle the same products.&lt;/li&gt;
  &lt;li&gt;Limit how many filters can be stacked into one crawlable URL.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Pagination&lt;/h2&gt;
&lt;p&gt;Paginated category pages are different from filters: they list different products, and they're how crawlers reach products deep in a catalog. Keep them crawlable with plain links, let each page declare itself as canonical, and keep pages from going on forever by giving categories sensible sizes.&lt;/p&gt;

&lt;h2&gt;Measure the change&lt;/h2&gt;
&lt;p&gt;Before changing anything, record the share of Googlebot requests going to parameter URLs, the count of discovered and unindexed pages, and how long new products take to be indexed. After the change, watch the same numbers. On large catalogs the shift in crawl toward products shows up within weeks, and the indexing of new products speeds up soon after.&lt;/p&gt;

&lt;p class="related"&gt;Crawl control for big catalogs is the focus of &lt;a href="https://gerriscorp.com/services/large-sites/"&gt;large-site SEO&lt;/a&gt; and &lt;a href="https://gerriscorp.com/services/ecommerce/"&gt;ecommerce SEO&lt;/a&gt;. See also the &lt;a href="https://gerriscorp.com/case-studies/programmatic-indexing/"&gt;programmatic indexing case study&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/crawling-managing-faceted-navigation"&gt;Google Search Central: Managing crawling of faceted navigation URLs&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls"&gt;Google Search Central: How to specify a canonical URL&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Why business email lands in spam, and how to find the cause</title>
    <link>https://gerriscorp.com/guides/email-deliverability/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/email-deliverability/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>How to find out why legitimate business email lands in spam or bounces: authentication results, reputation, blocklists, content, list hygiene, and tools.</description>
    <content:encoded>&lt;h1&gt;Why business email lands in spam, and how to find the cause&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Invoices disappear, proposals land in junk folders, and customers say they never got the reset link. Deliverability failures look random from the outside. They have specific causes, and the evidence for each one is sitting in the message headers and a few free tools.&lt;/p&gt;

&lt;h2&gt;Start with one message's headers&lt;/h2&gt;
&lt;p&gt;Send a message to a Gmail address you control, open it, and choose Show original. Gmail prints three verdicts at the top: SPF, DKIM, and DMARC, each with pass or fail, and the domain each one checked. In Outlook, the same information sits in the message properties under the &lt;code&gt;Authentication-Results&lt;/code&gt; header. Read it before changing anything; it tells you which of the following sections applies.&lt;/p&gt;

&lt;h2&gt;Authentication failures&lt;/h2&gt;
&lt;p&gt;A failing result usually has one of four causes.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;A sender left out of SPF.&lt;/strong&gt; The CRM, the invoicing tool, or the website's contact form sends as your domain and isn't listed in the SPF record.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;SPF over its lookup limit.&lt;/strong&gt; SPF allows ten DNS lookups. Every &lt;code&gt;include&lt;/code&gt; costs at least one, and a record that exceeds the limit fails for every message, including the ones from legitimate senders.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;DKIM never switched on.&lt;/strong&gt; Many services sign with their own domain until you publish their DKIM records and enable signing in their dashboard.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;No alignment.&lt;/strong&gt; DMARC needs SPF or DKIM to pass for the same domain shown in the From line. A service that passes SPF for its own bounce domain still fails DMARC for yours.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;The &lt;a href="https://gerriscorp.com/guides/email-authentication/"&gt;SPF, DKIM, and DMARC guide&lt;/a&gt; covers the records themselves. DMARC aggregate reports, sent daily by the big mailbox providers to the address in your DMARC record, list every source sending as your domain and whether it passed. They're the fastest way to find a sender you forgot.&lt;/p&gt;

&lt;h2&gt;Reputation&lt;/h2&gt;
&lt;p&gt;When authentication passes and mail still lands in spam, the problem is reputation: how the mailbox providers judge mail from your domain and your sending servers, based on how recipients treat it.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Google Postmaster Tools&lt;/strong&gt; shows Gmail's view of a domain that sends in volume: spam complaint rate, authentication rates, and delivery errors. Gmail expects bulk senders to keep reported spam below 0.3 percent, and recommends staying under 0.1.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Microsoft's Smart Network Data Services&lt;/strong&gt; shows how Outlook.com sees the IP addresses you send from, if you control them.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Blocklists.&lt;/strong&gt; Check the domain and the sending IPs against Spamhaus and the other major lists with a lookup tool such as MXToolbox. A listing usually comes with a reason and a removal process.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Shared sending services put many customers on the same IP addresses, so a neighbor's bad behavior can affect you. That's one reason marketing email and everyday business email belong on separate services, and often on separate subdomains, such as &lt;code&gt;news.example.com&lt;/code&gt; for newsletters. A newsletter complaint spike then can't drag down invoices.&lt;/p&gt;

&lt;h2&gt;The sending practices that cause it&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Old or purchased lists.&lt;/strong&gt; Addresses that bounce, and spam traps hidden on old lists, damage reputation quickly. Mail only people who asked for it, and remove addresses that bounce.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;No easy unsubscribe.&lt;/strong&gt; Since 2024 Gmail and Yahoo require bulk senders to support one-click unsubscribe. People who can't leave a list mark it as spam instead.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Sudden volume.&lt;/strong&gt; A domain that sends fifty messages a day and then ten thousand looks compromised. Increase volume over weeks.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Mail nobody opens.&lt;/strong&gt; Providers watch engagement. Stop mailing people who haven't opened anything in months.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Content and links&lt;/h2&gt;
&lt;p&gt;Content matters less than authentication and reputation, and it still matters. Link shorteners, links to domains with poor reputations, a single large image with no text, and attachments that commonly carry malware all raise suspicion. Every link in a message is judged by its domain, so a tracking domain shared with spammers hurts every message that uses it.&lt;/p&gt;

&lt;h2&gt;Bounces tell you exactly what happened&lt;/h2&gt;
&lt;p&gt;A bounce message carries a status code and usually a sentence from the receiving server. A &lt;code&gt;550 5.7.26&lt;/code&gt; from Gmail means the message failed authentication. Codes starting with 4 are temporary and the sending server retries. Codes starting with 5 are permanent. Search the exact code and text; the major providers document what each one means.&lt;/p&gt;

&lt;h2&gt;Testing&lt;/h2&gt;
&lt;p&gt;Services such as mail-tester.com score a single message against authentication, blocklists, and content rules. Seed tests, which send to test mailboxes at each major provider, show whether a campaign lands in the inbox, the promotions tab, or spam. Run one after every change to DNS or sending services.&lt;/p&gt;

&lt;p class="related"&gt;I trace delivery problems and fix the authentication and DNS behind them as part of &lt;a href="https://gerriscorp.com/services/email-and-dns/"&gt;email authentication and DNS work&lt;/a&gt;. If a domain is already listed, the &lt;a href="https://gerriscorp.com/guides/email-blocklist-recovery/"&gt;blocklist recovery guide&lt;/a&gt; covers what comes next.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/a/answer/81126"&gt;Google Workspace Admin Help: Email sender guidelines&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://senders.yahooinc.com/best-practices/"&gt;Yahoo Sender Hub: Best practices&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Core Web Vitals explained: LCP, INP, CLS, and the fixes that work</title>
    <link>https://gerriscorp.com/guides/core-web-vitals/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/core-web-vitals/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>What Google measures with Largest Contentful Paint, Interaction to Next Paint, and Cumulative Layout Shift, how to read field data, and the fixes for each one.</description>
    <content:encoded>&lt;h1&gt;Core Web Vitals explained: LCP, INP, CLS, and the fixes that work&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Core Web Vitals are three measurements of how a page feels to real visitors: how fast the main content appears, how quickly the page responds, and how much it jumps around. Google uses them in ranking as part of page experience. They matter more to visitors than to rankings, and that alone makes them worth fixing.&lt;/p&gt;

&lt;h2&gt;The three measurements&lt;/h2&gt;
&lt;table&gt;
&lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Metric&lt;/th&gt;&lt;th scope="col"&gt;What it measures&lt;/th&gt;&lt;th scope="col"&gt;Good&lt;/th&gt;&lt;th scope="col"&gt;Poor&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;&lt;td&gt;LCP, Largest Contentful Paint&lt;/td&gt;&lt;td&gt;Time until the biggest image or text block in view finishes rendering&lt;/td&gt;&lt;td&gt;2.5 seconds or less&lt;/td&gt;&lt;td&gt;Over 4 seconds&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;INP, Interaction to Next Paint&lt;/td&gt;&lt;td&gt;Delay between a tap, click, or key press and the screen visibly responding, across the whole visit&lt;/td&gt;&lt;td&gt;200 milliseconds or less&lt;/td&gt;&lt;td&gt;Over 500 milliseconds&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;CLS, Cumulative Layout Shift&lt;/td&gt;&lt;td&gt;How much visible content moves unexpectedly while the page is in use&lt;/td&gt;&lt;td&gt;0.1 or less&lt;/td&gt;&lt;td&gt;Over 0.25&lt;/td&gt;&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;INP replaced First Input Delay in March 2024. FID only measured the first interaction; INP watches every interaction and reports one of the slowest, which is why many sites that passed under FID fail under INP.&lt;/p&gt;

