What we threw out
These are not claims that are false. They are answers the research wrote that Crawlspace would not publish, because the sentence said more than the quote behind it actually proved. Several of them are true and badly sourced. They are here so you can judge whether the checking is real, which you cannot do from a site that only shows you what passed.
Read them as notes about our wording, not as findings about the law. If you want the answer to something, the subjects are where it is - everything on this page was kept out of them on purpose.
203 discarded drafts. A further 50 are not shown at all: each was rejected because it put a statement in the wrong organisation’s mouth, and printing it here would repeat that about a real business to make a point about our pipeline.
AI Mode10
AI Mode's new buttons on the Google home page are a small test on desktop.
Why it was not published: The quote confirms a small desktop test but does not identify the buttons as AI Mode's.
The part the quote did not support: AI Mode's new buttons
What it rested on
Update: Google's VP of Product, Robby Stein, has confirmed this is a test saying, "This is a small test on desktop to help people find new things they can do with Search. No impact to how the core search box works – just add your question and hit enter as usual (which is how most people search anyways!)"
AI Mode's new buttons on the Google home page help people find new things they can do with Search.
Why it was not published: The quote confirms a small desktop test to help people find new things they can do with Search, but it does not mention AI Mode or buttons.
The part the quote did not support: AI Mode's new buttons on the Google home page
What it rested on
Update: Google's VP of Product, Robby Stein, has confirmed this is a test saying, "This is a small test on desktop to help people find new things they can do with Search. No impact to how the core search box works – just add your question and hit enter as usual (which is how most people search anyways!)"
AI Mode is being tested or rolled out on the Google home page.
Why it was not published: The quote supports AI features/buttons on the Google home page but does not name or establish the specific 'AI Mode'.
The part the quote did not support: AI Mode
What it rested on
The Google home page is testing or rolling out now, new AI features and buttons in place of the Google Search and I'm Feeling Lucky buttons. There is now "Create images," "Ask about files," and "Brainstorm" buttons below the big search box.
AI Mode’s buttons are “Create images”, “Ask about files” and “Brainstorm”.
Why it was not published: The quote states those buttons appear on the Google home page, but it never names or establishes them as belonging to 'AI Mode'.
The part the quote did not support: AI Mode’s
What it rested on
The Google home page is testing or rolling out now, new AI features and buttons in place of the Google Search and I'm Feeling Lucky buttons. There is now "Create images," "Ask about files," and "Brainstorm" buttons below the big search box.
AI Mode’s buttons appear below the search box.
Why it was not published: The quote mentions new AI feature buttons below the search box but never names 'AI Mode.'
The part the quote did not support: AI Mode’s
What it rested on
The Google home page is testing or rolling out now, new AI features and buttons in place of the Google Search and I'm Feeling Lucky buttons. There is now "Create images," "Ask about files," and "Brainstorm" buttons below the big search box.
AI Mode’s buttons replace the Google Search and I'm Feeling Lucky buttons.
Why it was not published: The quote says new AI features and buttons replace the Google Search and I'm Feeling Lucky buttons, but it never mentions 'AI Mode.'
The part the quote did not support: AI Mode’s
What it rested on
The Google home page is testing or rolling out now, new AI features and buttons in place of the Google Search and I'm Feeling Lucky buttons. There is now "Create images," "Ask about files," and "Brainstorm" buttons below the big search box.
AI Mode's documented position is that robots.txt directives for Googlebot are the control for site owners to manage access to how their sites are crawled for Search, because AI is built into Search.
Why it was not published: The quote is from Google Search Central documentation and says AI is built into Search, but does not cite or mention AI Mode.
The part the quote did not support: AI Mode's documented position
What it rested on
AI is built into Search and integral to how Search functions, which is why robots.txt directives for Googlebot is the control for site owners to manage access to how their sites are crawled for Search.
AI Mode's documented position is that Google Search Central's guide to optimizing for generative AI features states that a site must be included in Search generative AI features in Search Console to be eligible for display in generative AI features on Google Search.
Why it was not published: The quote is from Google Search Central documentation and states the guide's requirement, but no cited quote is from or establishes AI Mode's documented position.
The part the quote did not support: AI Mode's documented position
What it rested on
In addition to the technical requirements for Search, a site must be included in Search generative AI features in Search Console to be eligible for display in generative AI features on Google Search.
AI Mode's documented position states that no third-party tool has access to Google's internal ranking or AI systems.
Why it was not published: The cited quote is from Google Search Central documentation, not identified as AI Mode's documented position.
The part the quote did not support: AI Mode's documented position
What it rested on
Be wary of third-party tools that promise ranking success or claim to use "internal" Google metrics. No third-party tool has access to our internal ranking or AI systems.
A study from SE Ranking analyzed more than 50,000 commercial prompts across 20 niches to understand how frequently AI Mode serves ads.
Why it was not published: The quote states the study analyzed ChatGPT ads, not AI Mode.
The part the quote did not support: AI Mode
What it rested on
A study from SE Ranking analyzed more than 50,000 commercial prompts across 20 niches to understand how frequently ChatGPT serves ads, where they appear and how relevant they are.
AI Overviews5
AI Overviews content performance in generative AI features is measurable via a Generative AI performance report in Search Console, to which Google's AI optimization guide directs site owners.
Why it was not published: The Google quote supports measuring content in generative AI features via the report, but does not specifically name AI Overviews as the measured content.
The part the quote did not support: AI Overviews content performance
What it rested on
To measure how your content is performing in generative AI features on Google Search and Discover, use the Generative AI performance report in Search Console.
It’s also telling that despite Google’s optimistic claim that “links included in AI Overviews get more clicks than if the page had appeared as a traditional web listing for that query”, there is still no way to disambiguate AI Overview clicks and impressions from the rest of your Search Console data.
Ahrefs re-ran its AI Overviews clickthrough study using data from December 2025 and found that the presence of an AI Overview now correlates with a 58% lower average clickthrough rate for the top-ranking page, up from the 34.5% it found in April 2025.
Why it was not published: The cited quotes do not mention December 2025 or any specific date for the re-run; they only say 'now.'
The part the quote did not support: using data from December 2025
What it rested on
Back in April 2025, we found that AI Overviews reduced clicks to top-ranking content by 34.5%, despite Google’s statements to the contrary.
We found that the presence of an AI Overview now correlates with a 58% lower average clickthrough rate for the top-ranking page.
AI Overviews documentation states that to limit the information shown from a site's pages in Search, site owners should use nosnippet, data-nosnippet, max-snippet, or noindex controls.
Why it was not published: The cited source is labeled 'AI Features and Your Website,' not 'AI Overviews' documentation, so that specific provenance is unsupported.
The part the quote did not support: AI Overviews documentation
What it rested on
To limit the information shown from your pages in Search, use nosnippet, data-nosnippet, max-snippet, or noindex controls.
Pew Research Center found that Wikipedia, YouTube and Reddit collectively accounted for 15% of the sources listed in the AI Overviews it examined.
Why it was not published: The quote says 'Collectively, they accounted for 15%' but does not name which sources 'they' refers to; the claim adds the specific platforms without direct support from the provided evidence.
The part the quote did not support: Wikipedia, YouTube and Reddit
What it rested on
Collectively, they accounted for 15% of the sources that were listed in the AI summaries we examined. They made up a similar share (17%) of the sources listed in standard search results.
Pew Research Center found that Wikipedia, YouTube and Reddit collectively accounted for 17% of the sources listed in standard search results, a share similar to their 15% share in AI Overviews.
Why it was not published: The quoted excerpt only refers to 'they' without naming Wikipedia, YouTube, and Reddit, so the claim adds specific entities not stated in the evidence.
The part the quote did not support: Wikipedia, YouTube and Reddit collectively
What it rested on
Collectively, they accounted for 15% of the sources that were listed in the AI summaries we examined. They made up a similar share (17%) of the sources listed in standard search results.
alt text12
Alt text documentation changed on March 2, 2026, according to Google's Search Central changelog.
Why it was not published: The quotes mention added preferred image metadata and image SEO best practices, but none state that alt text documentation changed.
The part the quote did not support: Alt text documentation
What it rested on
March 2026
March 2
Added preferred image best practices
What : Added a new section on specifying a preferred image with metadata to the image SEO best practices and Discover documentation .
The change to alt text documentation added a new section on specifying a preferred image with metadata to the image SEO best practices and Discover documentation.
Why it was not published: The quotes establish adding a new section on specifying a preferred image with metadata to image SEO and Discover docs, but none mention alt text documentation.
The part the quote did not support: alt text documentation
What it rested on
March 2026
March 2
Added preferred image best practices
What : Added a new section on specifying a preferred image with metadata to the image SEO best practices and Discover documentation .
WebAIM's measurement of the proportion of images missing alt text dropped from 18.5% in 2025 to 16.2% in 2026.
Why it was not published: The quotes support a drop from 18.5% in 2025 to 16.2%, but no cited quote states that the 16.2% figure is for 2026.
The part the quote did not support: in 2026
What it rested on
While there were notably more images, the proportion of images missing alt text dropped from 18.5% in 2025.
16.2% of all home page images (10.8 per page on average) had missing alternative text (not counting alt="").
Alt text is the subject of Accessibility Checker's claim that its analysis of 800 or more websites found that 73.4% experienced an increase in organic traffic after implementing accessibility improvements.
Why it was not published: The cited quote from Accessibility Checker does not mention alt text, only general accessibility improvements.
The part the quote did not support: Alt text is the subject of
What it rested on
Analyzing 800+ websites revealed that 73.4% experienced an increase in organic traffic after implementing accessibility improvements, with the average site seeing a 12% uplift in traffic within three months.
Alt text is the subject of Accessibility Checker's claim that its analysis of 800 or more websites found that the average site saw a 12% uplift in traffic within three months.
Why it was not published: The quote only mentions 'accessibility improvements', not alt text specifically.
The part the quote did not support: Alt text is the subject of
What it rested on
Analyzing 800+ websites revealed that 73.4% experienced an increase in organic traffic after implementing accessibility improvements, with the average site seeing a 12% uplift in traffic within three months.
The alt text states that seoscore.tools asserts that Google Images accounts for a larger share of Google searches than YouTube, Google Maps, and Google News combined.
Why it was not published: The cited quote is from a blog post, not an alt text, so it does not establish what the alt text states.
The part the quote did not support: The alt text states
What it rested on
Google Images accounts for 22.6% of all Google searches. That is more than YouTube, Google Maps, and Google News combined.
WebAIM's analysis of alt text in the top 1,000,000 home pages was conducted using the WAVE stand-alone API.
Why it was not published: The quotes establish the evaluation used the WAVE stand-alone API but do not mention alt text, so the specific claim about alt-text analysis is not grounded.
The part the quote did not support: analysis of alt text
What it rested on
The evaluation was conducted using the WAVE stand-alone API (with additional tools to collect site technology and category data).
The WAVE accessibility engine was used to analyze the rendered DOM (Document Object Model) of all pages after scripting and styles were applied.
WebAIM's analysis of alt text in the top 1,000,000 home pages was performed on the rendered DOM of all of the top 1,000,000 home pages after scripting and styles were applied.
Why it was not published: The quotes state the WAVE accessibility engine analyzed the rendered DOM of all pages, but none mentions alt text specifically.
The part the quote did not support: analysis of alt text
What it rested on
The evaluation was conducted using the WAVE stand-alone API (with additional tools to collect site technology and category data).
The WAVE accessibility engine was used to analyze the rendered DOM (Document Object Model) of all pages after scripting and styles were applied.
WebAIM found that missing alternative text for images was detected on 53.1% of the top 1,000,000 home pages.
Why it was not published: The quotes give the issue and 53.1% but do not state the surveyed population of top 1,000,000 home pages.
