Scorer changelog

13 August 2026 to 28 August 2026

The 2026-08-12 epoch stopped a knowledge disclaimer losing its veto to the remedy that followed it. Re-scoring against the labelled set showed six answers still counting as citations, every one of them an engine saying plainly that it did not know the brand. All three causes were vocabulary rather than logic. First, the disclaimer list held "could not find" and "cannot find" but not the contracted "couldn't find", so one spelling of a sentence was an absence and the other was a citation, decided by an apostrophe. Second, "Without public reviews or case studies on X, their effectiveness remains unclear" is an absence of evidence stated about the brand, and it was rescuing the mention because it is a clean clause that happens to name it. Third, a clause that puts a question back to the reader is part of the not-knowing response, not an answer: "do you have a website or company name associated with X?" is the engine admitting it does not know what X is. That rule only applies where the answer disclaims the brand somewhere, so an ordinary question in a real recommendation is untouched. This only ever REMOVES citations. Re-scoring all 5,977 stored cells, 11 stop counting and none start, moving the mention rate from 21.78% to 21.60%. Every one of the 11 was checked by hand and is an engine disclaiming knowledge of the brand.

12 August 2026 to 13 August 2026

Three corrections to what counts as a mention, found by measuring the classifier against a hand-labelled set for the first time. Two of them REMOVE citations, so scores may step down, and that is the correction rather than a drop in visibility. First, a knowledge disclaimer no longer loses its veto to the remedy that follows it. "I have no information about X. To find out, search for X on Google" names the brand twice, and the second clause was clean, so the answer counted as a citation; the same held for "could you tell me more about X?". That is the stock response to a brand an engine does not know, and it was the single largest error in the classifier. Second, a name attributed to a different vendor is that vendor's product: "SearchScore by SearchUnify" and "SearchNode (formerly SearchScore)" were both counted as ours, which inflated our own visibility, and any brand whose name is a common compound was exposed to it. Third, a multi-word brand now matches its own singular or plural in the final word only, so "Millennium Hotels" is found in "Millennium Hotel and Conference Centre". That one ADDS citations that were always real. A single-word brand is untouched, so Boots can never match Boot.

11 August 2026 to 12 August 2026

Three changes to how a citation is read. A rank is now stated only when the markdown reading and the classifier reading of the same answer agree on the same number; they agreed on 16% of the cells where either had an opinion, and the markdown reading alone was what customers saw. An uncorroborated rank is credited as a prose mention, never as a low rank. The markdown reading itself was also fixed: its list counter did not reset between lists, a prose mention above a list beat the rank below it, and a sentence after a list of rivals inherited the count, which reported a top-3 rank for a brand that was never in the list. Sentiment is now weighted by the confidence the classifier returned with it, which was previously stored and ignored, so a hedged verdict moves the score less than a certain one. Scores may step down slightly where ranks are no longer confirmed; that is the correction, not a change in visibility.

9 August 2026 to 11 August 2026

Adds a transactional intent, completing IAB's four intent types. It is decision-stage, being the closest of the four to a purchase, so 8 question-instances across the live sets move into the scored set. Prompt-type coverage goes from 3 of 4 to 4 of 4, which is the criteria-matrix row separating directional from decision-grade. Brand-named comparisons were NOT reclassified: that behaviour is deliberate and locked by tests/avs-brand-safety, and they are now marked by isBrandedComparison for separate reporting instead.

27 July 2026 to 9 August 2026

Provenance dropped from cell credit: it measured engine retrieval capability (Perplexity 99% vs Gemini 18%) rather than the brand. Sentiment and position only.

Until 27 July 2026

AVS weighted by sentiment, position and provenance.