&lt;h2&gt;Field data and lab data&lt;/h2&gt;
&lt;p&gt;Google's assessment uses field data: measurements from real Chrome users, collected in the Chrome User Experience Report over a rolling 28 days. A page passes when the 75th percentile of visits meets the good threshold for all three metrics. Mobile and desktop are judged separately.&lt;/p&gt;
&lt;p&gt;Lab data comes from a single simulated load, such as a Lighthouse run. It's useful for debugging and for testing a fix before it ships, and it can't measure INP at all, because nobody clicks anything during a lab test. Lighthouse reports Total Blocking Time as a stand-in. A perfect Lighthouse score with failing field data is common, and the field data is the one that counts.&lt;/p&gt;

&lt;h2&gt;Where to read it&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;PageSpeed Insights&lt;/strong&gt; shows field data for the URL and for the whole origin at the top, and a Lighthouse lab run below it.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Search Console's Core Web Vitals report&lt;/strong&gt; groups URLs with similar problems, usually by template, and tracks them over time. Fixing one template often fixes thousands of URLs.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Chrome DevTools&lt;/strong&gt;, in the Performance panel, records a real session on your own machine, shows each interaction's timing, and highlights layout shifts.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Real user monitoring&lt;/strong&gt;, through Google's web-vitals JavaScript library or a commercial tool, collects the metrics from your own visitors, with the element and script responsible attached.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Fixing LCP&lt;/h2&gt;
&lt;p&gt;Break LCP into its parts: time to first byte, the delay before the browser discovers the LCP resource, the time to download it, and the time to render it. Fix whichever part is largest.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Slow server response.&lt;/strong&gt; Cache HTML at a CDN, speed up database queries, and remove redirects in front of the page.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Late discovery.&lt;/strong&gt; Put the hero image in the HTML as an ordinary &lt;code&gt;img&lt;/code&gt; tag with &lt;code&gt;fetchpriority="high"&lt;/code&gt;. A background image set in CSS, or an image inserted by JavaScript, is found late.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Lazy loading the hero.&lt;/strong&gt; &lt;code&gt;loading="lazy"&lt;/code&gt; belongs on images below the fold. On the LCP image it delays the most important download on the page.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Heavy files.&lt;/strong&gt; Serve images at the size they're displayed, in WebP or AVIF, with &lt;code&gt;srcset&lt;/code&gt; for different screens.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Render blocking.&lt;/strong&gt; Large stylesheets and synchronous scripts in the head hold back the first paint. Inline the critical CSS and defer the rest.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Fixing INP&lt;/h2&gt;
&lt;p&gt;Poor INP almost always means the browser's main thread was busy when the visitor interacted. The culprits are long JavaScript tasks.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Third-party scripts.&lt;/strong&gt; Chat widgets, heatmaps, tag managers full of old tags, and ad scripts. Audit them, remove the ones nobody uses, and load the rest after the page is interactive.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Heavy event handlers.&lt;/strong&gt; A click that triggers a large re-render or a synchronous calculation blocks the next paint. Show visual feedback first, then do the work, and break long tasks up so the browser can paint in between.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Hydration.&lt;/strong&gt; Frameworks such as React and Vue attach their behavior after load, and interactions during that window wait. Smaller bundles, partial hydration, and server components shorten it.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Fixing CLS&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Give every image and video &lt;code&gt;width&lt;/code&gt; and &lt;code&gt;height&lt;/code&gt; attributes, or a CSS &lt;code&gt;aspect-ratio&lt;/code&gt;, so the browser reserves the space.&lt;/li&gt;
  &lt;li&gt;Reserve fixed space for ads, embeds, and cookie banners, or show banners as overlays that cover content instead of pushing it.&lt;/li&gt;
  &lt;li&gt;Load web fonts with &lt;code&gt;font-display: swap&lt;/code&gt; and a fallback font sized to match, or use system fonts, as this site does.&lt;/li&gt;
  &lt;li&gt;Never insert content above what the visitor is already reading, unless they asked for it.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;How much it matters&lt;/h2&gt;
&lt;p&gt;Page experience is one ranking signal among hundreds, and relevance wins. A slow page with the best answer still outranks a fast page with a weak one. The business case is visitors: faster pages convert better, and slow mobile pages lose people before they ever see the offer. I treat Core Web Vitals as a quality bar every template should clear, and I fix the templates in order of the traffic and revenue they carry.&lt;/p&gt;

&lt;p class="related"&gt;Speed work is its own service: &lt;a href="https://gerriscorp.com/services/site-speed/"&gt;site speed, Core Web Vitals, and Cloudflare&lt;/a&gt;. For the CDN side, see &lt;a href="https://gerriscorp.com/guides/cloudflare-seo/"&gt;Cloudflare settings for SEO&lt;/a&gt;, and for diagnosis with an AI assistant, &lt;a href="https://gerriscorp.com/guides/core-web-vitals-mcp/"&gt;Core Web Vitals with MCP&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/core-web-vitals"&gt;Google Search Central: Understanding Core Web Vitals and Google search results&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://web.dev/articles/vitals"&gt;web.dev: Web Vitals&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>Canonical tags: how Google chooses which URL to index</title>
    <link>https://gerriscorp.com/guides/canonical-tags/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/canonical-tags/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>How rel canonical works, why Google sometimes overrides it, the Search Console statuses that reveal canonical problems, and the mistakes that misdirect signals.</description>
    <content:encoded>&lt;h1&gt;Canonical tags: how Google chooses which URL to index&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Most sites publish the same content at more than one address: with and without tracking parameters, sorted and filtered, printed, paginated, or copied across domains. Google picks one address from each group to show in results. The canonical tag is how you tell Google which one you prefer.&lt;/p&gt;

&lt;h2&gt;What the tag does&lt;/h2&gt;
&lt;p&gt;A canonical tag is a line in the page's head:&lt;/p&gt;
&lt;pre&gt;&amp;lt;link rel="canonical" href="https://example.com/shoes/trail-runner/"&amp;gt;&lt;/pre&gt;
&lt;p&gt;It says: this page is a copy, or the original, of the content at that address, and that address is the one to index. Google consolidates ranking signals such as links onto the chosen URL and shows it in results. The duplicates stay reachable for visitors.&lt;/p&gt;
&lt;p&gt;For files without a head, such as PDFs, the same instruction can go in an HTTP header: &lt;code&gt;Link: &amp;lt;https://example.com/guide.pdf&amp;gt;; rel="canonical"&lt;/code&gt;.&lt;/p&gt;

&lt;h2&gt;A hint, weighed against other signals&lt;/h2&gt;
&lt;p&gt;Google treats the canonical tag as a strong hint. It weighs the tag together with redirects, internal links, the URLs listed in the XML sitemap, hreflang annotations, and a preference for HTTPS and for shorter, cleaner URLs. When the signals agree, Google follows the tag almost every time. When they disagree, Google decides for itself.&lt;/p&gt;
&lt;p&gt;A typical conflict: every product page declares a clean canonical URL, while the navigation, the sitemap, and the breadcrumbs all link to a version with a parameter. Three signals point one way and one points the other, and Google may side with the majority.&lt;/p&gt;

&lt;h2&gt;The Search Console statuses&lt;/h2&gt;
&lt;p&gt;The Page indexing report shows canonical decisions in three statuses.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Alternate page with proper canonical tag.&lt;/strong&gt; The page points elsewhere, and Google agreed. This is working as intended.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Duplicate without user-selected canonical.&lt;/strong&gt; Google found duplicates with no canonical tag and picked one itself. Add tags so the choice is yours.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Duplicate, Google chose different canonical than user.&lt;/strong&gt; You named one URL and Google picked another. Look for conflicting signals, or for pages that are more different than you think.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;URL Inspection shows both the user-declared canonical and the Google-selected canonical for any page, which settles individual cases.&lt;/p&gt;

&lt;h2&gt;The rules I apply&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Every indexable page declares itself.&lt;/strong&gt; A self-referencing canonical guards against parameter copies that other sites and ad platforms create.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Absolute URLs.&lt;/strong&gt; Include the protocol and host. Relative canonicals are valid, and they break the moment a page is served from a staging host or a second domain.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;The target returns a 200.&lt;/strong&gt; A canonical pointing at a redirect, a 404, or a noindexed page sends contradictory instructions.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;One tag per page, in the head.&lt;/strong&gt; Two different canonicals cancel each other, and Google ignores canonical tags in the body. Plugins and themes that each add their own tag are the usual cause.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;In the server's HTML.&lt;/strong&gt; A canonical inserted or changed by JavaScript may be read late or not at all. If the raw HTML and the rendered HTML disagree, fix the template.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Internal links match.&lt;/strong&gt; Navigation, breadcrumbs, and sitemaps link to the canonical version.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Common mistakes&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Pagination pointing to page one.&lt;/strong&gt; Page two of a category holds different products from page one. Each paginated page should declare itself. Google stopped using rel prev and next as an indexing signal in 2019, so plain crawlable links between the pages do that job.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Canonical and noindex together.&lt;/strong&gt; One says "index that other page instead," the other says "index nothing." Pick one.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Every page pointing at the home page.&lt;/strong&gt; A template bug that copies the home URL into every canonical can remove most of a site from results within weeks.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Canonicals to the HTTP version, or to the old domain.&lt;/strong&gt; Leftovers from a migration that nobody updated.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Product variants.&lt;/strong&gt; Color and size variants that differ only by a parameter usually canonicalize to the main product. Variants that people search for by name, such as a specific model, may deserve their own indexable pages.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Across domains&lt;/h2&gt;
&lt;p&gt;A canonical can point to another domain, which is how a business that runs two sites with overlapping content can concentrate signals on one. For syndicated articles republished by partners, Google's current advice is that the partner should block indexing of its copy with noindex, because a cross-domain canonical on a large partner site doesn't always win against that site's own authority.&lt;/p&gt;