The part the quote did not support: top 1,000,000 home pages
WebAIM found that 45% of the images missing alternative text on the top 1,000,000 home pages were linked images.
Why it was not published: The quote states the 45% figure but does not specify the top 1,000,000 home pages sample.
The part the quote did not support: on the top 1,000,000 home pages
What it rested on
45% of the images missing alternative text were linked images—resulting in links that were not descriptive. One out of every four linked images was missing alternative text.
All automated accessibility tools including WAVE have limitations for alt text.
Why it was not published: The quote states automated tools including WAVE have general limitations but does not mention alt text specifically.
The part the quote did not support: for alt text
What it rested on
All automated tools, including WAVE, have limitations—not all conformance failures can be automatically detected. Absence of detected errors does not indicate that a page is accessible or conformant.
Not all conformance failures of alt text can be automatically detected.
Why it was not published: The quote only states a general limitation about conformance failures, not specifically alt text failures.
The part the quote did not support: of alt text
What it rested on
All automated tools, including WAVE, have limitations—not all conformance failures can be automatically detected. Absence of detected errors does not indicate that a page is accessible or conformant.
Backlinks7
Google's Search documentation updates log records that the 'Link spam' section for backlinks consolidates previous pages on Paid links and Link schemes.
Why it was not published: The quotes establish a Link spam update consolidating Paid links and Link schemes, but do not state the section is for backlinks.
The part the quote did not support: for backlinks
What it rested on
Refreshed and renamed the Webmaster Guidelines. Notable changes include:
Link spam : Consolidates previous pages on Paid links and Link schemes.
Google's documented position on backlinks, as stated in its March 2024 spam policies blog post, is that third-party 'reputation' or 'authority' scores for sites do not correspond to any of Google's own signals.
Why it was not published: The quote addresses third-party reputation/authority scores for sites but never mentions backlinks, so framing this as Google's documented position on backlinks overreaches.
The part the quote did not support: on backlinks
What it rested on
Please note: some third-party services provide "reputation" or "authority" scores for sites. These don't correspond to any of Google's own signals nor come from Google.
Google's documented position on backlinks, as stated in its March 2024 spam policies blog post, is that third-party 'reputation' or 'authority' scores for sites do not come from Google.
Why it was not published: The quote states third-party reputation/authority scores for sites don't come from Google, but it does not mention backlinks.
The part the quote did not support: on backlinks
What it rested on
Please note: some third-party services provide "reputation" or "authority" scores for sites. These don't correspond to any of Google's own signals nor come from Google.
Google's spam policies documentation defines link spam as the practice of creating backlinks to or from a site primarily for the purpose of manipulating search rankings.
Why it was not published: The cited quote says 'links,' not 'backlinks,' so the claim narrows the definition without support.
The part the quote did not support: backlinks
What it rested on
Link spam is the practice of creating links to or from a site primarily for the purpose of manipulating search rankings.
Backlinks have a documented position in a post by Google's Gary Illyes that was reproduced by Search Engine Roundtable, reading 'I shouldn't have said that... I definitely shouldn't have said that'.
Why it was not published: The quote only contains the stated text and does not indicate the post is about backlinks.
The part the quote did not support: Backlinks have a documented position
What it rested on
I shouldn't have said that... I definitely shouldn't have said that — Gary 鯨理／경리 Illyes (so official, trust me) (@methode) April 19, 2024
Backlinko's 11.8 million search results study about backlinks states that its research question was which factors correlate with first-page search engine rankings.
Why it was not published: The cited quotes do not mention the study's size (11.8 million search results); they only mention the research question.
The part the quote did not support: 11.8 million search results
What it rested on
Which factors correlate with first-page search engine rankings?
With the help of our data partner Ahrefs , we uncovered some interesting findings.
Backlinko's 11.8 million search results study about backlinks states that Ahrefs was its data partner.
Why it was not published: The quotes confirm Ahrefs as a data partner but do not mention that the study is about backlinks.
The part the quote did not support: about backlinks
What it rested on
Which factors correlate with first-page search engine rankings?
With the help of our data partner Ahrefs , we uncovered some interesting findings.
Bing Webmaster Tools1
The Bing Webmaster blog states that the grounding queries data shown in AI Performance represents a sample of overall citation activity.
Why it was not published: The quote says the key-phrase data shown represents a sample of citation activity, but it never names that data 'grounding queries' or explicitly ties the sample statement to a 'grounding queries' metric in AI Performance.
The part the quote did not support: grounding queries data shown in AI Performance
What it rested on
Shows the key phrases the AI used when retrieving content that was referenced in AI-generated answers. The data shown represents a sample of overall citation activity. We will continue to refine this metric as additional data is processed.
canonical tag2
Google's Search Central documentation on specifying a canonical URL lists redirects, rel="canonical" link annotations and sitemap inclusion in order of how strongly they can influence canonicalization, with redirects listed first.
Why it was not published: The quotes only mention that methods are ordered by canonicalization influence and list redirects first, not that rel="canonical" annotations and sitemap inclusion are named in that list.
The part the quote did not support: rel="canonical" link annotations and sitemap inclusion
What it rested on
To specify a canonical URL for duplicate or very similar pages to Google Search, you can indicate your preference using a number of methods. These are, in order of how strongly they can influence canonicalization:
Redirects: A strong signal that the target of the redirect should become canonical.
Google's Search Central documentation on specifying a canonical URL describes specifying canonicals in a sitemap as a less powerful signal to Google than the rel="canonical" mapping technique.
Why it was not published: The quote states a less powerful signal than rel="canonical" mapping but does not mention sitemaps, so the specific sitemap assertion is not established.
The part the quote did not support: specifying canonicals in a sitemap
What it rested on
Less powerful signal to Google than the rel="canonical" mapping technique.
changefreq3
An archived version of Google's Webmaster Tools help page on sitemaps described <changefreq> as providing a hint about how frequently the page is likely to change.
Why it was not published: The quote gives the hint sentence but does not link it to <changefreq>.
The part the quote did not support: <changefreq>
An archived version of Google's Webmaster Tools help page on sitemaps stated that the sitemap <priority> value does not affect a site's ranking in Google search results.
Why it was not published: The quote says only that something 'does not affect your site's ranking' but does not mention the sitemap <priority> value.
The part the quote did not support: sitemap <priority> value
Content Powered's article tells readers that changefreq is not going to make or break your website
Why it was not published: The quote only says 'It's still something you should tune, but it isn't going to make or break your website,' using the pronoun 'it' without explicitly mentioning 'changefreq'; the claim overreaches by asserting the specific term.
The part the quote did not support: changefreq
What it rested on
It's still something you should tune, but it isn't going to make or break your website.
Chrome UX Report (CrUX)1
Chrome UX Report (CrUX) is described in Google's page-experience documentation as having no single page experience signal that Google Search uses for ranking.
Why it was not published: The quote says there is no single page experience signal, but it does not mention or describe Chrome UX Report (CrUX) at all.
The part the quote did not support: Chrome UX Report (CrUX)
What it rested on
Is there a single "page experience signal" that Google Search uses for ranking? There is no single signal. Our core ranking systems look at a variety of signals that align with overall page experience.
Common Crawl9
Common Crawl announced in a blog post that the Common Crawl blog index dates June 1, 2026.
Why it was not published: The quotes show a blog post titled "Introducing the AI Visibility Audit" dated June 1, 2026, but do not mention a "Common Crawl blog index" or state that such an index dates June 1, 2026.
The part the quote did not support: Common Crawl blog index
What it rested on
Introducing the AI Visibility Audit June 1, 2026
Today Common Crawl is publishing The AI Visibility Audit, a free field guide built for the SEOs and GEOs who are already doing this work and want a concrete framework rather than theory.
Common Crawl's published erratum states that the truncation threshold for archived content was increased to 5 MiB from the March 2025 crawl onwards.
Why it was not published: The quote states the threshold change but does not identify it as a published erratum.
The part the quote did not support: published erratum
What it rested on
Prior to March 2025 (CC-MAIN-2025-13), the truncation threshold was 1 MiB. From the March 2025 crawl onwards, this limit has been increased to 5 MiB.
Common Crawl's FAQ states that CCBot currently does not execute JavaScript.
Why it was not published: The quote says JavaScript is not executed but does not name CCBot.
The part the quote did not support: CCBot
A Search Engine Journal article dated August 10, 2026 states that BuzzStream measured deliberate blocking of Common Crawl at 75% of the top U.S. and UK publishers.
Why it was not published: The quote says BuzzStream measured blocking of training crawlers generally at 75%, not Common Crawl specifically.
The part the quote did not support: deliberate blocking of Common Crawl
What it rested on
The BBC robots.txt blocks 13 of the 14 AI crawlers the checker tracks, and other top news sites have been blocking training crawlers on purpose since 2023. BuzzStream measured it at 75% of the top U.S. and UK publishers.
Common Crawl echoed the claim that a Search Engine Journal article dated August 10, 2026 states that since July 1 2025, Cloudflare blocks AI crawlers by default on every new domain it serves.
Why it was not published: The quote supports the Cloudflare statement from a Search Engine Journal article but does not show Common Crawl echoing it or the August 10, 2026 date.
The part the quote did not support: Common Crawl echoed the claim that a Search Engine Journal article dated August 10, 2026 states
What it rested on
Since July 1 2025, Cloudflare blocks AI crawlers by default on every new domain it serves.
A Search Engine Journal article dated August 10, 2026 restates the Mozilla figure as 64% of LLMs released between 2019 and 2023 having trained on Common Crawl data in some form.
Why it was not published: The quote restates the Mozilla figure in a Search Engine Journal article but does not establish the article's publication date.
The part the quote did not support: dated August 10, 2026
What it rested on
When Mozilla audited the LLMs released between 2019 and 2023, 64% had trained on Common Crawl data in some form, and GPT-3 was roughly 60% filtered Common Crawl by training weight.
A Search Engine Journal article dated August 10, 2026 states that GPT-3 was roughly 60% filtered Common Crawl by training weight.
Why it was not published: The quote supports the GPT-3 statistic but does not establish the article's date.
The part the quote did not support: dated August 10, 2026
What it rested on
When Mozilla audited the LLMs released between 2019 and 2023, 64% had trained on Common Crawl data in some form, and GPT-3 was roughly 60% filtered Common Crawl by training weight.
A Search Engine Journal article dated August 10, 2026 states that being in Common Crawl does not put a site in a model.
Why it was not published: The cited quote states the core sentence, but no cited quote or source label provides the August 10, 2026 date.
The part the quote did not support: dated August 10, 2026
What it rested on
Being in Common Crawl doesn’t put you in a model.
Every training set filters the crawl, and the frontier labs stopped disclosing their mixes around 2023.
A Search Engine Journal article dated August 10, 2026 states that every training set filters the crawl.
Why it was not published: The quote supports the statement and Search Engine Journal source, but no quote provides the article date.
The part the quote did not support: dated August 10, 2026
What it rested on
Being in Common Crawl doesn’t put you in a model.
Every training set filters the crawl, and the frontier labs stopped disclosing their mixes around 2023.
core update2
Google's documentation page "Google Search's core updates and your website" calls those significant, broad changes to its search algorithms and systems core updates.
Why it was not published: The cited quotes show the page title as 'Google Search's Core Updates', not 'Google Search's core updates and your website'.
The part the quote did not support: Google Search's core updates and your website
What it rested on
Several times a year, Google makes significant, broad changes to our search algorithms and systems
We refer to these as core updates
Google Search's Core Updates | Google Search Central | Documentation | Google for Developers
Amsive's analysis of Google's March 2026 core update reports that YouTube dropped 567 visibility points in less than two weeks.