From 27 October 2026 · scores move in both directions

The terms-of-service check now recognises more ways sites label that link, and only applies to sites that take orders, bookings or sign-ups. The checks for named experts, original data and a distinct point of view now read the other pages the scanner fetched, not only the homepage. An image with an empty description now counts as having no description unless it is decorative. A domain is treated as parked only when the scanner sees parking evidence itself, and a sign-in screen or parking page that appears only once the page has loaded is no longer scored as the website. Three checks that look for phrases about named experts, original data and a distinct point of view are now shown as best-practice tips and no longer count towards the score. Text from neighbouring parts of a page is no longer run together, so phrases at the start of a section are recognised. Privacy-policy links labelled in Korean, Chinese or Japanese are now recognised.

From 27 October 2026 · scores move in both directions

When a homepage answers with an error, a maintenance notice, a redirect page, a near-empty page or a certificate problem, the scanner now follows the redirect or tries again in a full browser, and scores the site if it reaches it. If it still cannot reach the site, it says the site could not be reached for now, instead of saying it is not a website. A parked domain that moves visitors to a rotating parking address is recognised as parked. A site that answers an address that does not exist with an ordinary page, instead of a "not found" error, now fails the soft-404 check. One slow page no longer stops the scanner checking the rest of a site's footer pages, and a page that is still loading gets a few more seconds before it is read.

From 27 October 2026 · scores move down

The scanner now recognises more pages that stand in for a website without being it, such as a parking page's browser check, a link-tracking domain that answers with its own notice, and an empty holding page. These are no longer scored as the business's website. A homepage that refuses the scanner with HTTP 401 is now handled the same way as one that refuses it with HTTP 403: as a block, not as the site's own page. No check has changed how it scores: a page that is now refused gets no score instead of one.

From 27 October 2026 · scores move in both directions

A bot-check page that answers instead of the site is now treated as not measured, not as a site that needs JavaScript. The HTTPS redirect is now tested by visiting the http:// address, so a site that still serves pages over http:// is told so. Parked pages, closed shops, maintenance notices and bare sign-in screens are no longer scored as websites. The scanner now finds posts behind a blog or news index, including news pages in other languages, and counts an FAQ built as a list of expandable questions. Scores can move in either direction.

From 24 October 2026 · scores move up

The scanner now reads more of the pages that say who a business is and what it offers. It follows the site's own link to its imprint, legal notice or contact page, and reads one page from the section of the site that holds what it sells, such as a care home or a service page. Contact details and structured data on those pages now count, and a strip of client logos introduced by a line that counts customers, such as "200+ brands", is recognised as client logos.

Measured on the same live pages from 132 sites: 38 scores rose, by a median of 0.9 GEO points and up to 6.5, and 94 were unchanged. None fell.

From 24 October 2026 · scores move up

When a site links its blog or articles and they do not answer in time, the scanner now asks once more, with a longer wait, while the audit still has time for it. If they still do not load, Article markup is left out of the score instead of being marked as missing: the structured data score is worked out from the checks that remain, so an unreachable blog neither costs nor earns points. The Fix List then says we could not reach the blog, rather than asking for markup the site may already have. The FAQ card also now asks for FAQ structured data, not a new FAQ section, when that is all that is missing.

From 23 October 2026 · scores move in both directions

Blog posts and FAQ pages now get the credit they have earned. The scanner goes straight to a site's real posts: it tries the blog first, skips index and archive pages and file downloads, recognises posts published at the top level of the site, and reads the news feed a site declares when its blog loads its posts in the browser. FAQ, HowTo and Q&A markup on a site's own FAQ page now counts too.

Measured on the same live pages from 132 sites: 20 scores rose, by a median of 4.6 GEO points and up to 9.9, and 111 were unchanged. One fell, by under 1 point, where the larger posts now read used up the audit's download allowance before its sitemap was checked.

From 21 October 2026 · scores move up

We no longer score a Grok crawler check. The name we checked for is not one xAI publishes, so no robots.txt names it, and the check could only fail on a site whose robots.txt blocks every crawler it does not name. That blanket block is still scored, by its own check and by the named AI crawlers.