&lt;h2&gt;Canonical or redirect&lt;/h2&gt;
&lt;p&gt;When visitors never need the duplicate URL, redirect it. A 301 is a stronger signal, and nobody lands on the copy. Use a canonical when the duplicate has to keep working, such as a filtered view, a tracking link, or a print version.&lt;/p&gt;

&lt;p class="related"&gt;Canonical audits are part of &lt;a href="https://gerriscorp.com/services/technical-seo/"&gt;technical SEO forensics&lt;/a&gt; and &lt;a href="https://gerriscorp.com/services/large-sites/"&gt;large-site SEO&lt;/a&gt;. For the related indexing status, see &lt;a href="https://gerriscorp.com/guides/crawled-not-indexed/"&gt;Crawled, currently not indexed&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/canonicalization"&gt;Google Search Central: What is canonicalization&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls"&gt;Google Search Central: How to specify a canonical URL&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>ADA website compliance for businesses: lower risk, better SEO</title>
    <link>https://gerriscorp.com/guides/ada-website-compliance/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/ada-website-compliance/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>What the ADA expects of business websites, the six fixes that cover most of the risk, what to skip, and why the same work improves search, AI answers, trust.</description>
    <content:encoded>&lt;h1&gt;ADA website compliance for businesses: lower risk, better SEO&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Making a website accessible is the right thing to do, and it's also cheap insurance. Most of the legal exposure comes from a short list of common, fixable failures. Fix those, document what you did, and you'll also get a faster, clearer, better-ranking site that AI assistants understand. No badges, no widgets, and no press release required.&lt;/p&gt;
&lt;p&gt;A note before the details: this is a practitioner's guide, not legal advice. If you've received a demand letter, call a lawyer first.&lt;/p&gt;

&lt;h2&gt;What the law expects&lt;/h2&gt;
&lt;p&gt;The Americans with Disabilities Act covers businesses open to the public, and the Department of Justice's position, set out in its guidance on ADA.gov, is that this includes their websites. For private businesses there's no technical standard written into the regulations, so courts, settlements, and auditors almost always measure sites against the Web Content Accessibility Guidelines, WCAG, at level AA.&lt;/p&gt;
&lt;p&gt;For state and local governments, the rules are firmer. A 2024 Justice Department rule requires WCAG 2.1 level AA. An interim rule in April 2026 extended the deadlines: April 26, 2027, for governments serving 50,000 people or more, and April 26, 2028, for smaller ones and special districts. If you build or run sites for public agencies, those dates belong on your calendar.&lt;/p&gt;

&lt;h2&gt;Where the risk actually comes from&lt;/h2&gt;
&lt;p&gt;Demand letters and lawsuits tend to cite problems anyone can detect with a free automated scan, because that's how many of them are found. WebAIM's 2026 survey of the top million home pages found detectable failures on 95.9% of them, and six kinds of errors accounted for 96% of everything detected:&lt;/p&gt;
&lt;table&gt;
&lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Failure&lt;/th&gt;&lt;th scope="col"&gt;Share of home pages&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;&lt;td&gt;Low contrast text&lt;/td&gt;&lt;td&gt;83.9%&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Missing image alt text&lt;/td&gt;&lt;td&gt;53.1%&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Form fields without labels&lt;/td&gt;&lt;td&gt;51%&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Empty links&lt;/td&gt;&lt;td&gt;46.3%&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Empty buttons&lt;/td&gt;&lt;td&gt;30.6%&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Missing page language&lt;/td&gt;&lt;td&gt;13.5%&lt;/td&gt;&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;Fixing those six puts a site ahead of almost all of the web. That's the 80 to 90 percent that matters most, and it's usually days of work, not months.&lt;/p&gt;

&lt;h2&gt;The fix list, in order&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Contrast.&lt;/strong&gt; Darken pale text and light brand colors until normal text reaches 4.5 to 1 against its background. This is usually a handful of color values in one stylesheet.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Alt text.&lt;/strong&gt; Describe every meaningful image; give decorative ones an empty alt attribute.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Form labels.&lt;/strong&gt; Every field gets a visible label tied to it. Placeholder text alone doesn't count.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Links and buttons that say something.&lt;/strong&gt; Icon-only buttons need an accessible name; a link wrapped around nothing but an icon needs one too.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Page language.&lt;/strong&gt; One attribute on the html element.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Keyboard and focus.&lt;/strong&gt; Tab through every page template; everything should be reachable, in order, with a visible focus outline.&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;Then cover the rest of WCAG AA over time, starting with headings, captions for video, and error messages on forms.&lt;/p&gt;

&lt;h2&gt;What to skip&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Overlay widgets.&lt;/strong&gt; Scripts that promise one-line compliance don't fix the underlying markup, many users of assistive technology find them obstructive, and sites using them have still been sued.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Badges and certificates.&lt;/strong&gt; A "fully compliant" seal invites scrutiny and proves nothing. The work speaks for itself.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Grand announcements.&lt;/strong&gt; A short accessibility statement is useful; a marketing campaign about it isn't.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Document it&lt;/h2&gt;
&lt;p&gt;Keep a simple record: which pages were checked, with which tools, what was fixed, and when. Publish a brief accessibility statement with a way to report problems, and answer those reports promptly. A business that can show steady, good-faith effort is in a far better position than one that can't.&lt;/p&gt;

&lt;h2&gt;The dividends you collect anyway&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Search.&lt;/strong&gt; Alt text feeds image search, real headings and descriptive links are exactly what Google uses to understand a page, and the page language helps search engines serve the right audience.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;AI answers and AI agents.&lt;/strong&gt; Assistants quote clean, well-structured text, and the browsing agents now arriving use a page's accessibility information to find and operate buttons and forms. Lighthouse even added an Agentic Browsing check.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Trust.&lt;/strong&gt; A site that works for everyone reads as careful and current, which is what quality raters, and customers, look for.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Conversions.&lt;/strong&gt; Readable text, labeled forms, and usable buttons help every visitor, including the ones on a phone in bright sunlight.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Keep it from sliding back&lt;/h2&gt;
&lt;p&gt;Accessibility erodes with every new page, plugin, and campaign. Add a quick automated scan and a keyboard pass to your launch routine, check new templates before they ship, and review the site once a quarter. My guide to &lt;a href="https://gerriscorp.com/guides/lynx-qa/"&gt;Lynx for website QA&lt;/a&gt; covers a text-browser pass that catches a surprising number of these problems in minutes.&lt;/p&gt;
&lt;p&gt;For help auditing and fixing a site, see &lt;a href="https://gerriscorp.com/services/technical-seo/"&gt;technical SEO&lt;/a&gt; or &lt;a href="https://gerriscorp.com/contact/"&gt;get in touch&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://www.ada.gov/resources/web-guidance/"&gt;ADA.gov: Guidance on web accessibility and the ADA&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.ada.gov/resources/2024-03-08-web-rule/"&gt;ADA.gov: Fact sheet on the Title II web rule&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.w3.org/TR/WCAG22/"&gt;W3C: Web Content Accessibility Guidelines (WCAG) 2.2&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://webaim.org/projects/million/"&gt;WebAIM: The WebAIM Million, 2026 report&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://overlayfactsheet.com/"&gt;Overlay Fact Sheet&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to plan 301 redirects that keep your rankings and links</title>
    <link>https://gerriscorp.com/guides/301-redirects/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/301-redirects/</guid>
    <pubDate>Tue, 06 Oct 2026 12:00:00 +0000</pubDate>
    <description>Which redirect to use, how to map old URLs to new ones, where to implement redirects on common platforms, how to test them, and how long to keep them live.</description>
    <content:encoded>&lt;h1&gt;How to plan 301 redirects that keep your rankings and links&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-06"&gt;October 6, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A redirect tells browsers and search engines that a page has moved. Done well, it carries the old page's links, bookmarks, and search history to the new address. Done badly, it quietly throws all of that away. Here's how I plan them.&lt;/p&gt;

&lt;h2&gt;Which redirect to use&lt;/h2&gt;
&lt;table&gt;
&lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Type&lt;/th&gt;&lt;th scope="col"&gt;Meaning&lt;/th&gt;&lt;th scope="col"&gt;Use it for&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;&lt;td&gt;301&lt;/td&gt;&lt;td&gt;Moved permanently&lt;/td&gt;&lt;td&gt;Almost every move: renamed pages, merged pages, domain changes, HTTP to HTTPS&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;308&lt;/td&gt;&lt;td&gt;Moved permanently, and the request method stays the same&lt;/td&gt;&lt;td&gt;The same jobs as a 301; some servers and APIs prefer it&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;302 or 307&lt;/td&gt;&lt;td&gt;Moved temporarily&lt;/td&gt;&lt;td&gt;A page that will come back: a short promotion, maintenance, a geographic or login handoff&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Meta refresh at zero seconds&lt;/td&gt;&lt;td&gt;An HTML instruction to load another page&lt;/td&gt;&lt;td&gt;Hosts that can't send server redirects, such as GitHub Pages&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;JavaScript redirect&lt;/td&gt;&lt;td&gt;A script that changes the location&lt;/td&gt;&lt;td&gt;A last resort; it only works once the page renders&lt;/td&gt;&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;Google treats 301 and 308 as strong signals that the new URL should replace the old one in its index. A temporary redirect is a weaker signal, though Google may eventually treat a 302 that never goes away as permanent. An instant meta refresh is read as a permanent move too, which is how this site handles the hundreds of old gerriscorp.com addresses it inherited: GitHub Pages can't send a real 301, so each old path holds a tiny page that forwards immediately and names the new page as its canonical.&lt;/p&gt;