Why it was not published: The quote discusses a domain at YouTube's scale, not explicitly stating that YouTube itself dropped 567 visibility points.
The part the quote did not support: YouTube
What it rested on
A drop of -567 visibility points in less than two weeks is staggering - there is essentially nothing comparable in the recent SISTRIX historical record for a domain at YouTube's scale.
Google March 2026 Core Update: Winners, Losers & Analysis | Amsive
Core Web Vitals6
Core Web Vitals, according to Google's documentation page 'Understanding page experience in Google Search results', has no single page experience signal that Google Search uses for ranking.
Why it was not published: The quotes establish that there is no single page experience signal, but they do not mention Core Web Vitals.
The part the quote did not support: Core Web Vitals
What it rested on
Is there a single "page experience signal" that Google Search uses for ranking?
There is no single signal. Our core ranking systems look at a variety of signals that align with overall page experience.
The Google Ad Manager blog post "The need for mobile speed." sources Core Web Vitals' 53% abandonment figure to aggregated, anonymized Google Analytics data.
Why it was not published: The quote identifies the data source but does not mention Core Web Vitals or a 53% abandonment figure.
The part the quote did not support: Core Web Vitals' 53% abandonment figure
What it rested on
6 Google Data, Aggregated, anonymized Google Analytics data from a sample of mWeb sites opted into sharing benchmark data, n=3.7K, Global, March 2016
The Google Analytics data for Core Web Vitals' 53% abandonment figure comes from a sample of mWeb sites opted into sharing benchmark data.
Why it was not published: The quote describes a sample of mWeb sites opted into sharing benchmark data but does not mention Core Web Vitals or the 53% abandonment figure.
The part the quote did not support: Core Web Vitals' 53% abandonment figure
What it rested on
6 Google Data, Aggregated, anonymized Google Analytics data from a sample of mWeb sites opted into sharing benchmark data, n=3.7K, Global, March 2016
The sample for Core Web Vitals' 53% abandonment figure is global, has n=3.7K, and is from March 2016.
Why it was not published: The quote states the n, global scope, and date but does not mention Core Web Vitals or a 53% abandonment figure.
The part the quote did not support: the sample for Core Web Vitals' 53% abandonment figure
What it rested on
6 Google Data, Aggregated, anonymized Google Analytics data from a sample of mWeb sites opted into sharing benchmark data, n=3.7K, Global, March 2016
Core Web Vitals has an average load time of 19 seconds for mobile sites over 3G connections, as reported by the Google Ad Manager blog post 'The need for mobile speed.'
Why it was not published: The quote states the average mobile site load time over 3G is 19 seconds, but it never mentions Core Web Vitals.
The part the quote did not support: Core Web Vitals
What it rested on
Unfortunately, based on our analysis of 10,000+ mobile web domains, we found that most mobile sites don't meet this bar: the average load time for mobile sites is 19 seconds over 3G connections. 1
Core Web Vitals' average load time is based on an analysis of 10,000+ mobile web domains, as reported by the Google Ad Manager blog post 'The need for mobile speed.'
Why it was not published: The quote mentions average mobile-site load time from 10,000+ domains but never ties it to Core Web Vitals.
The part the quote did not support: Core Web Vitals' average load time
What it rested on
Unfortunately, based on our analysis of 10,000+ mobile web domains, we found that most mobile sites don't meet this bar: the average load time for mobile sites is 19 seconds over 3G connections. 1
crawl budget3
Google migrated the Optimize your crawl budget documentation from Google Search Central to Google's crawling infrastructure site on December 18, 2025.
Why it was not published: The quotes confirm the migration and destination but do not mention the original location as Google Search Central.
The part the quote did not support: from Google Search Central
What it rested on
December 2025
December 18
What: Migrated the following documentation to Google's crawling infrastructure site. The functionality hasn't changed, only the location of the documentation and some minor wording changes to clarify that some guidance applies to both Google Search and other Google products.
Optimize your crawl budget
Backlinko states that for a website with 10,000 or more pages, crawl budget can cause Google to have trouble finding all the pages.
Why it was not published: The quote only states that Google might have trouble finding all pages on a 10k+ page site, not that crawl budget is the cause.
The part the quote did not support: crawl budget can cause
What it rested on
You run a larger site: If you have a website (like an ecommerce site) with 10k+ pages, Google can have trouble finding them all.
TwoSquares characterises the majority of websites, for which crawl budget is not a limiting factor, as those under roughly 50,000 URLs with stable architecture and reasonable performance.
Why it was not published: The quotes do not mention stable architecture or reasonable performance, so the claim adds unsupported conditions.
The part the quote did not support: with stable architecture and reasonable performance
What it rested on
The uncomfortable truth: most sites do not have a crawl budget problem
under ~50,000 URLs
Crawl budget is not a limiting factor.
Cumulative Layout Shift5
Google's "Understanding page experience in Google Search results" documentation states that Cumulative Layout Shift is used by Google's ranking systems.
Why it was not published: The quote states Core Web Vitals generally are used by ranking systems, but does not specifically state Cumulative Layout Shift is used.
The part the quote did not support: Cumulative Layout Shift
What it rested on
Core Web Vitals are used by our ranking systems.
Understanding page experience in Google Search results
The Web Almanac 2025 performance chapter states that Cumulative Layout Shift's primary analysis is based on measurements from July 2025 drawn from the HTTP Archive and the Chrome UX Report.
Why it was not published: The quotes describe the chapter's overall primary analysis, but never specify that the statement applies to Cumulative Layout Shift.
The part the quote did not support: Cumulative Layout Shift's
What it rested on
This chapter draws on data from the HTTP Archive and the Chrome UX Report (CrUX) , combining lab-based measurements with real-user performance data.
The primary analysis is based on measurements from July 2025 and spans millions of websites and a large volume of page loads across the web.
The Web Almanac 2025 performance chapter states that Cumulative Layout Shift's primary analysis spans millions of websites.
Why it was not published: The quote states the primary analysis spans millions of websites but does not tie it to Cumulative Layout Shift specifically.
The part the quote did not support: Cumulative Layout Shift's
What it rested on
This chapter draws on data from the HTTP Archive and the Chrome UX Report (CrUX) , combining lab-based measurements with real-user performance data.
The primary analysis is based on measurements from July 2025 and spans millions of websites and a large volume of page loads across the web.
Cumulative Layout Shift's good results in Core Web Vitals reports do not guarantee that a site's pages will rank at the top of Google Search results, according to Google's "Understanding page experience in Google Search results" documentation.
Why it was not published: The quote only says good results in Core Web Vitals reports don't guarantee top rankings; it does not single out Cumulative Layout Shift.
The part the quote did not support: Cumulative Layout Shift's good results
What it rested on
Keep in mind that getting good results in reports like Search Console's Core Web Vitals report or third-party tools doesn't guarantee that your pages will rank at the top of Google Search results; there's more to great page experience than Core Web Vitals scores alone.
Google's "Understanding page experience in Google Search results" documentation states that beyond Core Web Vitals (including Cumulative Layout Shift), other page experience aspects don't directly help a website rank higher in search results.
Why it was not published: The cited quote states Core Web Vitals but does not specify Cumulative Layout Shift.
The part the quote did not support: including Cumulative Layout Shift
What it rested on
Beyond Core Web Vitals, other page experience aspects don't directly help your website rank higher in search results.
Digital PR6
Digital PR was reported by Search Engine Roundtable to be one of three new topics cited in Google's updated link schemes document.
Why it was not published: The cited quote mentions 'advertorials or native advertising', not 'Digital PR'; the claim substitutes an unsupported terminology.
The part the quote did not support: Digital PR
What it rested on
In short, Google has updated their link schemes document specifically citing three new topics: (1) Large-scale article marketing or guest posting campaigns with keyword-rich anchor text links (2) Advertorials or native advertising where payment is received for articles that include links that pass PageRank (3) Links with optimized anchor text in articles or press releases distributed on other sites.
Jul 29, 2013 - 9:22 am
Digital PR's documented position is that Google's spam policies documentation lists as link spam links with optimized anchor text in articles, guest posts, or press releases distributed on other sites.
Why it was not published: The quote is from Google's documentation and does not state or establish anything about Digital PR's documented position.
The part the quote did not support: Digital PR's documented position is that
What it rested on
Advertorials or native advertising where payment is received for articles that include links that pass ranking credit, or links with optimized anchor text in articles, guest posts, or press releases distributed on other sites.
Digital PR's documented position is that Search Engine Journal reported that Google's John Mueller, asked whether brand mentions without a link help with SEO rankings, said "From my point of view, I don't think we use those at all for things like PageRank or understanding the link graph of a website."
Why it was not published: Quotes are from Search Engine Journal, not Digital PR; no mention of Digital PR's position.
The part the quote did not support: Digital PR's documented position
What it rested on
“Do brand mentions without a link help with SEO rankings?”
“From my point of view, I don’t think we use those at all for things like PageRank or understanding the link graph of a website. And just a plain mention is sometimes kind of tricky to figure out anyway.”
The Reporter Outreach digital PR statistics page cites only Reporter Outreach as the source.
Why it was not published: The quote shows one citation to Reporter Outreach but does not establish that the page cites no other sources.
The part the quote did not support: only
What it rested on
Digital PR is the lowest-risk approach because editorial decisions sit with the journalist — making campaigns effectively immune to Google algorithm penalties. Reporter Outreach
The Reporter Outreach digital PR statistics page provides no linked study for the figure.
Why it was not published: The quote provides the statistic and names Semrush but does not state whether the figure has a linked study.
The part the quote did not support: provides no linked study
What it rested on
53% of Google penalties involve paid links with keyword-rich anchor text — a risk that disappears when journalists choose their own anchors. Semrush
An Ahrefs study of 75,000 brands reports that digital PR (brand web mentions) shows a correlation of 0.664 with AI Overview brand visibility.
Why it was not published: The quotes mention 'brand web mentions' but not 'digital PR'; the claim adds an unsupported label.
The part the quote did not support: digital PR
What it rested on
we’ve analyzed 75,000 brands to see which search factors are most likely to influence brand mentions in AI Overviews
Brand web mentions show the strongest correlation (0.664) with AI Overview brand visibility.
E-E-A-T3
E-E-A-T's documented position, as stated in Google's 'Creating helpful, reliable, people-first content' documentation, is that rater data is not used directly in Google's ranking algorithms.
Why it was not published: The quotes state Google documentation says rater data is not used directly, but none mention E-E-A-T.
The part the quote did not support: E-E-A-T's documented position
What it rested on
Rater data is not used directly in our ranking algorithms.
Creating Helpful, Reliable, People-First Content | Google Search Central
Search Engine Roundtable reported that John Mueller of Google said there is no E-A-T score that Google uses.
Why it was not published: The quote only says 'John said,' and does not identify him as John Mueller of Google.
The part the quote did not support: John Mueller of Google
What it rested on
John said that while there is no E-A-T score that Google uses, E-A-T is important and is in the search ranking algorithm in an indirect way.
Feb 21, 2022
Search Engine Roundtable published a transcript in which John Mueller says he wouldn't see there being an E-A-T score that you have to get five or something on.
Why it was not published: The quote matches the E-A-T score remark but does not identify John Mueller or a transcript, only a Search Engine Roundtable page.
The part the quote did not support: published a transcript in which John Mueller says
What it rested on
But I wouldn't see like there's like an EAT score and you have to get five or something like that on it.
Expired domain abuse1
Google's spam policies documentation lists "Casino-related content on a former elementary school site" as an illustrative example of expired domain abuse.
Why it was not published: The cited quotes show it is an illustrative example in the spam policies documentation, but do not state the specific category 'expired domain abuse'.