Measured on the 6,000 most recent audits: 5,817 scores were unchanged and 183 rose, by 0.5 or 0.6 GEO points. Those 183 are the sites with a blanket block.

From 20 October 2026 · scores move down

A link to a page listing several policies still counts as both a privacy policy and terms, as it has since geo-162. That reading now applies only to a page of several policies, such as /policies or a link labelled "Policies". A link to one document, such as "Privacy policy" at /policy, counts as that document alone and no longer as terms too.

Measured on the same live pages from 139 sites: 138 scores were unchanged and 1 fell, by 0.30 GEO points, where a link to a single privacy policy had been counted as terms too.

From 20 October 2026 · scores unchanged

A number after an offer word such as "Save", "Buy" or "Over" is no longer read as a house number, however the word is capitalised, so "SAVE 5 Market Place deals" is not mistaken for a postal address.

Measured on the same live pages from 142 sites: every score was unchanged.

From 20 October 2026 · scores move up

A postal address is now found when a page prints it in separate pieces of markup, as many site builders do, and a street name that begins with "The" or contains a hyphen is recognised. A title such as "5 Easy Ways To Drive Traffic" is no longer mistaken for an address.

Measured on the same live pages from 141 sites: 138 scores were unchanged and 3 rose, each by 0.91 GEO points, where an address was already on the page. None fell.

From 20 October 2026 · scores move up

Links to an About page, a contact page and a privacy policy or terms page are now recognised however a site writes them, including relative links such as about.html and plain menu labels such as "About" and "Contact". A single footer link to a policies page now counts as the privacy policy and the terms it leads to.

Contact details now include an email address protected by Cloudflare, which hides it from automated readers, and a postal address written out on the page in any of the languages we measure, not only a UK one.

Measured on the same live pages from 140 sites: 123 scores were unchanged and 17 rose. None fell, and no score moved by more than 1.96 GEO points.

From 9 October 2026 · scores move in both directions

Each check now reads a page the way visitors and search crawlers actually experience it, so scores reflect what is really there.

Privacy policies are now recognised in other languages, such as a German /datenschutz page. Cloudflare's protected email links are no longer mistaken for broken links. Images marked as decorative, which is good accessibility practice, no longer count as missing alt text. Only the headings a visitor can actually see are read as the page headline. And the notice that a site needs JavaScript now appears only on pages that are genuinely empty without it.

Sitemaps are now checked properly: a sitemap must list at least one page, so an empty file or an ordinary web page sitting at the sitemap address no longer earns credit. Images with no alt text at all are now counted consistently on every page.

Measured on the same live pages from 141 sites: 109 scores were unchanged, 11 rose, and 21 dipped slightly where a sitemap was empty or images lacked alt text. No score moved by more than 0.73 GEO points.

From 26 August 2026 · scores move in both directions

Text extraction included the contents of script and style tags, so on a typical page about a third of what was measured as content was actually code, and sites were charged for the JavaScript they ship. Off-site presence now reads links and structured data rather than any appearance of a name in the source, so a page that merely used the word "Reddit" is no longer counted as having a Reddit presence. And the JavaScript-gating check compared a page against itself, which reported no gating for every site ever measured; it now compares what a browser sees against what a crawler sees. Measurement changed in BOTH directions: sites gain where they were charged for their own code, and lose where a presence was counted that was never linked, or where content genuinely is hidden from crawlers. Measured across 446 sites the typical score did not move and the largest single change was 1.6 points.

From 26 August 2026 · scores move up

The answer-first check walked SIBLINGS of the h1, so any page builder that nests the heading in a div made a perfectly good lead paragraph unreachable and the check read an empty string. The social-proof checks never fetched /reviews/, /testimonials/, /press/ or /case-studies/, which are the pages that hold reviews and case studies. And the copyright proxy demanded a two-year span, so an accurate "2025-2026" failed while a stale "2025" passed. Measurement moves UP only, and only for sites that already had the signal.