&lt;h2&gt;Map every old URL to its closest equivalent&lt;/h2&gt;
&lt;p&gt;The redirect map is a two-column list: old URL, new URL. The work is in the second column. Each old page goes to the page that answers the same question its visitors arrived with.&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;One to one when you can.&lt;/strong&gt; A product, article, or service that still exists goes to its new address.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Merged pages go to the survivor.&lt;/strong&gt; Three thin pages folded into one strong page all point at it.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Retired pages go to the nearest parent.&lt;/strong&gt; A discontinued product goes to its category, and a retired service goes to the services overview.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Junk can stay gone.&lt;/strong&gt; Spam URLs, old tag archives, and test pages with no links and no traffic can return a 404 or 410.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Send old pages to relevant pages, and keep the home page for the home page. Google treats a mass redirect to an unrelated page as a soft 404, so the old page's signals are dropped anyway, and visitors land somewhere that doesn't answer what they came for.&lt;/p&gt;

&lt;h2&gt;Where the old URLs come from&lt;/h2&gt;
&lt;p&gt;Build the list from every source you have, because no single one is complete. Crawl the current site with a tool such as Screaming Frog. Export pages from Search Console's Performance report and from analytics landing pages. Pull linked pages from a backlink tool. For an old site that's already gone, the Internet Archive's Wayback Machine keeps a record of URLs it captured, which is how I recovered more than six hundred old paths for this domain.&lt;/p&gt;

&lt;h2&gt;Where to implement them&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;At the edge.&lt;/strong&gt; Cloudflare's Redirect Rules and Bulk Redirects answer before the request ever reaches your server, which is fast and survives a change of host or platform.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;On the web server.&lt;/strong&gt; Apache uses &lt;code&gt;Redirect 301&lt;/code&gt; or &lt;code&gt;RewriteRule&lt;/code&gt; in the configuration or &lt;code&gt;.htaccess&lt;/code&gt;; Nginx uses &lt;code&gt;return 301&lt;/code&gt; inside a &lt;code&gt;location&lt;/code&gt; block.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;In the platform.&lt;/strong&gt; WordPress handles them with a plugin such as Redirection or the redirect manager in Rank Math. Shopify has a URL redirects screen in the admin, and it only applies a redirect when the old path would otherwise return a 404. Squarespace, Wix, and Webflow each have a redirects panel in their settings.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;In the framework.&lt;/strong&gt; Next.js, Netlify, and Vercel take redirects in a configuration file that ships with the code, which keeps them under version control.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Pick one layer and keep all the redirects there. Rules split across Cloudflare, the server, and a plugin are how loops and chains happen.&lt;/p&gt;

&lt;h2&gt;Chains and loops&lt;/h2&gt;
&lt;p&gt;A chain is a redirect that lands on another redirect: A goes to B, and B goes to C. Googlebot follows up to ten hops, but each hop costs time for visitors and makes the signal weaker. When a site has been redesigned more than once, old redirects often point at URLs that are themselves redirected now. Update the old rules so every one points straight at the final address.&lt;/p&gt;
&lt;p&gt;A loop is A to B and B back to A. Browsers give up with an error. Loops usually come from two layers fighting, such as an HTTPS rule in Cloudflare and a different HTTPS rule on the server.&lt;/p&gt;

&lt;h2&gt;How to test them&lt;/h2&gt;
&lt;p&gt;From a terminal, &lt;code&gt;curl -sIL https://example.com/old-page&lt;/code&gt; prints every hop with its status code, so you can see a 301 followed by a 200 and nothing in between. For the full map, feed the old URL list to a crawler in list mode and check three things for each row: the first response is a 301 or 308, the final response is a 200, and the final URL matches the target in the map. Then spot check the same URLs in a private browser window, since your own browser caches permanent redirects aggressively.&lt;/p&gt;

&lt;h2&gt;How long to keep them&lt;/h2&gt;
&lt;p&gt;Google's advice is at least a year. I keep them indefinitely. Backlinks from other sites never get updated, and a redirect costs almost nothing to keep. Removing redirects after a year is how a business loses links it earned a decade ago.&lt;/p&gt;

&lt;section class="faq"&gt;
&lt;h2&gt;Questions about redirects&lt;/h2&gt;
&lt;h3&gt;Do 301 redirects lose PageRank?&lt;/h3&gt;
&lt;p&gt;No. Google has said for years that 301, 302, and 308 redirects pass link value without a penalty. The losses people see come from redirecting to the wrong page, from chains, or from removing redirects too soon.&lt;/p&gt;
&lt;h3&gt;Should I redirect a page I deleted on purpose?&lt;/h3&gt;
&lt;p&gt;Only if something on the site now answers the same need. Otherwise a 410 tells search engines the removal was deliberate, and they drop the URL a little faster than with a 404.&lt;/p&gt;
&lt;h3&gt;How many redirects is too many?&lt;/h3&gt;
&lt;p&gt;There's no practical limit on the number of rules. Tens of thousands are normal on large sites. What matters is that each one is a single hop to a relevant page.&lt;/p&gt;
&lt;/section&gt;

&lt;p class="related"&gt;Redirect mapping is part of every &lt;a href="https://gerriscorp.com/services/migrations/"&gt;website migration&lt;/a&gt; I run. For the whole process, see the &lt;a href="https://gerriscorp.com/guides/migration-checklist/"&gt;migration checklist&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/301-redirects"&gt;Google Search Central: Redirects and Google Search&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/site-move-with-url-changes"&gt;Google Search Central: Site moves with URL changes&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to remove your personal information from data brokers and search</title>
    <link>https://gerriscorp.com/guides/personal-information-removal/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/personal-information-removal/</guid>
    <pubDate>Mon, 05 Oct 2026 12:00:00 +0000</pubDate>
    <description>How data brokers and people search sites get your information, how to opt out, how Google Results about you works, and what California's DROP changes.</description>
    <content:encoded>&lt;h1&gt;How to remove your personal information from data brokers and search&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-05"&gt;October 5, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Your home address, phone numbers, relatives, and age are probably for sale on dozens of people-search sites. Removing them is tedious and never permanent, and it's still worth doing, especially for executives, public figures, and anyone with a reason to be careful.&lt;/p&gt;

&lt;h2&gt;Where the information comes from&lt;/h2&gt;
&lt;p&gt;Data brokers compile profiles from public records such as property deeds, voter files, court records, and business filings, combined with marketing data, social media, and other brokers. People-search sites such as Spokeo, Whitepages, BeenVerified, Radaris, and many others sell access to those profiles, and search engines index the result pages, which is how your address ends up in Google results for your name.&lt;/p&gt;

&lt;h2&gt;Opting out, site by site&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Most people-search sites have an opt-out form, usually linked from the footer as "Do not sell," "Opt out," or "Privacy."&lt;/li&gt;
  &lt;li&gt;Search each site for every version of your name, past and current addresses, and relatives' listings that include you.&lt;/li&gt;
  &lt;li&gt;Use a dedicated email address for opt-out requests, since many sites require confirmation by email.&lt;/li&gt;
  &lt;li&gt;Keep a log of each site, the date, and the confirmation. Listings reappear when brokers refresh their data, so the work repeats every few months.&lt;/li&gt;
  &lt;li&gt;Commercial removal services automate much of this for a subscription. They cover the major sites well and the long tail less well.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Google's "Results about you"&lt;/h2&gt;
&lt;p&gt;Google's Results about you tool lets you request removal of search results that show your phone number, email address, or home address, and it can alert you when new results with your contact information appear. Removal from Google doesn't delete the page from the website itself, so pair it with an opt-out at the source. Google also has separate removal policies for doxxing, non-consensual explicit images, and certain other harmful content.&lt;/p&gt;

&lt;h2&gt;Your legal rights&lt;/h2&gt;
&lt;h3&gt;California's Delete Act and DROP&lt;/h3&gt;
&lt;p&gt;California's Delete Act created the Delete Request and Opt-out Platform, known as DROP, run by the California Privacy Protection Agency. Since January 2026, California residents can submit one request that goes to every registered data broker, and brokers must begin processing those requests in August 2026 and keep checking for new ones on a regular schedule. It's the most significant change to data broker removal in years.&lt;/p&gt;
&lt;h3&gt;Other states&lt;/h3&gt;
&lt;p&gt;California, Vermont, Texas, and Oregon require data brokers to register, which makes the industry easier to identify. A growing number of state privacy laws give residents a right to delete personal information held by businesses, with exceptions, so check what your state provides.&lt;/p&gt;
&lt;h3&gt;Outside the US&lt;/h3&gt;
&lt;p&gt;In the European Union and the United Kingdom, data protection law gives a right to erasure, and search engines handle delisting requests for results about you that are inaccurate, inadequate, or no longer relevant.&lt;/p&gt;