The part the quote did not support: of expired domain abuse
What it rested on
Illustrative examples include, but are not limited to:
Casino-related content on a former elementary school site
FAQPage structured data9
Google's Search Central documentation changelog records that on September 14, 2023 the FAQ structured data documentation was updated.
Why it was not published: The quotes confirm a September 14 update to the FAQ structured data documentation but do not specify the year 2023.
The part the quote did not support: 2023
What it rested on
September 14:
Updated the FAQ structured data documentation to state that the feature is only shown for well-known, authoritative government and health websites.
Google's Search Central documentation changelog entry for FAQPage structured data is titled "Deprecating the FAQ rich result feature".
Why it was not published: The quotes establish the changelog title but do not state that the entry is for FAQPage structured data.
The part the quote did not support: for FAQPage structured data
What it rested on
May 8
Deprecating the FAQ rich result feature
Why: This feature will no longer appear in Google Search starting May 7, 2026.
Google's Search Central documentation changelog entry for FAQPage structured data, titled "Deprecating the FAQ rich result feature", is dated May 8, 2026.
Why it was not published: Quotes mention the FAQ rich result feature and a May 8 date, but none states FAQPage structured data.
The part the quote did not support: FAQPage structured data
What it rested on
May 8
Deprecating the FAQ rich result feature
Why: This feature will no longer appear in Google Search starting May 7, 2026.
Google's Search Central documentation changelog entry for FAQPage structured data, titled "Deprecating the FAQ rich result feature", states that FAQPage structured data's FAQ rich result feature will no longer appear in Google Search starting May 7, 2026.
Why it was not published: The quotes mention the FAQ rich result feature and its discontinuation date, but they never mention FAQPage structured data.
The part the quote did not support: FAQPage structured data's
What it rested on
May 8
Deprecating the FAQ rich result feature
Why: This feature will no longer appear in Google Search starting May 7, 2026.
Google's Search Central documentation changelog entry for FAQPage structured data is titled 'Deprecating the FAQ rich result feature'.
Why it was not published: The quotes show an entry titled 'Deprecating the FAQ rich result feature' but do not mention or establish 'FAQPage structured data'.
The part the quote did not support: FAQPage structured data
What it rested on
May 8
Deprecating the FAQ rich result feature
What: Added a deprecation notice to the FAQ rich result documentation.
Google's Search Central documentation changelog entry for FAQPage structured data is dated May 8, 2026.
Why it was not published: The quotes give 'May 8' and the FAQ deprecation entry but do not state the year 2026.
The part the quote did not support: May 8, 2026
What it rested on
May 8
Deprecating the FAQ rich result feature
What: Added a deprecation notice to the FAQ rich result documentation.
Google removed FAQPage structured data's documentation for the FAQ rich result feature.
Why it was not published: The quotes state documentation for the FAQ rich result feature was removed, but none mention FAQPage structured data.
The part the quote did not support: FAQPage structured data's
What it rested on
June 15
Removing documentation for the FAQ rich result feature
What: Removed documentation for the FAQ rich result feature.
The removal of FAQPage structured data's documentation for the FAQ rich result feature was recorded in Google's Search Central documentation changelog entry titled "Removing documentation for the FAQ rich result feature" dated June 15, 2026.
Why it was not published: The quotes show the title and 'June 15' but do not state the year 2026, so the dated entry as specified is not fully established.
The part the quote did not support: June 15, 2026
What it rested on
June 15
Removing documentation for the FAQ rich result feature
What: Removed documentation for the FAQ rich result feature.
FAQPage structured data is no longer shown in Google Search results, according to Google's Search Central documentation changelog.
Why it was not published: The quote says the FAQ rich result feature is no longer shown, not that FAQPage structured data itself is no longer shown.
The part the quote did not support: FAQPage structured data is no longer shown
What it rested on
Why: The FAQ rich result feature is no longer shown in Google Search results, as announced in the changelog entry in May 2026.
generative engine optimization8
Google Search Central's AI optimization guide states that creating separate content for every possible variation of how people might search violates Google's scaled content abuse spam policy.
Why it was not published: The quote only says this violates policy when done primarily to manipulate rankings or generative AI responses, so the claim's unconditional version overreaches.
The part the quote did not support: creating separate content for every possible variation of how people might search violates Google's scaled content abuse spam policy
What it rested on
While it might be tempting to create separate content for every possible variation of how people might search (for example, by focusing on other queries that people have asked, or fan-out queries), doing so primarily to manipulate rankings or generative AI responses in Google Search violates Google's scaled content abuse spam policy.
Generative engine optimization has a documented position, reported by Search Engine Roundtable, that is expressed in John Mueller of Google's Bluesky post: "FWIW no AI system currently uses llms.txt."
Why it was not published: The quote shows John Mueller's Bluesky statement as reported by Search Engine Roundtable, but it does not state or clearly entail that this is a documented position about generative engine optimization.
The part the quote did not support: Generative engine optimization has a documented position
What it rested on
Well, John Mueller of Google chimed in and wrote on Bluesky, "FWIW no AI system currently uses llms.txt."
Search Engine Roundtable quotes John Mueller of Google saying that consumer LLMs and chatbots will fetch your pages for training and grounding.
Why it was not published: The quote does not identify the speaker as John Mueller of Google; it only says "He went on to say."
The part the quote did not support: John Mueller of Google
What it rested on
He went on to say, "It's super-obvious if you look at your server logs. The consumer LLMs / chatbots (the ones that SEOs want traffic from) will fetch your pages - for training and grounding, but none of them fetch the llms.txt file. Maybe they will tomorrow? Maybe I'll win in the lottery tomorrow?"
Search Engine Roundtable quotes John Mueller of Google saying that none of the consumer LLMs and chatbots fetch the llms.txt file.
Why it was not published: The quote establishes the statement via Search Engine Roundtable but does not identify the speaker as John Mueller.
The part the quote did not support: John Mueller of Google
What it rested on
He went on to say, "It's super-obvious if you look at your server logs. The consumer LLMs / chatbots (the ones that SEOs want traffic from) will fetch your pages - for training and grounding, but none of them fetch the llms.txt file. Maybe they will tomorrow? Maybe I'll win in the lottery tomorrow?"
For generative engine optimization, Chrome for Developers' Lighthouse documentation describes llms.txt as used to provide a machine-readable summary of a website's content.
Why it was not published: The cited quote says llms.txt is designed for LLMs and AI agents, not generative engine optimization.
The part the quote did not support: For generative engine optimization
What it rested on
The llms.txt file is an emerging convention used to provide a machine-readable summary of a website's content, specifically designed for LLMs and AI agents.
Chrome for Developers' Lighthouse documentation states that, for generative engine optimization, providing an llms.txt file is optional at the moment.
Why it was not published: The quote states providing the file is optional, but does not mention generative engine optimization.
The part the quote did not support: for generative engine optimization
What it rested on
Lighthouse flags the pages if a server error occurs when attempting to retrieve the llms.txt file. If the file is not provided by the server (resulting in a 404), the audit is marked as Not Applicable (N/A), as providing the file is optional at the moment.
Search Engine Land quotes Wil Reynolds asserting that in generative engine optimization, AI is 100x more likely to force you to have receipts to back up your claims than in the old days of ranking on domain authority.
Why it was not published: The quoted assertion does not mention generative engine optimization; the claim inserts a context not stated in the quotes.
The part the quote did not support: in generative engine optimization
What it rested on
Wil Reynolds said it best:
“In the old days you could just muscle your way to the top of Google with zero receipts that you are “top 3” at XYZ thing. If you had enough domain authority you could rank. AI is 100x more likely to force you to have receipts to back up your claims.”
Search Engine Journal reports that, in generative engine optimization, Google's guidance on llms.txt is split between different Google developer sites.
Why it was not published: The cited quote does not mention generative engine optimization; it only states guidance is split between developer sites, so this contextual framing is not supported.
The part the quote did not support: in generative engine optimization
What it rested on
Guidance is split between different Google developer sites, which can lead to conflicting instructions when comparing Lighthouse or its llms.txt documentation with Google’s Search docs.
Google-Extended5
A Google blog post announced that Google was beginning to test Google-Extended, a new toggle in Search Console.
Why it was not published: The quote describes a new toggle but does not name it 'Google-Extended' or mention testing.
The part the quote did not support: beginning to test Google-Extended
What it rested on
With this new toggle in Search Console, website owners can decide if they want their site to appear in and help ground responses in our generative AI Search features (like AI Overviews, AI Mode or AI Overviews in Discover). Sites that opt out will not receive traffic or impressions from our generative AI features.
Google-Extended lets website owners decide if they want their site to appear in and help ground responses in Google's generative AI Search features such as AI Overviews and AI Mode.
Why it was not published: The quote refers to a 'new toggle in Search Console' without naming it 'Google-Extended', so the claim's specific product name is unsupported.
The part the quote did not support: Google-Extended
What it rested on
With this new toggle in Search Console, website owners can decide if they want their site to appear in and help ground responses in our generative AI Search features (like AI Overviews, AI Mode or AI Overviews in Discover). Sites that opt out will not receive traffic or impressions from our generative AI features.
Google-Extended's documented position in Google's 'AI features and your website' documentation is that robots.txt directives for Googlebot are the control for site owners to manage access to how their sites are crawled for Search.
Why it was not published: The quote states the documentation's position about robots.txt directives for Googlebot, but it never mentions Google-Extended.
The part the quote did not support: Google-Extended's documented position
What it rested on
AI is built into Search and integral to how Search functions, which is why robots.txt directives for Googlebot is the control for site owners to manage access to how their sites are crawled for Search. To limit the information shown from your pages in Search, use nosnippet, data-nosnippet, max-snippet, or noindex controls.
Google's Search Console help page for the Search generative AI control states that Google-Extended only affects whether content can appear in certain Search generative AI features.
Why it was not published: The quoted text refers only to 'this control' without naming 'Google-Extended', so the claim overreaches by inserting the specific product name.
The part the quote did not support: Google-Extended
What it rested on
This control only affects whether your content can appear in certain Search generative AI features; this control isn't used as a ranking or inclusion signal affecting other parts of Search.
Google's Search Console help page for the Search generative AI control states that Google-Extended is not used as a ranking or inclusion signal affecting other parts of Search.
Why it was not published: The quote refers to 'this control' but never explicitly names it 'Google-Extended', so the claim adds a specific name not stated.
The part the quote did not support: Google-Extended
What it rested on
This control only affects whether your content can appear in certain Search generative AI features; this control isn't used as a ranking or inclusion signal affecting other parts of Search.
Helpful content system6
The Google blog post about the Helpful content system was published on blog.google in March 2024.
Why it was not published: The cited quote discusses low-quality, unoriginal content but does not identify the Helpful content system.
The part the quote did not support: about the Helpful content system
What it rested on
Update April 26, 2024: As of April 19, we’ve completed the rollout of these changes. You’ll now see 45% less low-quality, unoriginal content in search results versus the 40% improvement we expected across this work.
The Google blog post about the Helpful content system was later updated.
Why it was not published: The cited quote shows a Google blog post was later updated, but it never identifies that post as about the Helpful content system.
The part the quote did not support: about the Helpful content system
What it rested on
Update April 26, 2024: As of April 19, we’ve completed the rollout of these changes. You’ll now see 45% less low-quality, unoriginal content in search results versus the 40% improvement we expected across this work.
The updated Google blog post about the Helpful content system stated that users would see 45% less low-quality, unoriginal content in search results, versus the 40% improvement Google had expected.
Why it was not published: The quote directly states the 45% vs 40% figures but does not identify the blog post as being about the Helpful content system.
The part the quote did not support: about the Helpful content system
What it rested on
Update April 26, 2024: As of April 19, we’ve completed the rollout of these changes. You’ll now see 45% less low-quality, unoriginal content in search results versus the 40% improvement we expected across this work.