From 25 August 2026 · scores move in both directions

Two readings of robots.txt were sharpened. A field our parser did not recognise, such as the Content-Signal line Cloudflare now adds by default, no longer ends the rule group it sits in, so a following Allow line is honoured rather than discarded. And a group is now closed by the next User-agent line, as the standard requires, so a rule written for one crawler can no longer apply to a crawler named earlier in the file. Measured by parsing the same 448 live robots.txt files with both versions: 94.0% of sites are unchanged, 21 moved up and 6 moved down, every move being 5.3 GEO points or less. Scores move in BOTH directions and some sites are newly reported as blocking, which is the correction working rather than a regression: a file that disallows everything and then whitelists named crawlers really does block the AI crawlers it leaves out, and the old grouping leaked those whitelist Allow lines to every agent in the file.

From 25 August 2026 · scores move up

A YouTube channel is now recognised when it is linked by its @handle, the format YouTube has issued since 2022, and when it is declared in Organization structured data rather than as a link. Measurement improved: worth 3 points of brand authority, and sites that had a channel all along are now credited for it, so scores move up only.

From 25 August 2026 · scores move in both directions

Claude-SearchBot, which indexes content for search, is now scored separately from ClaudeBot, which collects content that may contribute to training. The existing weight is split rather than added to, so no total moves. Measurement changed in BOTH directions: a site that allows search while declining training gains, and a site blocking the search crawler now sees the visibility gap it previously was not shown.

From 20 August 2026 · scores move up

The four trust-page signals now recognise contact, about, privacy and terms pages in languages whose words do not contain the English spelling, and the canonical-domain check is sharpened again. Measurement improved: worth up to 15 points of brand authority, and non-English sites gain most. Both changes move scores up only.

From 20 August 2026 · scores move up

The rendering signal now distinguishes a page that hides its content behind JavaScript from one whose crawler request was refused by a firewall or timed out. Measurement improved: a site that was never gated is no longer told that it was, and scores move up for every site whose audit had been blocked rather than gated.

From 19 August 2026 · scores move down

The llms.txt signal now assesses the file against the specification - a Markdown document with a heading and at least one link - rather than recording only that a file exists. Measurement changed and this one moves DOWN: a site whose file exists but does not meet the spec now scores lower than under geo-152. Adoption figures published before this generation were measured on presence and continue to say so.

From 19 August 2026 · scores move up

Article detection now distinguishes a published article from a grid of cards that happens to use the same markup. Measurement improved: homepages laid out as card grids are no longer asked for a publication date they were never going to carry, and gain the 6 points that question was costing them.

From 19 August 2026 · scores move up

Breadcrumb, Article and FAQPage signals in the SEO score are now judged across the same pages the AI-search score already used. Measurement improved: the two halves of a report now agree, and a site credited by one is no longer queried by the other.

From 19 August 2026 · scores move up

FAQ and speakable signals are now judged across every page the audit fetches rather than the homepage alone, and the audit reaches /faq, /faqs and /pricing alongside /about. Measurement improved: a site publishing questions away from its homepage is now credited for them, so scores move up only.

From 19 August 2026 · scores move up

The canonical-domain signal now compares the host the alternate address resolves to, so a site that correctly redirects one address to the other is recognised as consistent. Measurement improved: sites with a single correct canonical gain 3 points they were previously not credited with.

From 13 August 2026 · scores move up

The quotable-statistics signal now recognises money outside dollars, ordinary English sentences where the figure follows the object, verb-final languages, and a space as the thousands separator. It is worth 8 points. Measurement improved: scores read higher than geo-146 for any site previously unread, and non-anglophone sites gain most.

From 1 August 2026 · scores move up

The audit now reads authorship and publication dates from structured data, follows a site's own navigation to find its content, and no longer counts a soft 404 as a content page. Measurement improved: scores read higher than geo-145 for the same site, so figures must name the generation they were measured under.