&lt;h2&gt;What removal can't do&lt;/h2&gt;
&lt;p&gt;Public records themselves usually stay public, though some states let certain people, such as judges, law enforcement officers, and domestic violence survivors, shield their addresses. Legitimate news coverage is rarely removable. Where removal isn't available, the practical answer is a stronger, accurate public record that ranks above what you can't remove.&lt;/p&gt;

&lt;h2&gt;Protecting the people around you&lt;/h2&gt;
&lt;p&gt;Broker profiles link relatives, so a spouse's or parent's listing can expose your address even after your own is removed. For executives and public figures, the work should cover the household, old addresses, and any properties held in a personal name. Consider holding new property in a trust or entity where that's appropriate, and use a mailing address that isn't your home for registrations and subscriptions.&lt;/p&gt;

&lt;h2&gt;Keeping it clean&lt;/h2&gt;
&lt;p&gt;Brokers refresh their data from public records and other brokers constantly, so removed listings come back. A quarterly sweep across all the major sites catches relistings early, and monitoring tools, including Results about you, flag new appearances between sweeps.&lt;/p&gt;

&lt;h2&gt;A sensible routine&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;Search your name, with and without your city, in Google and Bing, and on the largest people-search sites.&lt;/li&gt;
  &lt;li&gt;Opt out of every listing you find, and log each request.&lt;/li&gt;
  &lt;li&gt;Set up Results about you, and use DROP if you're a California resident.&lt;/li&gt;
  &lt;li&gt;Repeat every three months.&lt;/li&gt;
  &lt;li&gt;For families and executives, cover spouses, children, and older addresses as well.&lt;/li&gt;
&lt;/ol&gt;

&lt;p class="related"&gt;I handle this for clients as part of &lt;a href="https://gerriscorp.com/services/reputation/"&gt;reputation and perception management&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://consumer.ftc.gov/articles/what-know-about-people-search-sites-sell-your-information"&gt;FTC Consumer Advice: People search sites that sell your information&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://cppa.ca.gov/data_broker_registry/"&gt;California Privacy Protection Agency: Data broker registry&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>The website migration SEO checklist</title>
    <link>https://gerriscorp.com/guides/migration-checklist/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/migration-checklist/</guid>
    <pubDate>Mon, 05 Oct 2026 12:00:00 +0000</pubDate>
    <description>A step-by-step website migration SEO checklist: URL inventory, 301 redirect map, template review, launch day checks, and the weeks of monitoring after.</description>
    <content:encoded>&lt;h1&gt;The website migration SEO checklist&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-05"&gt;October 5, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Migrations lose traffic for predictable reasons: URLs that weren't mapped, redirects that weren't tested, templates that changed what crawlers see, and nobody watching afterward. This checklist covers each stage.&lt;/p&gt;

&lt;h2&gt;Before you start&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Record a baseline: organic sessions and conversions by landing page from analytics, and clicks, impressions, and indexed pages from Search Console. Export them, because you'll compare against them for months.&lt;/li&gt;
  &lt;li&gt;Decide the launch window. Avoid your busiest season.&lt;/li&gt;
  &lt;li&gt;Make sure Search Console and Bing Webmaster Tools are verified for both the old and new sites, and that you'll keep access to the old domain if it's changing.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Build the URL inventory&lt;/h2&gt;
&lt;p&gt;Combine every source, because each one misses something:&lt;/p&gt;
&lt;ul&gt;
  &lt;li&gt;A full crawl of the current site.&lt;/li&gt;
  &lt;li&gt;Every URL in the XML sitemaps.&lt;/li&gt;
  &lt;li&gt;Pages with clicks or impressions in Search Console over the last 16 months.&lt;/li&gt;
  &lt;li&gt;Landing pages with traffic in analytics.&lt;/li&gt;
  &lt;li&gt;URLs with external backlinks, from Ahrefs, Semrush, or Search Console's Links report.&lt;/li&gt;
  &lt;li&gt;Images, PDFs, and other files that rank or receive links.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Build the redirect map&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Map every old URL to its closest equivalent on the new site with a 301 or 308 permanent redirect.&lt;/li&gt;
  &lt;li&gt;Redirect to the most relevant page, never to the home page in bulk. Google treats mass redirects to the home page like soft 404s.&lt;/li&gt;
  &lt;li&gt;Make deliberate decisions about pages you're retiring: redirect them to a close substitute, or let them return 404 or 410 if nothing fits.&lt;/li&gt;
  &lt;li&gt;Avoid chains. Every old URL should reach its final destination in one hop, including URLs that were already redirecting.&lt;/li&gt;
  &lt;li&gt;Keep the map in a spreadsheet: old URL, new URL, status, reason. It's the document every later question goes back to.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Review the new site before launch&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Crawl staging and compare titles, descriptions, headings, canonicals, and word counts with the old site, template by template.&lt;/li&gt;
  &lt;li&gt;Check that the main content is in the server's HTML, so it doesn't depend on JavaScript.&lt;/li&gt;
  &lt;li&gt;Check structured data, internal links, image alt text, and hreflang if you have more than one language.&lt;/li&gt;
  &lt;li&gt;Test Core Web Vitals on the new templates.&lt;/li&gt;
  &lt;li&gt;Keep staging blocked from search with authentication or noindex, and write down exactly how, so it can be removed at launch.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Launch day&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Remove the staging block, and confirm production robots.txt and meta robots tags allow crawling.&lt;/li&gt;
  &lt;li&gt;Test the redirect map in bulk against the live site: status codes, single hops, correct destinations.&lt;/li&gt;
  &lt;li&gt;Submit the new XML sitemaps. Keep a sitemap of the old URLs available for a few weeks so Google recrawls them and discovers the redirects faster.&lt;/li&gt;
  &lt;li&gt;For a domain change, use Google's Change of Address tool in Search Console and Bing's Site Move tool.&lt;/li&gt;
  &lt;li&gt;Confirm analytics and tag tracking work on every template.&lt;/li&gt;
  &lt;li&gt;If DNS changes, confirm that mail records were left untouched and email still arrives.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;The weeks after&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Watch Search Console daily for the first two weeks: page indexing, crawl stats, and 404s.&lt;/li&gt;
  &lt;li&gt;Compare organic landing pages with the baseline, and investigate every important page that lost traffic.&lt;/li&gt;
  &lt;li&gt;Fix redirect gaps as they appear in 404 reports and server logs.&lt;/li&gt;
  &lt;li&gt;Update internal links that still point at old URLs.&lt;/li&gt;
  &lt;li&gt;Ask the most valuable sites linking to you to update their links.&lt;/li&gt;
  &lt;li&gt;Keep the redirects in place for at least a year, and ideally indefinitely.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Special cases&lt;/h2&gt;
&lt;h3&gt;HTTP to HTTPS and www changes&lt;/h3&gt;
&lt;p&gt;Pick one canonical host and protocol, redirect every other combination to it in a single hop, and update canonicals, sitemaps, and internal links to match.&lt;/p&gt;
&lt;h3&gt;Merging several sites into one&lt;/h3&gt;
&lt;p&gt;Map each old site's pages to their best home on the new one, and keep every old domain registered and redirecting. Old domains carry links for years.&lt;/p&gt;
&lt;h3&gt;Moving to a JavaScript framework&lt;/h3&gt;
&lt;p&gt;Check that the new templates render their content on the server. A migration from a traditional CMS to a client-rendered app is a common way to lose content from Google and AI crawlers. See the &lt;a href="https://gerriscorp.com/guides/javascript-seo/"&gt;JavaScript SEO guide&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;What to expect&lt;/h2&gt;
&lt;p&gt;A well-run migration often shows some fluctuation for a few weeks while Google reprocesses the site, then settles. Domain changes take longer than platform changes on the same domain. A drop that deepens after the first month, or that's concentrated in one template, means something specific went wrong, and it's worth diagnosing rather than waiting out.&lt;/p&gt;

&lt;p class="related"&gt;I run this process as a &lt;a href="https://gerriscorp.com/services/migrations/"&gt;website migration SEO service&lt;/a&gt;, before launch or after it. The kit is in &lt;a href="https://gerriscorp.com/guides/migration-tools/"&gt;SEO tools for a website migration&lt;/a&gt;, and if traffic has already fallen, start with &lt;a href="https://gerriscorp.com/guides/migration-recovery/"&gt;migration recovery&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/site-move-with-url-changes"&gt;Google Search Central: Site moves with URL changes&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/301-redirects"&gt;Google Search Central: Redirects and Google Search&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How to tell if JavaScript is hiding your content from Google and AI</title>
    <link>https://gerriscorp.com/guides/javascript-seo/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/javascript-seo/</guid>
    <pubDate>Mon, 05 Oct 2026 12:00:00 +0000</pubDate>
    <description>A practical test for JavaScript SEO: compare raw and rendered HTML, check Search Console, find content that needs a scroll or click, and fix it on the server.</description>
    <content:encoded>&lt;h1&gt;How to tell if JavaScript is hiding your content from Google and AI&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-05"&gt;October 5, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;A page can look complete in a browser and nearly empty to a crawler. When the important content depends on JavaScript, search engines may see less of it, later, and AI crawlers may not see it at all.&lt;/p&gt;