Helpful content system echoed that a Google blog post was published on blog.google in March 2024.
Why it was not published: The quote is from a Google blog post about search updates but never mentions a 'Helpful content system' or that it echoed the publication.
The part the quote did not support: Helpful content system echoed
What it rested on
We believe these updates will reduce the amount of low-quality content on Search and send more traffic to helpful and high-quality sites. Based on our evaluations, we expect that the combination of this update and our previous efforts will collectively reduce low-quality, unoriginal content in search results by 40%.
Helpful content system echoed that the Google blog post published on blog.google in March 2024 states that Google expected the combination of the March 2024 update and its previous efforts to collectively reduce low-quality, unoriginal content in search results by 40%.
Why it was not published: The quote establishes the blog post's 40% statement, but does not show any 'Helpful content system' echoed it.
The part the quote did not support: Helpful content system echoed that
What it rested on
We believe these updates will reduce the amount of low-quality content on Search and send more traffic to helpful and high-quality sites. Based on our evaluations, we expect that the combination of this update and our previous efforts will collectively reduce low-quality, unoriginal content in search results by 40%.
The helpful content system recovery study is published on thestacc.com.
Why it was not published: The quote only mentions HCU recovery discussions and a statistic, not a study published on thestacc.com.
The part the quote did not support: study
What it rested on
Context: The 22% figure has become the most cited statistic in HCU recovery discussions. It represents the ceiling, not the floor.
hreflang2
Google's documentation on localized versions states that even without taking action on hreflang, Google might still find alternate language versions of a page.
Why it was not published: The quote says 'without taking action' but does not mention hreflang, so the claim adds an unsupported specific.
The part the quote did not support: on hreflang
What it rested on
Note that even without taking action, Google might still find alternate language versions of your page, but it is usually best for you to explicitly indicate your language- or region-specific pages.
Google's documentation on localized versions describes that hreflang helps Google Search point users to the most appropriate version of the page by language or region.
Why it was not published: The cited quote says telling Google about variations helps point users to the appropriate version, but it does not mention hreflang.
The part the quote did not support: hreflang
What it rested on
If you have multiple versions of a page for different languages or regions, tell Google about these different variations. Doing so will help Google Search point users to the most appropriate version of your page by language or region.
Interaction to Next Paint (INP)4
Google's page-experience documentation states that Interaction to Next Paint (INP) is used by Google's ranking systems.
Why it was not published: The cited quote says Core Web Vitals are used by ranking systems but never mentions Interaction to Next Paint (INP).
The part the quote did not support: Interaction to Next Paint (INP) is used by Google's ranking systems
What it rested on
Core Web Vitals are used by our ranking systems. We recommend site owners achieve good Core Web Vitals for success with Search and to ensure a great user experience generally.
A corgenx.com article about Interaction to Next Paint (INP) asserts that Google has stated that Core Web Vitals (CWV) are used as a tiebreaker when two pages have similar relevance and content quality.
Why it was not published: The quote is from a corgenx.com article but does not mention INP or indicate the article's topic is about INP.
The part the quote did not support: about Interaction to Next Paint (INP)
What it rested on
Google has stated that CWV are used as a tiebreaker when two pages have similar relevance and content quality.
Google's page-experience documentation states that getting good Interaction to Next Paint (INP) results in reports like Search Console's Core Web Vitals report or third-party tools does not guarantee that pages will rank at the top of Google Search results.
Why it was not published: The cited quote does not mention INP; it only refers to good results in Core Web Vitals reports or third-party tools.
The part the quote did not support: Interaction to Next Paint (INP)
What it rested on
Keep in mind that getting good results in reports like Search Console's Core Web Vitals report or third-party tools doesn't guarantee that your pages will rank at the top of Google Search results; there's more to great page experience than Core Web Vitals scores alone.
Google's page-experience documentation states that trying to get a perfect Core Web Vitals score, which includes Interaction to Next Paint (INP), just for SEO reasons may not be the best use of a site owner's time.
Why it was not published: The quote says only 'these scores' and does not identify them as Core Web Vitals or mention INP.
The part the quote did not support: Core Web Vitals score, which includes Interaction to Next Paint (INP)
What it rested on
These scores are meant to help you to improve your site for your users overall, and trying to get a perfect score just for SEO reasons may not be the best use of your time.
JavaScript rendering2
According to Vercel's blog post on AI crawlers, JavaScript rendering's primary data comes from monitoring nextjs.org and the Vercel network.
Why it was not published: The quote says 'Our primary data' but never links it to JavaScript rendering.
The part the quote did not support: JavaScript rendering's primary data
What it rested on
Our primary data comes from monitoring nextjs.org and the Vercel network for the past few months. To validate our findings across different technology stacks, we also analyzed two job board websites: Resume Library , built with Next.js, and CV Library , which uses a custom monolithic framework.
According to Vercel's blog post on AI crawlers, JavaScript rendering's primary data is validated against two job board websites, Resume Library and CV Library.
Why it was not published: The quote says primary data comes from nextjs.org and the Vercel network, while the two job board websites were analyzed to validate findings, not as the primary data source.
The part the quote did not support: JavaScript rendering's primary data is validated against two job board websites
What it rested on
Our primary data comes from monitoring nextjs.org and the Vercel network for the past few months. To validate our findings across different technology stacks, we also analyzed two job board websites: Resume Library , built with Next.js, and CV Library , which uses a custom monolithic framework.
Largest Contentful Paint16
The Chrome 130 transparent-text change to Largest Contentful Paint launched the week of Oct 21, 2024.
Why it was not published: The quote states a launch date but does not mention 'transparent-text change'; that specific detail is not supported.
The part the quote did not support: transparent-text change
What it rested on
Launch was the week of Oct 21, 2024, when m130 was in stable, but a limited number of early adopters might have seen this change as early as m126.
web.dev states that an element with a background image loaded using the url() function is considered for Largest Contentful Paint, as opposed to a CSS gradient.
Why it was not published: The quote only names the element type and does not state that it is considered for LCP.
The part the quote did not support: is considered for Largest Contentful Paint
What it rested on
An element with a background image loaded using the url() function, (as opposed to a CSS gradient)
web.dev's Web Vitals article states that all Core Web Vitals, of which Largest Contentful Paint is one, are first and foremost field metrics.
Why it was not published: The quote establishes that all Core Web Vitals are field metrics but does not state that Largest Contentful Paint is one of them.
The part the quote did not support: of which Largest Contentful Paint is one
What it rested on
While all of the Core Web Vitals are, first and foremost, field metrics, many of them are also measurable in the lab.
web.dev's Web Vitals article states that many Core Web Vitals, of which Largest Contentful Paint is one, are also measurable in the lab.
Why it was not published: The quote says many Core Web Vitals are measurable in the lab but does not state that Largest Contentful Paint is among them.
The part the quote did not support: of which Largest Contentful Paint is one
What it rested on
While all of the Core Web Vitals are, first and foremost, field metrics, many of them are also measurable in the lab.
web.dev's Web Vitals article states that for Largest Contentful Paint, lab measurement is an essential part of delivering great experiences.
Why it was not published: The quote states lab measurement is essential generally but does not tie it to Largest Contentful Paint.
The part the quote did not support: for Largest Contentful Paint
What it rested on
While lab measurement is an essential part of delivering great experiences, it is not a substitute for field measurement.
web.dev's Web Vitals article states that for Largest Contentful Paint, lab measurement is not a substitute for field measurement.
Why it was not published: The quote makes a general statement about lab versus field measurement but does not specifically mention Largest Contentful Paint.
The part the quote did not support: for Largest Contentful Paint
What it rested on
While lab measurement is an essential part of delivering great experiences, it is not a substitute for field measurement.
web.dev's Core Web Vitals thresholds article states that a site is classified as having good Largest Contentful Paint performance if at least 75 percent of page views meet the good Largest Contentful Paint threshold.
Why it was not published: The quote states the 75% 'good' threshold for 'that metric' but does not specify Largest Contentful Paint.
The part the quote did not support: Largest Contentful Paint
What it rested on
if at least 75 percent of page views to a site meet the "good" threshold, the site is classified as having "good" performance for that metric
The documented position of Largest Contentful Paint is that Google's Search Central documentation on page experience states that Core Web Vitals are used by Google's ranking systems.
Why it was not published: The quote mentions Core Web Vitals generally but does not mention Largest Contentful Paint or its documented position.
The part the quote did not support: documented position of Largest Contentful Paint
What it rested on
What aspects of page experience are used in ranking? Core Web Vitals are used by our ranking systems.
Largest Contentful Paint's documented position, as stated in Google's Search Central documentation on page experience, is that there is no single page experience signal used by Google Search for ranking.
Why it was not published: The quote states there is no single page experience signal but never mentions Largest Contentful Paint.
The part the quote did not support: Largest Contentful Paint's documented position
What it rested on
Is there a single "page experience signal" that Google Search uses for ranking? There is no single signal. Our core ranking systems look at a variety of signals that align with overall page experience.
web.dev's Core Web Vitals thresholds article reports that at a 4 second Largest Contentful Paint threshold, roughly 26% of phone origins would be classified as poor.
Why it was not published: The quote only says a 4 second threshold, not that it is the Largest Contentful Paint threshold.
The part the quote did not support: Largest Contentful Paint
What it rested on
For a 4 second threshold, roughly 26% of phone origins, and 21% of desktop origins, would be classified as poor.
web.dev's Core Web Vitals thresholds article reports that at a 4 second Largest Contentful Paint threshold, roughly 21% of desktop origins would be classified as poor.
Why it was not published: The quote gives the 21% desktop figure for a 4 second threshold but does not state that this threshold is Largest Contentful Paint.
The part the quote did not support: Largest Contentful Paint threshold
What it rested on
For a 4 second threshold, roughly 26% of phone origins, and 21% of desktop origins, would be classified as poor.
web.dev's Core Web Vitals thresholds article reports that the 2 second Largest Contentful Paint threshold is not consistently achievable
Why it was not published: The quote says the 2 second threshold is not consistently achievable but does not identify it as the Largest Contentful Paint threshold.
The part the quote did not support: Largest Contentful Paint
What it rested on
We find that the 1.5 and 2 second thresholds are not consistently achievable, while 2.5 seconds is consistently achievable.
web.dev's Core Web Vitals thresholds article reports that the 2.5 second Largest Contentful Paint threshold is consistently achievable
Why it was not published: The quote states 2.5 seconds is consistently achievable but does not specify that this threshold is for Largest Contentful Paint.
The part the quote did not support: Largest Contentful Paint
What it rested on
We find that the 1.5 and 2 second thresholds are not consistently achievable, while 2.5 seconds is consistently achievable.
The Chrome 83 change stopping Largest Contentful Paint measurement at input or scroll in any frame produced a roughly 40% drop at the 99th percentile across all sites that use frames.
Why it was not published: The source states this drop occurred on Android only, but the claim omits that restriction.
The part the quote did not support: produced a roughly 40% drop at the 99th percentile across all sites that use frames
What it rested on
Two bugs in the implementation of Largest Contentful Paint (LCP) were fixed in Chrome 83:
For the change that stops measuring LCP at input or scroll in any frame, we see a reduction in LCP at the higher percentiles on Android only (not desktop). Across all sites that use frames, we see the following impact.
At the 99th percentile, a ~40% drop
Google's Search Central documentation on page experience states that getting good Largest Contentful Paint results in reports like Search Console's Core Web Vitals report or third-party tools does not guarantee that pages will rank at the top of Google Search results.
Why it was not published: The quote says good results in Core Web Vitals reports or third-party tools, not specifically Largest Contentful Paint results.
The part the quote did not support: Largest Contentful Paint
What it rested on
Keep in mind that getting good results in reports like Search Console's Core Web Vitals report or third-party tools doesn't guarantee that your pages will rank at the top of Google Search results; there's more to great page experience than Core Web Vitals scores alone.
For Largest Contentful Paint, Google's Search Central documentation on page experience states that Google Search always seeks to show the most relevant content even if the page experience is sub-par.
Why it was not published: The cited quote states the general page experience principle but does not tie it to Largest Contentful Paint.
The part the quote did not support: For Largest Contentful Paint
What it rested on
Google Search always seeks to show the most relevant content, even if the page experience is sub-par.
Lighthouse2
Lighthouse echoed that Vercel's blog post is about how Google handles JavaScript.
Why it was not published: The cited quote is from Vercel and never mentions Lighthouse echoing anything.
The part the quote did not support: Lighthouse echoed that
What it rested on
Plus, page speed is linked to good user experience—with every 100ms of load time saved correlated to an 8% uptick in website conversion.
Lighthouse echoed that Vercel's blog post states that every 100ms of load time saved is correlated to an 8% uptick in website conversion.
Why it was not published: The quote is from Vercel's blog and states the conversion statistic, but no cited quote shows Lighthouse echoed it.
The part the quote did not support: Lighthouse echoed that
What it rested on
Plus, page speed is linked to good user experience—with every 100ms of load time saved correlated to an 8% uptick in website conversion.
llms.txt10
Google's Search Central changelog states that the June 2026 llms.txt clarification was made to address questions from the community.
Why it was not published: The quotes mention 'these files' and the reason, but none explicitly names llms.txt, so the specific llms.txt clarification is not established by the cited text.
The part the quote did not support: llms.txt
What it rested on
June 2026
June 15
Why: To address questions from the community and clarify that while these files aren't needed for Google Search (and won't negatively or positively impact your visibility or rankings), it's fine if you want to maintain these files for other services or systems that use them.
Google's Search Central changelog states that the June 2026 llms.txt clarification clarifies that llms.txt files aren't needed for Google Search.
Why it was not published: The quoted clarification says 'these files' but never names llms.txt.
The part the quote did not support: llms.txt
What it rested on
June 2026
June 15
Why: To address questions from the community and clarify that while these files aren't needed for Google Search (and won't negatively or positively impact your visibility or rankings), it's fine if you want to maintain these files for other services or systems that use them.
Google's Search Central changelog states that the June 2026 llms.txt clarification clarifies that llms.txt files won't negatively or positively impact visibility or rankings.
Why it was not published: The quotes establish a June 2026 clarification that 'these files' won't affect visibility or rankings, but none of the cited text names llms.txt.
The part the quote did not support: llms.txt
What it rested on
June 2026
June 15
Why: To address questions from the community and clarify that while these files aren't needed for Google Search (and won't negatively or positively impact your visibility or rankings), it's fine if you want to maintain these files for other services or systems that use them.
Google added its guide on optimizing for generative AI features on Search to llms.txt on May 15, 2026.
Why it was not published: The quotes establish the guide and date but never mention llms.txt.
The part the quote did not support: to llms.txt
What it rested on
May 2026
May 15
Adding a new guide on optimizing for generative AI features
What: Added a new guide on optimizing for generative AI features on Search. Notable new sections include guidance on the importance of providing non-commodity content, tips about providing local, shopping, image, and video content, mythbusting common "AEO/GEO" misconceptions, initial guidance on AI agents, and more information on why SEO best practices continue to be relevant for success in our generative AI features on Search.
The guide Google added to llms.txt includes mythbusting of common AEO/GEO misconceptions.
Why it was not published: The quote says Google added a guide on Search with mythbusting, but does not mention llms.txt.
The part the quote did not support: to llms.txt
What it rested on
May 2026
May 15
Adding a new guide on optimizing for generative AI features
What: Added a new guide on optimizing for generative AI features on Search. Notable new sections include guidance on the importance of providing non-commodity content, tips about providing local, shopping, image, and video content, mythbusting common "AEO/GEO" misconceptions, initial guidance on AI agents, and more information on why SEO best practices continue to be relevant for success in our generative AI features on Search.
The llms.txt proposal published at llmstxt.org was originally published in September 2024.
Why it was not published: Quote confirms original publication in September 2024, but does not mention llmstxt.org as the publication venue.
The part the quote did not support: published at llmstxt.org
What it rested on
v2 (August 2026)
The original llms.txt proposal was published in September 2024, when the idea that language models would routinely read websites was still speculative.
Version 2 of the llms.txt proposal removed the context-expansion tooling from the proposal.
Why it was not published: Quote says tooling is no longer part, but does not mention Version 2.
The part the quote did not support: Version 2
What it rested on
The context-expansion tooling is no longer part of the proposal, and with it goes the special meaning of the Optional section, which told those tools what to omit. Optional sections are still allowed, and remain a useful convention for secondary links, but they no longer carry mechanical semantics.
Search Engine Journal described the Ahrefs llms.txt log analysis with a sample size of 137,000 domains in a June 2026 news article.
Why it was not published: The quotes support SEJ reporting the 137,000-domain Ahrefs log analysis, but none gives the June 2026 date the claim asserts.
The part the quote did not support: June 2026 news article
What it rested on
Ahrefs analyzed logs from 137,000 domains and found 97% of llms.txt files got zero requests. No bots, no humans.
Ahrefs ran the numbers across 100,000 domains and found that the file is, in practice, largely ignored by the crawlers it is meant to court, a finding since echoed by other large studies showing no measurable citation advantage for sites that add one.
Search Engine Journal described the Ahrefs llms.txt log analysis with a sample size of 100,000 domains in an August 2026 column.
Why it was not published: The quotes confirm Search Engine Journal described an Ahrefs analysis across 100,000 domains, but none states or entails an August 2026 date.
The part the quote did not support: August 2026 column
What it rested on
Ahrefs analyzed logs from 137,000 domains and found 97% of llms.txt files got zero requests. No bots, no humans.
Ahrefs ran the numbers across 100,000 domains and found that the file is, in practice, largely ignored by the crawlers it is meant to court, a finding since echoed by other large studies showing no measurable citation advantage for sites that add one.
Search Engine Journal published an account of a deliberately invented file format called cats.txt being tested against the four proofs commonly cited for llms.txt.
Why it was not published: The quote shows something checked against the four proofs but does not state that cats.txt was a deliberately invented file format.
The part the quote did not support: deliberately invented file format called cats.txt
What it rested on
Then, I checked it against the exact four "proofs" people cite for llms.txt. It passed all four.
meta description3
Portent's study of meta descriptions states that in December 2017 Google increased the meta description length to 320 characters.
Why it was not published: The cited quote does not state that it is from a 'study'; it is a blog post.
The part the quote did not support: study
What it rested on
Remember back in December 2017, when Google increased the meta description length to 320 characters and then rapidly changed their minds? Looks like the 320 character display limit is still there, but only for snippets Google writes.
Google Search Central's documentation-updates changelog records that on April 20, 2026 a new section on "read more" deep links was added to the meta description snippet documentation.
Why it was not published: The quote only says "April 20" without any year, so the specific date "April 20, 2026" is not established.
The part the quote did not support: April 20, 2026
What it rested on
April 20 Added a section about "read more" deep links What: Added a new section on "read more" deep links to the snippet documentation. Why: To explain how to increase the likelihood of your content appearing with a "read more" deeplink in Google Search results.
Portent's study of meta descriptions reports a first-page rewrite rate higher than the 63% figure it attributes to Ahrefs.
Why it was not published: The quote only mentions a rewrite rate higher than Ahrefs' 63%, not that it is specifically the first-page rewrite rate.
The part the quote did not support: first-page
What it rested on
This rewrite rate is a little higher than Ahref's number (63%), and that's not due to featured snippets being included in these figures; we took them out. This difference between our overall rewrite rates could be explained by us including more keywords that are rewritten more often.
noindex1
Google announced on September 1, 2019 that it was retiring all code that handles unsupported and unpublished robots.txt rules, including noindex.
Why it was not published: The quote states the retirement of the code happens on September 1, 2019, not that Google made the announcement on that date.
The part the quote did not support: announced on September 1, 2019
What it rested on
In the interest of maintaining a healthy ecosystem and preparing for potential future open source releases, we're retiring all code that handles unsupported and unpublished rules (such as noindex) on September 1, 2019.
nosnippet5
nosnippet is documented on Google's Search Console Help page "Search generative AI control".
Why it was not published: The cited quotes do not mention 'nosnippet' or any documentation about it.
The part the quote did not support: nosnippet
What it rested on
Use the Search generative AI control to manage your site's inclusion in the following generative AI features on Google Search. We expect to update this list over time as we develop Google Search.
See and manage the Search generative AI control for your property under Settings > Search generative AI in Search Console.
nosnippet manages a site's inclusion in AI Overviews, AI Mode, and generative AI features in Google Discover.
Why it was not published: The quotes discuss a 'Search generative AI control', not nosnippet.
The part the quote did not support: nosnippet
What it rested on
Use the Search generative AI control to manage your site's inclusion in the following generative AI features on Google Search. We expect to update this list over time as we develop Google Search.
See and manage the Search generative AI control for your property under Settings > Search generative AI in Search Console.
The nosnippet meta tag is not used as a ranking signal affecting other parts of Search.
Why it was not published: The quote refers to "this control" without specifying it is the nosnippet meta tag.
The part the quote did not support: The nosnippet meta tag
What it rested on
This control only affects whether your content can appear in certain Search generative AI features; this control isn't used as a ranking or inclusion signal affecting other parts of Search.
The nosnippet meta tag is not used as an inclusion signal affecting other parts of Search.
Why it was not published: The quote refers to 'This control' without naming it, so it does not establish that the statement applies specifically to the nosnippet meta tag.
The part the quote did not support: nosnippet meta tag
What it rested on
This control only affects whether your content can appear in certain Search generative AI features; this control isn't used as a ranking or inclusion signal affecting other parts of Search.
nosnippet echoes Search Engine Journal's note that NewzDash has not published sample sizes or collection dates alongside the figures.
Why it was not published: Quotes show SEJ's statement, but no evidence that nosnippet echoed it.
The part the quote did not support: nosnippet echoes Search Engine Journal's note
What it rested on
John Shehata, CEO and founder of NewzDash and GDdash, put a number on it : “Nearly 1 in 6 U.S. trending news queries now place Top Stories inside AI Overviews.” The 15.5% rate applies to tracked results where Google displayed Top Stories, not to all queries NewzDash tracked. The UK figure is 17.46%.
NewzDash hasn't published sample sizes or collection dates alongside the figures.
robots.txt6
Google's Search Central blog states that the Robots Exclusion Protocol draft introduced a requirement that developers must parse at least the first 500 kibibytes of a robots.txt file.
Why it was not published: The quote states the parsing requirement but does not say the Robots Exclusion Protocol draft introduced it.
The part the quote did not support: the Robots Exclusion Protocol draft introduced
What it rested on
Developers must parse at least the first 500 kibibytes of a robots.txt. Defining a maximum file size ensures that connections are not open for too long, alleviating unnecessary strain on servers.
Cloudflare announced that the default of robots.txt will change to block AI crawlers unless those AI crawlers pay creators for their content.
Why it was not published: The quote states changing the default to block AI crawlers, but does not mention robots.txt specifically.
The part the quote did not support: robots.txt
What it rested on
That changes today, July 1, what we’re calling Content Independence Day. Cloudflare, along with a majority of the world's leading publishers and AI companies, is changing the default to block AI crawlers unless they pay creators for their content.
robots.txt's documented position is that subdomains are treated as separate sites with separate crawl budgets.