&lt;h2&gt;How crawlers handle JavaScript&lt;/h2&gt;
&lt;p&gt;Googlebot renders pages with an up-to-date version of Chromium, so it can run JavaScript. Rendering happens in a queue after the initial crawl, though, and the renderer behaves like a very patient visitor who never clicks, never types, and never scrolls the way a person does. It loads the page with a tall viewport and takes a snapshot. Content that waits for a click, a real scroll event, or a user action can be missing from that snapshot.&lt;/p&gt;
&lt;p&gt;Most AI crawlers are simpler. GPTBot, ClaudeBot, PerplexityBot, and similar crawlers generally fetch the HTML the server sends and read it without running JavaScript. If your product details, prices, reviews, or answers exist only after scripts run, those crawlers see a shell.&lt;/p&gt;

&lt;h2&gt;Five tests that settle it&lt;/h2&gt;
&lt;h3&gt;1. Compare view-source with the rendered page&lt;/h3&gt;
&lt;p&gt;Open the page, then open view-source. View-source shows the HTML your server sent; the Elements panel in developer tools shows the DOM after JavaScript ran. Search view-source for a sentence from the content you care about. If it's missing, crawlers that don't render can't see it.&lt;/p&gt;
&lt;h3&gt;2. Fetch it from the command line&lt;/h3&gt;
&lt;pre&gt;&lt;code&gt;curl -sL https://example.com/products/widget-200 | grep -c "Specifications"&lt;/code&gt;&lt;/pre&gt;
&lt;p&gt;A count of zero means the text isn't in the server's HTML.&lt;/p&gt;
&lt;h3&gt;3. Disable JavaScript&lt;/h3&gt;
&lt;p&gt;Turn off JavaScript in your browser's developer tools and reload. What remains is roughly what a non-rendering crawler gets.&lt;/p&gt;
&lt;h3&gt;4. Ask Google directly&lt;/h3&gt;
&lt;p&gt;In Search Console, run URL Inspection on the page, click "Test live URL," then "View tested page" and the HTML tab. That's the HTML Google rendered. If your content is missing there, Google doesn't have it.&lt;/p&gt;
&lt;h3&gt;5. Crawl both ways&lt;/h3&gt;
&lt;p&gt;Crawl a sample of pages in Screaming Frog with JavaScript rendering off, then on, and compare word counts by template. In one diagnosis I ran, category pages yielded thousands of words while flagship product pages yielded 250 to 400 words despite weighing more than a megabyte, which pointed straight at the product template.&lt;/p&gt;

&lt;h2&gt;Common causes&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Client-side rendering.&lt;/strong&gt; Single-page apps that send an empty HTML shell and build the page in the browser.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Content that loads on scroll or interaction.&lt;/strong&gt; Sections rendered only when they enter the viewport after a real scroll, such as Alpine.js &lt;code&gt;x-if&lt;/code&gt; combined with &lt;code&gt;x-intersect&lt;/code&gt;, or reviews fetched when a visitor scrolls near them.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Content that loads on click.&lt;/strong&gt; Tabs and accordions that fetch their content when opened. Tabs that merely hide content already in the HTML are fine.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Links without real URLs.&lt;/strong&gt; Navigation built with click handlers instead of &lt;code&gt;&amp;lt;a href&amp;gt;&lt;/code&gt; links, which crawlers can't follow.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Infinite scroll without paginated URLs.&lt;/strong&gt; Products beyond the first batch never get a crawlable link.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Metadata changed by JavaScript.&lt;/strong&gt; Titles, canonical tags, or robots directives rewritten after load, so the raw HTML and the rendered page disagree.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Blocked resources.&lt;/strong&gt; JavaScript or CSS files disallowed in robots.txt, which prevents Google from rendering the page properly.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Fixes&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Render on the server.&lt;/strong&gt; Server-side rendering or static generation puts the content in the initial HTML. Next.js, Nuxt, SvelteKit, Astro, and similar frameworks do this well when pages are configured for it.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Keep content in the HTML and hide it with CSS.&lt;/strong&gt; In Alpine.js, &lt;code&gt;x-show&lt;/code&gt; keeps an element in the DOM and only toggles visibility, while &lt;code&gt;x-if&lt;/code&gt; removes it entirely until the condition is true. Use animation as progressive enhancement on content that's already present.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Use real links and paginated URLs&lt;/strong&gt; for navigation and long lists.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Put titles, canonicals, and structured data in the server's HTML&lt;/strong&gt;, and keep them unchanged after load.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Avoid dynamic rendering&lt;/strong&gt; as a long-term answer. Google describes serving crawlers a separately rendered version as a workaround, and server rendering is the durable fix.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Writing the ticket&lt;/h2&gt;
&lt;p&gt;Developers fix these problems fastest with evidence: the affected template, example URLs, the word count with and without rendering, the Search Console live test result, and a testable definition of done, such as "a curl of the product URL returns the specification table." There's a &lt;a href="https://gerriscorp.com/services/developers/"&gt;sample ticket on the developers page&lt;/a&gt;.&lt;/p&gt;

&lt;p class="related"&gt;I diagnose rendering problems as part of &lt;a href="https://gerriscorp.com/services/technical-seo/"&gt;technical SEO forensics&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics"&gt;Google Search Central: JavaScript SEO basics&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://platform.openai.com/docs/bots"&gt;OpenAI: Overview of OpenAI crawlers&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>SPF, DKIM, and DMARC for every domain that sends email</title>
    <link>https://gerriscorp.com/guides/email-authentication/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/email-authentication/</guid>
    <pubDate>Mon, 05 Oct 2026 12:00:00 +0000</pubDate>
    <description>What SPF, DKIM, and DMARC records do, example records, the Gmail, Yahoo, and Microsoft sender requirements, and how to reach DMARC enforcement safely.</description>
    <content:encoded>&lt;h1&gt;SPF, DKIM, and DMARC for every domain that sends email&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-05"&gt;October 5, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Three DNS records decide whether mail from your domain is trusted. Since 2024, the largest mailbox providers require them, and mail without them increasingly lands in spam or bounces.&lt;/p&gt;

&lt;h2&gt;The current requirements&lt;/h2&gt;
&lt;p&gt;Since February 2024, Gmail and Yahoo require every sender to authenticate mail with SPF or DKIM, to have valid forward and reverse DNS for sending servers, and to keep spam complaint rates low. Bulk senders, those sending around 5,000 or more messages a day to Gmail accounts, must also have SPF, DKIM, and a DMARC policy, align the visible From domain with SPF or DKIM, offer one-click unsubscribe on marketing mail, and keep the spam rate reported in Google Postmaster Tools below 0.3%. Since May 2025, Microsoft's Outlook.com has required SPF, DKIM, and DMARC from senders of more than 5,000 messages a day to its consumer addresses.&lt;/p&gt;
&lt;p&gt;Even small businesses are affected, because a newsletter tool, CRM, invoicing system, or help desk sending as your domain counts as a sender.&lt;/p&gt;

&lt;h2&gt;SPF: who may send&lt;/h2&gt;
&lt;p&gt;SPF (Sender Policy Framework) is a TXT record on your domain listing the servers and services allowed to send mail as you.&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;example.com.  TXT  "v=spf1 include:_spf.google.com include:sendgrid.net ~all"&lt;/code&gt;&lt;/pre&gt;
&lt;ul&gt;
  &lt;li&gt;Publish exactly one SPF record per domain. Two records break SPF.&lt;/li&gt;
  &lt;li&gt;SPF allows at most ten DNS lookups. Each &lt;code&gt;include&lt;/code&gt; counts, and nested includes count too. Too many services can push you over.&lt;/li&gt;
  &lt;li&gt;&lt;code&gt;~all&lt;/code&gt; (soft fail) is the common ending; &lt;code&gt;-all&lt;/code&gt; (fail) is stricter. With DMARC in place, either works.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;DKIM: a signature on every message&lt;/h2&gt;
&lt;p&gt;DKIM (DomainKeys Identified Mail) signs each message with a private key; receivers check the signature against a public key published in DNS under a selector.&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;google._domainkey.example.com.  TXT  "v=DKIM1; k=rsa; p=MIIBIjANBgkq..."&lt;/code&gt;&lt;/pre&gt;
&lt;ul&gt;
  &lt;li&gt;Every service that sends as your domain needs its own DKIM setup, usually a TXT or CNAME record the service provides.&lt;/li&gt;
  &lt;li&gt;Use 2048-bit keys where the service offers them.&lt;/li&gt;
  &lt;li&gt;DKIM survives forwarding better than SPF, which makes it the more reliable of the two for DMARC alignment.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;DMARC: the policy that ties them together&lt;/h2&gt;
&lt;p&gt;DMARC (Domain-based Message Authentication, Reporting, and Conformance) tells receivers what to do when a message claiming to be from your domain fails authentication, and asks them to send you reports.&lt;/p&gt;
&lt;pre&gt;&lt;code&gt;_dmarc.example.com.  TXT  "v=DMARC1; p=none; rua=mailto:dmarc@example.com"&lt;/code&gt;&lt;/pre&gt;
&lt;ul&gt;
  &lt;li&gt;A message passes DMARC when SPF or DKIM passes and the domain it passed for aligns with the visible From address.&lt;/li&gt;
  &lt;li&gt;&lt;code&gt;p=none&lt;/code&gt; monitors only, &lt;code&gt;p=quarantine&lt;/code&gt; sends failures to spam, and &lt;code&gt;p=reject&lt;/code&gt; refuses them.&lt;/li&gt;
  &lt;li&gt;&lt;code&gt;rua&lt;/code&gt; sets where daily aggregate reports go. They're XML files listing every source that sent mail as your domain and whether it passed.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Reaching enforcement safely&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;Inventory every service that sends mail as your domain.&lt;/li&gt;
  &lt;li&gt;Set up SPF and DKIM for each.&lt;/li&gt;
  &lt;li&gt;Publish DMARC at &lt;code&gt;p=none&lt;/code&gt; with reporting, and read the reports for two to four weeks, using a report viewer to make them readable.&lt;/li&gt;
  &lt;li&gt;Fix every legitimate source that fails.&lt;/li&gt;
  &lt;li&gt;Move to &lt;code&gt;p=quarantine&lt;/code&gt;, optionally with &lt;code&gt;pct&lt;/code&gt; below 100 to phase it in, and keep reading reports.&lt;/li&gt;
  &lt;li&gt;Move to &lt;code&gt;p=reject&lt;/code&gt; once failures are only spoofing.&lt;/li&gt;
&lt;/ol&gt;
&lt;p&gt;Enforcement is what stops criminals sending invoices and password resets in your name. It also unlocks BIMI, which can show your logo next to your mail in supporting inboxes when combined with a verified mark certificate.&lt;/p&gt;