Why it was not published: The quote states Google's crawl-budget documentation treats subdomains as separate sites, but it is not a robots.txt documented position.
The part the quote did not support: robots.txt's documented position
What it rested on
In this context, Google's crawling infrastructure defines a site as a unique hostname. For example, https://www.example.com/ and https://code.example.com/ are treated as separate sites and have separate crawl budgets.
The robots.txt finding reported by Search Engine Journal states that Mozilla audited the LLMs released between 2019 and 2023.
Why it was not published: The cited quote states Mozilla audited those LLMs but does not mention a robots.txt finding.
The part the quote did not support: robots.txt finding
What it rested on
When Mozilla audited the LLMs released between 2019 and 2023, 64% had trained on Common Crawl data in some form, and GPT-3 was roughly 60% filtered Common Crawl by training weight.
The robots.txt finding reported by Search Engine Journal states that 64% of the audited LLMs had trained on Common Crawl data in some form.
Why it was not published: The quote supports the 64% statistic from Search Engine Journal but does not describe it as a robots.txt finding.
The part the quote did not support: robots.txt finding
What it rested on
When Mozilla audited the LLMs released between 2019 and 2023, 64% had trained on Common Crawl data in some form, and GPT-3 was roughly 60% filtered Common Crawl by training weight.
Google's crawl budget documentation states that URLs blocked by robots.txt are recrawled when the block is removed.
Why it was not published: The quote says blocked URLs will be recrawled when the block is removed, but does not state the block is by robots.txt.
The part the quote did not support: by robots.txt
What it rested on
Google won't forget a URL that it knows about, but a 404 status code is a strong signal not to crawl that URL again. Blocked URLs, however, will stay part of your crawl queue much longer, and will be recrawled when the block is removed.
scaled content abuse10
The Google blog post announcing the March 2024 Search updates was updated on April 26, 2024 to state that users would see 45% less scaled content abuse (low-quality, unoriginal content) in search results.
Why it was not published: The quote states 45% less low-quality, unoriginal content but does not call it 'scaled content abuse,' so that specific label is unsupported.
The part the quote did not support: scaled content abuse
What it rested on
Update April 26, 2024: As of April 19, we've completed the rollout of these changes. You'll now see 45% less low-quality, unoriginal content in search results versus the 40% improvement we expected across this work.
Google's 'Spam policies for Google web search' documentation states that scaled content abuse is detected through both automated systems and, as needed, human review.
Why it was not published: The quote refers generally to 'policy-violating practices' but does not specifically mention 'scaled content abuse.'
The part the quote did not support: scaled content abuse
What it rested on
We detect policy-violating practices both through automated systems and, as needed, human review that can result in a manual action .
Google's 'Spam policies for Google web search' documentation states that human review of scaled content abuse can result in a manual action.
Why it was not published: The cited quote only refers to 'policy-violating practices' broadly and does not mention scaled content abuse.
The part the quote did not support: scaled content abuse
What it rested on
We detect policy-violating practices both through automated systems and, as needed, human review that can result in a manual action .
The documented position on scaled content abuse in Google's 'Spam policies for Google web search' is that sites which violate Google's policies may rank lower in results.
Why it was not published: The quote states sites violating Google's policies may rank lower, but it does not mention scaled content abuse specifically.
The part the quote did not support: on scaled content abuse
What it rested on
Sites that violate our policies may rank lower in results or not appear in results at all.
The documented position on scaled content abuse in Google's 'Spam policies for Google web search' is that sites which violate Google's policies may not appear in results at all.
Why it was not published: The cited quote states the general consequence for violating Google's policies but does not mention or specifically document the position on scaled content abuse.
The part the quote did not support: scaled content abuse
What it rested on
Sites that violate our policies may rank lower in results or not appear in results at all.
Scaled content abuse is echoed in a Search Engine Journal article published on March 5, 2024.
Why it was not published: The quotes mention reducing low-quality, unoriginal content, but do not mention or entail the specific phrase 'scaled content abuse'.
The part the quote did not support: Scaled content abuse
What it rested on
Google March 2024 Core Update: Reducing “Unhelpful” Content By 40%
The update aims to reduce low-quality, unoriginal content in search results by 40%.
March 5, 2024
Scaled content abuse is echoed in a Search Engine Journal article headlined "Google March 2024 Core Update: Reducing "Unhelpful" Content By 40%".
Why it was not published: No cited quote mentions 'scaled content abuse'; they only discuss reducing unhelpful content.
The part the quote did not support: scaled content abuse
What it rested on
Google March 2024 Core Update: Reducing “Unhelpful” Content By 40%
The update aims to reduce low-quality, unoriginal content in search results by 40%.
March 5, 2024
Scaled content abuse is echoed in a Search Engine Journal article that restates Google's forward-looking 40% expectation about low-quality, unoriginal content as a reduction in unhelpful content.
Why it was not published: The cited quotes do not mention 'scaled content abuse'; they only discuss reducing low-quality, unoriginal content and unhelpful content.
The part the quote did not support: Scaled content abuse
What it rested on
Google March 2024 Core Update: Reducing “Unhelpful” Content By 40%
The update aims to reduce low-quality, unoriginal content in search results by 40%.
March 5, 2024
The Google blog post announcing the March 2024 Search updates states that Google expected the combination of that update and previous efforts to collectively reduce scaled content abuse (low-quality, unoriginal content) in search results by 40%.
Why it was not published: The quote states a 40% reduction in low-quality, unoriginal content, but it does not use or support the term 'scaled content abuse.'
The part the quote did not support: scaled content abuse
What it rested on
Based on our evaluations, we expect that the combination of this update and our previous efforts will collectively reduce low-quality, unoriginal content in search results by 40%.
Google's "Spam policies for Google web search" documentation instructs site owners hosting scaled content abuse material to exclude that material from Search.
Why it was not published: The quote says 'such content' but does not identify it as scaled content abuse material.
The part the quote did not support: scaled content abuse material
What it rested on
If you're hosting such content on your site, exclude it from Search .
schema.org structured data5
Google Search Central's documentation-updates changelog records that a deprecation notice was added to the schema.org structured data FAQ rich result documentation.
Why it was not published: The quote only mentions the FAQ rich result documentation, not that it is schema.org structured data.
The part the quote did not support: schema.org structured data
What it rested on
Deprecating the FAQ rich result feature What: Added a deprecation notice to the FAQ rich result documentation. Why: This feature will no longer appear in Google Search starting May 7, 2026.
Google Search Central's guide to optimizing for generative AI features states that using structured data helps with being eligible for rich results on Google Search.
Why it was not published: The quote says 'using it' helps with rich results but does not identify 'it' as structured data.
The part the quote did not support: structured data
What it rested on
However, it's a good idea to continue using it as part of your overall SEO strategy, as it helps with being eligible for rich results on Google Search.
Schema.org structured data for FAQ rich results has an average click-through-rate of 87%, according to Sixth City Marketing's page of schema markup statistics.
Why it was not published: The cited quote only says FAQ rich results have an average CTR of 87%, not that Schema.org structured data does.
The part the quote did not support: Schema.org structured data for
What it rested on
FAQ rich results have an average CTR (click-through-rate) of 87%
schema.org structured data (the FAQ schema) had sections added to the page's content before its implementation.
Why it was not published: The quote mentions adding sections to content for SEO terms, but does not reference the FAQ schema or its implementation.
The part the quote did not support: schema.org structured data (the FAQ schema) had sections added to the page's content before its implementation
What it rested on
For the phrases above that were listed as “not in text”, we looked again to Content Fusion to add sections to our content to ensure that all of the suggested terms and phrases were used within the content.
The Food Network experienced a 35% increase in visits after converting its pages to enable search features with schema.org structured data.
Why it was not published: The quote says pages were converted to enable search features but does not specify schema.org structured data.
The part the quote did not support: with schema.org structured data
What it rested on
The Food Network has converted 80% of their pages to enable search features, and has seen a 35% increase in visits.
Search Console generative AI report6
On June 16, 2025, data from AI Mode began counting towards the totals in the Search Console Performance report.
Why it was not published: The quotes support June 16 and AI Mode counting toward Search Console Performance totals but do not state the year 2025.
The part the quote did not support: 2025
What it rested on
June 16
AI Mode is now counting towards totals in Search Console
Why: Data from AI Mode is now counting towards the totals in the Search Console Performance report.
The Search Console generative AI report was introduced by Microsoft in Bing Webmaster Tools in a Bing Webmaster Blog post dated February 10 2026.
Why it was not published: The quotes introduce 'AI Performance in Bing Webmaster Tools,' not a 'Search Console generative AI report.'
The part the quote did not support: Search Console generative AI report
What it rested on
February 10 2026
Introducing AI Performance in Bing Webmaster Tools Public Preview
We are happy to introduce AI Performance in Bing Webmaster Tools, a new set of insights that shows how publisher content appears across Microsoft Copilot, AI-generated summaries in Bing, and select partner integrations.
The Search Console generative AI report was described by Microsoft as a new set of insights showing how publisher content appears across Microsoft Copilot, AI-generated summaries in Bing, and select partner integrations.
Why it was not published: The quote describes 'AI Performance in Bing Webmaster Tools,' not a 'Search Console generative AI report.'
The part the quote did not support: Search Console generative AI report
What it rested on
February 10 2026
Introducing AI Performance in Bing Webmaster Tools Public Preview
We are happy to introduce AI Performance in Bing Webmaster Tools, a new set of insights that shows how publisher content appears across Microsoft Copilot, AI-generated summaries in Bing, and select partner integrations.
Google's Search Console Help page for the generative AI performance report (Search) states that Search Console does not include data from experiments in Search Labs in the generative AI performance report.
Why it was not published: The quote does not specify that this note is from the generative AI performance report page, only that Search Console excludes data from Search Labs.
The part the quote did not support: Google's Search Console Help page for the generative AI performance report (Search)
What it rested on
Note: Search Console doesn't include data from experiments in Search Labs, as these experiments are still in active development.
Google's Search Console Help page for the generative AI performance report (Discover) states that a user who scrolls past a card and then scrolls back has only one impression recorded.
Why it was not published: The quote directly supports the scroll-back impression statement, but no cited quote or source label mentions Discover.
The part the quote did not support: (Discover)
What it rested on
Only one impression is counted per result per session; if a user scrolls past a card and then scrolls back, only one impression is recorded.
Microsoft's AI Performance in Bing Webmaster Tools reports a "Total Citations" metric, being the total number of citations displayed as sources in AI-generated answers during the selected time frame.
Why it was not published: Quotes define Total Citations for AI Performance in Bing Webmaster Tools, but none mention a Search Console report.
The part the quote did not support: Search Console generative AI report
What it rested on
Total Citations
Shows the total number of citations that are displayed as sources in AI-generated answers during the selected time frame. This highlights how often your content is referenced by AI systems, without indicating placement or presentation within a specific answer.
site reputation abuse8
Google announced its site reputation abuse spam policy on March 5, 2024.
Why it was not published: The quotes mention a 'new policy' and a March 5, 2024 date but never name it as the site reputation abuse spam policy.
The part the quote did not support: site reputation abuse spam policy
What it rested on
Tuesday, March 5, 2024
To allow time for site owners to prepare for this change, this new policy will take effect starting May 5, 2024.
Google stated its site reputation abuse spam policy would take effect starting May 5, 2024.
Why it was not published: The cited quote says a new policy takes effect May 5, 2024, but does not identify that policy as the site reputation abuse spam policy.
The part the quote did not support: site reputation abuse spam policy
What it rested on
Tuesday, March 5, 2024
To allow time for site owners to prepare for this change, this new policy will take effect starting May 5, 2024.