&lt;h2&gt;Common mistakes&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Two SPF records after a new service is added, which invalidates both.&lt;/li&gt;
  &lt;li&gt;A sending service set up without DKIM, so its mail fails DMARC once enforcement begins.&lt;/li&gt;
  &lt;li&gt;DMARC reports sent to a mailbox nobody reads.&lt;/li&gt;
  &lt;li&gt;Jumping straight to &lt;code&gt;p=reject&lt;/code&gt; and blocking legitimate invoices or password resets.&lt;/li&gt;
  &lt;li&gt;Mail records lost during a DNS provider change because only the web records were copied.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;Domains that don't send mail&lt;/h2&gt;
&lt;p&gt;Parked and secondary domains get spoofed too. Protect them with an SPF record of &lt;code&gt;v=spf1 -all&lt;/code&gt;, a DMARC record at &lt;code&gt;p=reject&lt;/code&gt;, and no MX records if they receive no mail.&lt;/p&gt;

&lt;h2&gt;Changing DNS without breaking mail&lt;/h2&gt;
&lt;p&gt;Website moves, CDN setups, and DNS provider changes are when mail records get lost. Export every record first, recreate them exactly, keep MX, SPF, DKIM, and DMARC untouched unless they're the point of the change, and send test messages afterward. Header analyzers show exactly which checks passed.&lt;/p&gt;

&lt;p class="related"&gt;I set this up and troubleshoot deliverability as &lt;a href="https://gerriscorp.com/services/email-and-dns/"&gt;email authentication and DNS&lt;/a&gt; work.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://www.rfc-editor.org/rfc/rfc7208"&gt;RFC 7208: Sender Policy Framework (SPF)&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://www.rfc-editor.org/rfc/rfc7489"&gt;RFC 7489: DMARC&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/a/answer/81126"&gt;Google Workspace Admin Help: Email sender guidelines&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>"Crawled, currently not indexed": what it means and how to fix it</title>
    <link>https://gerriscorp.com/guides/crawled-not-indexed/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/crawled-not-indexed/</guid>
    <pubDate>Mon, 05 Oct 2026 12:00:00 +0000</pubDate>
    <description>What Google Search Console's Crawled, currently not indexed status means, how it differs from Discovered, and a step-by-step method to find the real cause.</description>
    <content:encoded>&lt;h1&gt;"Crawled, currently not indexed": what it means and how to fix it&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-05"&gt;October 5, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;Google visited the page, read it, and decided not to put it in the index, at least for now. It's the most misunderstood status in Search Console, and the fix depends entirely on why Google made that call.&lt;/p&gt;

&lt;h2&gt;Where you see it&lt;/h2&gt;
&lt;p&gt;In Google Search Console, open Indexing, then Pages. The "Why pages aren't indexed" table lists every reason Google has for leaving URLs out. "Crawled, currently not indexed" means Googlebot fetched the URL successfully and chose not to index it. Google gives no specific reason, and the status alone isn't an error. Every large site has some URLs here, and many of them deserve to be.&lt;/p&gt;

&lt;h2&gt;How it differs from "Discovered, currently not indexed"&lt;/h2&gt;
&lt;p&gt;"Discovered, currently not indexed" means Google knows the URL exists and hasn't crawled it yet. That usually points to crawl priority: too many URLs, weak internal linking, a slow server, or a site Google doesn't consider worth crawling quickly. "Crawled, currently not indexed" means Google did crawl the page and judged it not worth indexing. One is about getting Google to look; the other is about what Google saw when it looked.&lt;/p&gt;

&lt;h2&gt;Common causes&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Thin or duplicate content.&lt;/strong&gt; Pages that say little, or say nearly the same thing as other pages: tag archives, near-identical location pages, product variants, and boilerplate-heavy templates.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Content hidden from the renderer.&lt;/strong&gt; The page looks rich to visitors, but Google's rendered HTML holds only the template, because the substance loads on scroll or click. The &lt;a href="https://gerriscorp.com/guides/javascript-seo/"&gt;JavaScript SEO guide&lt;/a&gt; shows how to check.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Weak signals of importance.&lt;/strong&gt; Few internal links, no external links, buried deep in the site.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Low-value URL patterns.&lt;/strong&gt; Filter and sort parameters, internal search results, and session URLs that multiply without adding anything.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Soft 404s and empty states.&lt;/strong&gt; Out-of-stock or "no results" pages that return a 200 status.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;A new or recently changed site.&lt;/strong&gt; Google is still deciding, and many URLs move into the index over the following weeks.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;A method that finds the real cause&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;&lt;strong&gt;Export the examples.&lt;/strong&gt; Click the status in the report and export the URL list. Search Console shows a sample of up to 1,000 URLs.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Group them by template.&lt;/strong&gt; Sort the URLs into page types: product, category, article, location, tag, parameter. A status spread evenly across the site tells a different story than a cluster in one template.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Ask whether each group deserves indexing.&lt;/strong&gt; Parameter URLs, tag pages, and near-duplicates are often correctly excluded. Handle those with canonicals, noindex, or by removing the links that create them, and move on.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Inspect the important ones.&lt;/strong&gt; For pages that should rank, run URL Inspection, test the live URL, and read the rendered HTML. Is the main content there? Is the canonical the URL you expect? Is the page returning 200?&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Compare with pages that are indexed.&lt;/strong&gt; Crawl both groups and compare word counts, internal links, duplication, and load time. The difference usually points at the cause.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Check the dates.&lt;/strong&gt; Line up when the count started rising with deployments, template changes, migrations, and plugin updates.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Keep other problems separate.&lt;/strong&gt; A backlink spike or a redirect error at the same time may be a coincidence. Investigate each on its own track until the data connects them.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;Fixes, by cause&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;Thin or duplicate pages: merge them, expand them with real substance, or canonicalize them to the strongest version.&lt;/li&gt;
  &lt;li&gt;Hidden content: render it in the server's HTML.&lt;/li&gt;
  &lt;li&gt;Weak signals: link to the pages from relevant, indexed pages and from navigation where it makes sense.&lt;/li&gt;
  &lt;li&gt;Low-value patterns: stop generating crawlable links to them, and keep them out with canonicals or noindex.&lt;/li&gt;
  &lt;li&gt;Soft 404s: return a real 404 or 410, or give the page useful content.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;After the fix ships, click "Validate fix" in the report. Validation takes days to weeks, and Google reprocesses pages at its own pace. Request indexing through URL Inspection for a handful of the most important pages.&lt;/p&gt;

&lt;h2&gt;When it's a crisis&lt;/h2&gt;
&lt;p&gt;A sudden rise concentrated in your most valuable template, with a matching traffic drop, is a technical emergency, and it's usually a rendering or template change. A slow, even scatter across low-value pages is housekeeping. Telling the two apart is the first job.&lt;/p&gt;

&lt;p class="related"&gt;I diagnose indexing problems as part of &lt;a href="https://gerriscorp.com/services/technical-seo/"&gt;technical SEO forensics&lt;/a&gt; and &lt;a href="https://gerriscorp.com/services/search-console/"&gt;Search Console cleanup&lt;/a&gt;.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://support.google.com/webmasters/answer/7440203"&gt;Search Console Help: Page indexing report&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics"&gt;Google Search Central: JavaScript SEO basics&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
  <item>
    <title>How AI assistants decide what to say about your business</title>
    <link>https://gerriscorp.com/guides/ai-assistants/</link>
    <guid isPermaLink="true">https://gerriscorp.com/guides/ai-assistants/</guid>
    <pubDate>Mon, 05 Oct 2026 12:00:00 +0000</pubDate>
    <description>Where ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews get their answers about a business, which crawlers feed them, and what you can do about it.</description>
    <content:encoded>&lt;h1&gt;How AI assistants decide what to say about your business&lt;/h1&gt;
&lt;p class="byline"&gt;By &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt; · Published &lt;time datetime="2026-10-05"&gt;October 5, 2026&lt;/time&gt;&lt;/p&gt;
&lt;p class="lead"&gt;When someone asks ChatGPT, Claude, Gemini, Perplexity, or Google's AI Overviews about your company, the answer comes from two places: what the model learned in training, and what it finds by searching the web right then. You can influence both, and the second is where most of the action is.&lt;/p&gt;

&lt;h2&gt;Two sources of knowledge&lt;/h2&gt;
&lt;h3&gt;Training data&lt;/h3&gt;
&lt;p&gt;Every large language model is trained on a huge snapshot of text, much of it crawled from the public web, and that knowledge stops at a training cutoff. What a model "knows" about you from training reflects whatever was public and widely repeated before that date. It changes only when the company trains a new model, so it lags reality by months or years. Older press coverage, Wikipedia, directories, and anything that was repeated often all weigh heavily here.&lt;/p&gt;
&lt;h3&gt;Live retrieval&lt;/h3&gt;
&lt;p&gt;Most assistants now search the web before answering questions about specific businesses, products, and recent events. The assistant turns the question into one or more searches, reads the top results, and writes an answer grounded in those pages, often with citations. This is called retrieval or grounding, and it means a page that ranks well and answers the question clearly can shape the answer within days of being published.&lt;/p&gt;

&lt;h2&gt;Which search indexes feed which assistants&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Google AI Overviews and AI Mode&lt;/strong&gt; are built on Google Search's own index, crawled by Googlebot.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Gemini&lt;/strong&gt; grounds its answers with Google Search.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Microsoft Copilot&lt;/strong&gt; uses Bing's index.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;ChatGPT search&lt;/strong&gt; uses OpenAI's own crawler, OAI-SearchBot, alongside third-party search providers.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Perplexity&lt;/strong&gt; runs its own crawler, PerplexityBot, and its own index.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Claude&lt;/strong&gt; searches the web through a search provider and fetches pages with Anthropic's crawlers.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;The practical lesson: being well indexed in both Google and Bing covers most of the AI landscape. Bing Webmaster Tools and IndexNow, which tells Bing and other engines about changed pages within minutes, matter more than they used to.&lt;/p&gt;

&lt;h2&gt;The crawlers, and what each one is for&lt;/h2&gt;
&lt;table&gt;
&lt;thead&gt;&lt;tr&gt;&lt;th scope="col"&gt;Company&lt;/th&gt;&lt;th scope="col"&gt;Training&lt;/th&gt;&lt;th scope="col"&gt;Search index&lt;/th&gt;&lt;th scope="col"&gt;Fetches for a user&lt;/th&gt;&lt;/tr&gt;&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;&lt;td&gt;OpenAI&lt;/td&gt;&lt;td&gt;GPTBot&lt;/td&gt;&lt;td&gt;OAI-SearchBot&lt;/td&gt;&lt;td&gt;ChatGPT-User&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Anthropic&lt;/td&gt;&lt;td&gt;ClaudeBot&lt;/td&gt;&lt;td&gt;Claude-SearchBot&lt;/td&gt;&lt;td&gt;Claude-User&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Perplexity&lt;/td&gt;&lt;td&gt;(none listed)&lt;/td&gt;&lt;td&gt;PerplexityBot&lt;/td&gt;&lt;td&gt;Perplexity-User&lt;/td&gt;&lt;/tr&gt;
&lt;tr&gt;&lt;td&gt;Google&lt;/td&gt;&lt;td&gt;Google-Extended (a robots.txt token)&lt;/td&gt;&lt;td&gt;Googlebot&lt;/td&gt;&lt;td&gt;Google's user-triggered fetchers&lt;/td&gt;&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;You can allow or block each in robots.txt. Blocking a training crawler keeps your future pages out of future models; blocking a search crawler keeps you out of that assistant's live answers. Google-Extended controls whether your content is used for Gemini, and it doesn't affect Google Search or AI Overviews, which follow Googlebot. Check your CDN, too: Cloudflare began blocking many AI crawlers by default on new setups in 2025, which can keep a business out of AI answers without anyone deciding that on purpose.&lt;/p&gt;

&lt;h2&gt;What assistants favor when they write an answer&lt;/h2&gt;
&lt;ul&gt;
  &lt;li&gt;&lt;strong&gt;Direct answers.&lt;/strong&gt; Pages that state the answer plainly near the top, in a sentence that still makes sense when quoted on its own.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Self-contained context.&lt;/strong&gt; Definitions, dates, locations, and names on the page itself, so the assistant doesn't have to guess.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Consistency.&lt;/strong&gt; The same facts about you on your site, your profiles, directories, and press. Conflicting facts produce hedged or wrong answers.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Third-party confirmation.&lt;/strong&gt; Reviews, news coverage, industry directories, and community discussions. Assistants treat what others say about you as evidence.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Freshness.&lt;/strong&gt; Visible "updated" dates and current information, especially for anything that changes.&lt;/li&gt;
  &lt;li&gt;&lt;strong&gt;Readable HTML.&lt;/strong&gt; Most AI crawlers fetch the HTML without running JavaScript, so content that appears only after scripts run is invisible to them. The &lt;a href="https://gerriscorp.com/guides/javascript-seo/"&gt;JavaScript SEO guide&lt;/a&gt; shows how to check.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;What you can do this quarter&lt;/h2&gt;
&lt;ol&gt;
  &lt;li&gt;Ask the major assistants the ten questions your buyers ask, and write down the answers and the sources cited.&lt;/li&gt;
  &lt;li&gt;Fix your own pages first: one clear page per service or product, with the facts stated plainly and an FAQ section with real answers.&lt;/li&gt;
  &lt;li&gt;Make your facts consistent everywhere: name, description, location, founding date, leadership, services, prices where public.&lt;/li&gt;
  &lt;li&gt;Add schema.org structured data for your organization, people, products, and services, matching the visible page.&lt;/li&gt;
  &lt;li&gt;Check robots.txt and your CDN's bot settings for the AI crawlers you want.&lt;/li&gt;
  &lt;li&gt;Submit your sitemap to Google Search Console and Bing Webmaster Tools, and turn on IndexNow.&lt;/li&gt;
  &lt;li&gt;Consider an &lt;a href="https://gerriscorp.com/guides/glossary/#llms-txt"&gt;llms.txt&lt;/a&gt; file. It's a proposed convention from 2024 that gives language models a plain summary of your site; adoption by AI companies is still uncertain, and it costs almost nothing.&lt;/li&gt;
  &lt;li&gt;Repeat the same questions in a month and compare.&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;Measuring it&lt;/h2&gt;
&lt;p&gt;AI answers vary with phrasing, location, account history, and time, so a single test proves little. Use a fixed set of prompts, run them on a schedule, and record the answers and citations. Add referral traffic from chatgpt.com, perplexity.ai, gemini.google.com, and copilot.microsoft.com in analytics, and watch which pages they land on.&lt;/p&gt;

&lt;p class="related"&gt;I do this work as &lt;a href="https://gerriscorp.com/services/ai-search/"&gt;AI search visibility&lt;/a&gt;, starting with an AI search readiness audit.&lt;/p&gt;
&lt;aside class="help-box" aria-label="Get help"&gt;
  &lt;h2&gt;Want me to handle this for you?&lt;/h2&gt;
  &lt;p&gt;This guide shows you how to do it yourself. If you'd rather hand it to someone who does this work for clients, I'm glad to help: tell me what's happening and I'll tell you where I'd start.&lt;/p&gt;
  &lt;p&gt;&lt;a class="button" href="https://calendly.com/chrisabraham/30"&gt;Book a 30 minute call&lt;/a&gt; &lt;a class="button" href="tel:+12023525051"&gt;Call +1 202-352-5051&lt;/a&gt; &lt;a class="button" href="mailto:chris@gerriscorp.com"&gt;Email chris@gerriscorp.com&lt;/a&gt;&lt;/p&gt;
  &lt;p&gt;If your company hires through Upwork, you'll find me &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;there as well&lt;/a&gt;.&lt;/p&gt;
&lt;/aside&gt;
&lt;h2&gt;References&lt;/h2&gt;
&lt;ul class="sources"&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/appearance/ai-features"&gt;Google Search Central: AI features and your website&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://developers.google.com/search/docs/fundamentals/creating-helpful-content"&gt;Google Search Central: Creating helpful, reliable, people-first content&lt;/a&gt;&lt;/li&gt;
  &lt;li&gt;&lt;a href="https://platform.openai.com/docs/bots"&gt;OpenAI: Overview of OpenAI crawlers&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;
&lt;aside class="author-box" aria-label="About the author"&gt;
  &lt;img src="https://gerriscorp.com/chris-abraham.jpg" width="72" height="72" alt=""&gt;
  &lt;p&gt;Written by &lt;a href="https://gerriscorp.com/about/"&gt;Chris Abraham&lt;/a&gt;, founder of Gerris Corp in Arlington, Virginia, from his own client engagements, with every technical claim checked against the documentation listed. Websites since 1994, digital marketing since 2002. &lt;a href="https://gerriscorp.com/standards/"&gt;Editorial standards&lt;/a&gt; · &lt;a href="https://www.upwork.com/freelancers/chrisjabraham"&gt;Upwork profile&lt;/a&gt;&lt;/p&gt;
&lt;/aside&gt;</content:encoded>
  </item>
</channel>
</rss>