Google states that moving content that received a site reputation abuse manual action to a subdirectory or subdomain within the same domain does not resolve the underlying issue.
Why it was not published: The quote does not mention manual action; it only says moving content doesn't resolve the underlying issue.
The part the quote did not support: that received a site reputation abuse manual action
What it rested on
Moving content to a subdirectory or subdomain within the same site's domain name: This doesn't resolve the underlying issue and may be viewed as an attempt to circumvent our spam policy, which may lead to broader actions against a site in Google Search.
Google states that moving content that received a site reputation abuse manual action to another established site will resolve the site reputation abuse issue for the site it was removed from.
Why it was not published: The quote does not mention manual action; it generally states that moving content resolves site reputation abuse.
The part the quote did not support: that received a site reputation abuse manual action
What it rested on
Moving content to another established site: This will resolve the site reputation abuse issue for the site it was removed from, as the site reputation of that site is no longer being abused. However, it may introduce a site reputation abuse issue to the site the content is moved to if the established site has its own reputation and the third-party nature is unchanged.
Google states that moving content that received a site reputation abuse manual action to another established site may introduce a site reputation abuse issue on the site it is moved to.
Why it was not published: The quote discusses moving content with a 'site reputation abuse issue' but does not specify the issue resulted from a manual action, so the claim adds unsupported specificity.
The part the quote did not support: that received a site reputation abuse manual action
What it rested on
Moving content to another established site: This will resolve the site reputation abuse issue for the site it was removed from, as the site reputation of that site is no longer being abused. However, it may introduce a site reputation abuse issue to the site the content is moved to if the established site has its own reputation and the third-party nature is unchanged.
Sites which removed offending content or implemented noindex tags on affected sections have started seeing their site reputation abuse manual actions lifted.
Why it was not published: The quote only says 'manual actions' without specifying they are for site reputation abuse, so the claim adds unsupported specificity.
The part the quote did not support: site reputation abuse
What it rested on
The recovery process has shown clear patterns: sites that removed offending content or implemented noindex tags on affected sections have started seeing their manual actions lifted. However, ranking recovery takes time as Google's crawlers need to process these changes.
According to Search Engine Journal's report of Aleyda Solis, site reputation abuse has the practical effect that Newsweek's 'Vault' experienced a decline in visibility.
Why it was not published: The quotes mention a decline in visibility but do not link it to site reputation abuse.
The part the quote did not support: site reputation abuse has the practical effect that
What it rested on
Analysis by Aleyda Solis reveals that more publishers' review sections have seen drops in organic search traffic and rankings.
According to Solis, USA Today's "Reviewed", Newsweek's Vault, and The Sun UK's Shopping section have all experienced declines in visibility.
According to Search Engine Journal's report of Aleyda Solis, site reputation abuse has the practical effect that The Sun UK's 'Shopping' section experienced a decline in visibility.
Why it was not published: The quotes confirm the decline but do not establish that it is caused by site reputation abuse.
The part the quote did not support: site reputation abuse has the practical effect
What it rested on
Analysis by Aleyda Solis reveals that more publishers' review sections have seen drops in organic search traffic and rankings.
According to Solis, USA Today's "Reviewed", Newsweek's Vault, and The Sun UK's Shopping section have all experienced declines in visibility.
unlinked brand mentions19
Google Search Central's guide to optimizing for generative AI features states that seeking inauthentic “mentions” across the web isn't as helpful as it might seem.
Why it was not published: The quote mentions inauthentic 'mentions' across the web but does not specify unlinked or brand mentions.
The part the quote did not support: unlinked brand mentions
What it rested on
However, seeking inauthentic "mentions" across the web isn't as helpful as it might seem.
Google Search Central's guide to optimizing for generative AI features states that unlinked brand mentions - what's being said about products and services across the web - can be shown by Google's generative AI features, including in blogs, videos and forum discussions.
Why it was not published: The quote says generative AI features can show what's being said about products and services, but it does not state or entail the specific phrase 'unlinked brand mentions'.
The part the quote did not support: unlinked brand mentions
What it rested on
Just like the rest of Google Search, our generative AI features can show what's being said about products and services across the web, including in blogs, videos, and forum discussions.
Unlinked brand mentions are defined in Google's patent US8682892B1, “Ranking search results”, as implied links: a reference to a target resource, such as a citation to the target resource, included in a source resource but not an express link to the target resource.
Why it was not published: The quote defines 'implied link' but never uses or defines the term 'unlinked brand mentions.'
The part the quote did not support: Unlinked brand mentions
What it rested on
An implied link is a reference to a target resource, e.g., a citation to the target resource, which is included in a source resource but is not an express link to the target resource.
Unlinked brand mentions are documented in Google's patent US8682892B1, 'Ranking search results', which states that links for a group of resources can include express links, implied links, or both.
Why it was not published: The quote states links can include implied links but never mentions unlinked brand mentions.
The part the quote did not support: Unlinked brand mentions are documented
What it rested on
Links for the group can include express links, implied links, or both. An express link, e.g., a hyperlink, is a link that is included in a source resource that a user can follow to navigate to a target resource.
Google's patent US8682892B1, “Ranking search results”, documents the position on unlinked brand mentions by defining a reference query for a group of resources as a search query that has been submitted to a search engine.
Why it was not published: The quote defines the reference query but never mentions unlinked brand mentions, so it does not establish that the patent documents that position.
The part the quote did not support: documents the position on unlinked brand mentions
What it rested on
A reference query for a group of resources is a search query that has been submitted to a search engine and has been classified as referring to a resource in the group.
Google's patent US8682892B1, “Ranking search results”, documents the position on unlinked brand mentions by defining a reference query for a group of resources as a search query that has been classified as referring to a resource in the group.
Why it was not published: The quote defines a reference query exactly, but does not mention unlinked brand mentions or tie the definition to that position.
The part the quote did not support: documents the position on unlinked brand mentions
What it rested on
A reference query for a group of resources is a search query that has been submitted to a search engine and has been classified as referring to a resource in the group.
The finding about unlinked brand mentions is that 92% of sites that experience significant organic growth produce their own proprietary assets.
Why it was not published: The cited quote states the 92% finding but does not mention unlinked brand mentions or tie the finding to that topic.
The part the quote did not support: about unlinked brand mentions
What it rested on
Cyrus Shepard found in a recent study that 92% of sites that experience significant organic growth produce their own proprietary assets.
The finding about unlinked brand mentions was made by Cyrus Shepard in a recent study.
Why it was not published: The quote describes a finding about proprietary assets, not about unlinked brand mentions.
The part the quote did not support: unlinked brand mentions
What it rested on
Cyrus Shepard found in a recent study that 92% of sites that experience significant organic growth produce their own proprietary assets.
Unlinked brand mentions echo the claim made by Search Engine Land's article on the traditional link building model that links still matter because without a certain baseline level of authority competitive SEO thresholds become hard to clear.
Why it was not published: The quoted evidence only establishes what Search Engine Land's article says, not that unlinked brand mentions echo that claim.
The part the quote did not support: Unlinked brand mentions echo the claim made by
What it rested on
Yes. Links still matter because, without a certain baseline level of authority, competitive SEO thresholds become hard to clear.
Similarweb's study 'The Downstream Impact of AI Visibility' reports that across Finance, Travel and Beauty, users who received an AI recommendation were 2.5 times more likely to visit the recommended brand's website in the seven days that followed.
Why it was not published: The quote says 'all three industries' but does not name Finance, Travel and Beauty.
The part the quote did not support: Finance, Travel and Beauty
What it rested on
Across all three industries, users who received an AI recommendation were 2.5 times more likely to visit that brand’s website in the seven days that followed.
Similarweb's study 'The Downstream Impact of AI Visibility' on unlinked brand mentions describes its method as tracking thousands of real user journeys across Finance, Travel and Beauty.
Why it was not published: The quote supports the tracking method but never mentions unlinked brand mentions.
The part the quote did not support: on unlinked brand mentions
What it rested on
To prove it, we tracked thousands of real user journeys across Finance, Travel, and Beauty. We followed users who asked ChatGPT an industry-relevant question, received a brand recommendation, and then watched what they did for the next seven days.
Similarweb's study 'The Downstream Impact of AI Visibility' on unlinked brand mentions follows users who asked ChatGPT an industry-relevant question and received a brand recommendation.
Why it was not published: The quote shows the study followed users who asked ChatGPT and received a brand recommendation, but does not mention unlinked brand mentions or the exact study title.
The part the quote did not support: on unlinked brand mentions
What it rested on
To prove it, we tracked thousands of real user journeys across Finance, Travel, and Beauty. We followed users who asked ChatGPT an industry-relevant question, received a brand recommendation, and then watched what they did for the next seven days.
For unlinked brand mentions, Similarweb's study “The Downstream Impact of AI Visibility” reports that search accounts for around 40% of traffic for standard visits.
Why it was not published: The quote states the 40% figure for standard visits generally but does not specify it applies to unlinked brand mentions.
The part the quote did not support: For unlinked brand mentions
What it rested on
For standard visits, search accounts for around 40% of traffic. For AI-influenced visits, it’s closer to 56%. AI is essentially priming people to search for brands by name.
SparkToro's post on the Similarweb research states that the study of unlinked brand mentions looks at three consumer industries: finance, travel and beauty.
Why it was not published: The quote states the study looks at finance, travel, and beauty but does not describe it as a study of unlinked brand mentions.
The part the quote did not support: unlinked brand mentions
What it rested on
The study looks at three consumer industries: finance, travel, beauty, and analyzes AI prompts that showed particular brands in these spaces and what happened on devices that saw those recommendations.
SparkToro's post on the Similarweb research states that the study of unlinked brand mentions analyzes AI prompts that showed particular brands in these spaces.
Why it was not published: The quote states the study analyzes AI prompts showing particular brands but does not say it is a study of unlinked brand mentions.
The part the quote did not support: unlinked brand mentions
What it rested on
The study looks at three consumer industries: finance, travel, beauty, and analyzes AI prompts that showed particular brands in these spaces and what happened on devices that saw those recommendations.
SparkToro's post on the Similarweb research states that the study of unlinked brand mentions analyzes what happened on devices that saw those recommendations.
Why it was not published: The quote says the study analyzed AI prompts that showed particular brands, not 'unlinked brand mentions'.
The part the quote did not support: unlinked brand mentions
What it rested on
The study looks at three consumer industries: finance, travel, beauty, and analyzes AI prompts that showed particular brands in these spaces and what happened on devices that saw those recommendations.
SparkToro's post on the Similarweb research states that the data on unlinked brand mentions is aggregate data.
Why it was not published: The quote says the data is aggregate but does not specify that it is data on unlinked brand mentions.
The part the quote did not support: unlinked brand mentions
What it rested on
I will say that my last bit of advice if you’re looking at this is once again, this is aggregate data. They are looking at some specific brands, big brands that you’ve probably heard of, Sephora, American Express, Capital One, right in spaces you’re probably familiar with.
SparkToro's post on the Similarweb research states that the data on unlinked brand mentions covers specific large brands including Sephora, American Express and Capital One.
Why it was not published: The quote mentions aggregate data and specific big brands but never states unlinked brand mentions.
The part the quote did not support: data on unlinked brand mentions
What it rested on
I will say that my last bit of advice if you’re looking at this is once again, this is aggregate data. They are looking at some specific brands, big brands that you’ve probably heard of, Sephora, American Express, Capital One, right in spaces you’re probably familiar with.
None of the unlinked brand mentions measured by Similarweb's study shows up properly in a site's analytics yet.
Why it was not published: The quote says only 'None of this shows up,' without specifying unlinked brand mentions or a Similarweb study.
The part the quote did not support: unlinked brand mentions measured by Similarweb's study