<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>SEO Madman - SEO reports</title>
    <link>https://seomadman.com</link>
    <description>Recurring data reports on search volatility, AI answer coverage and index behaviour.</description>
    <language>en</language>
    <atom:link href="https://seomadman.com/reports/rss.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Canonical tag adoption, 2025: 68% of desktop pages set one</title>
      <link>https://seomadman.com/reports/canonical-tag-adoption-web-almanac-2025</link>
      <guid isPermaLink="true">https://seomadman.com/reports/canonical-tag-adoption-web-almanac-2025</guid>
      <pubDate>Fri, 18 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>The 2025 Web Almanac SEO chapter puts rel=canonical on 68% of desktop and 67% of mobile pages, up from 65%, with mismatches under 1%.</description>
      <content:encoded><![CDATA[HTTP Archive's 2025 Web Almanac found canonical tags on 68% of desktop and 67% of mobile pages in its July 2025 crawl, up from 65% a year earlier. Almost all sit in the HTML head: HTTP header canonicals appear on only 0.58% of desktop sites, and canonical mismatches affect just 0.72% of them.

- Canonical tag usage rose to 68% of desktop pages and 67% of mobile pages in 2025, up from 65% on both in 2024.
- Pages that other pages treat as the canonical version, called canonicalized, rose to 7% on desktop and 9% on mobile.
- HTTP header canonicals remain rare, found on 0.58% of desktop sites and 0.50% of mobile sites, against 64.40% and 64.27% for raw HTML canonicals.
- Canonical mismatches were detected on 0.72% of desktop sites and 0.71% of mobile sites, a small drop from 2024.
- Rendered-change canonicals, where JavaScript alters the tag after load, affect 2.71% of desktop sites and 3.02% of mobile sites.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
      <category>indexing</category>
    </item>
    <item>
      <title>AI training now drives 52% of crawler requests, Cloudflare says</title>
      <link>https://seomadman.com/reports/cloudflare-ai-training-crawler-share-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/cloudflare-ai-training-crawler-share-2026</guid>
      <pubDate>Fri, 18 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Cloudflare reports AI training made up 52% of identified crawler requests in June 2026, up from 22% in Spring 2025, with search crawling shrinking.</description>
      <content:encoded><![CDATA[Cloudflare reports that 52% of crawler requests it identifies by purpose were for AI training as of June 2026, up from 22% in Spring 2025. Mixed-use crawlers add over 36% more. Pure search crawling, the kind that can send readers back to a site, is now a small and declining share of crawler activity.

- Cloudflare says 52% of crawler requests were for AI training as of June 2026, up from 22% in Spring 2025.
- Mixed-use crawlers, which blend search, agent use and training, account for over 36% of crawler activity.
- Cloudflare describes pure search crawling as a small and declining share, despite remaining critical for publisher visibility.
- More than one-third of crawler activity still comes from mixed-use bots, so purpose labels cannot fully separate training from search.
- Cloudflare defines its crawl-to-refer ratio as HTML requests from a platform's crawlers divided by HTML requests referred from that platform, normalized to one referral.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>Where desktop Google searches end: Datos Q4 2025 data</title>
      <link>https://seomadman.com/reports/desktop-search-click-distribution-datos-q4-2025</link>
      <guid isPermaLink="true">https://seomadman.com/reports/desktop-search-click-distribution-datos-q4-2025</guid>
      <pubDate>Fri, 18 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Datos and SparkToro&apos;s Q4 2025 clickstream shows 24.5% of US desktop Google searches ended with no click in December, versus 25.1% a year earlier.</description>
      <content:encoded><![CDATA[Datos and SparkToro's Q4 2025 report found that 24.5% of US Google desktop searches ended with no click in December 2025, against 25.1% in December 2024. Organic clicks reached 42.0% and searches staying inside Google-owned properties 16.6%. The report describes zero-click behavior as stabilizing on desktop rather than accelerating.

- Datos found that 24.5% of US Google desktop searches ended with no click in December 2025, down slightly from 25.1% in December 2024.
- In the same December 2025 data, 42.0% of US searches ended in an organic click on a non-Google result and 2.1% in a paid click.
- Another 16.6% of US searches ended on a Google-owned property such as maps, weather or hotels, up from 14.5% in December 2024.
- The report states that zero-click behavior fluctuated but did not accelerate in Q4 2025, and that its data covers desktop only.
- Fewer Google searches per desktop searcher is the report's larger shift, which SparkToro's Rand Fishkin puts at a nearly 20% decline in the US from 2024 to 2025.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>INP in 2025: 77% of mobile sites pass, but the top 1,000 lag at 63%</title>
      <link>https://seomadman.com/reports/inp-good-scores-by-site-rank-2025</link>
      <guid isPermaLink="true">https://seomadman.com/reports/inp-good-scores-by-site-rank-2025</guid>
      <pubDate>Fri, 18 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>The 2025 Web Almanac&apos;s INP data: 77% of mobile sites have good INP, popular sites trail at 63%, and mobile home pages beat inner pages 80% to 69%.</description>
      <content:encoded><![CDATA[About 77% of mobile websites had good Interaction to Next Paint in the 2025 Web Almanac, up from 74% in 2024. Popular sites trail: the top 1,000 reached 63%, up from 53%. Desktop stood at 97%. On mobile, home pages hit 80% good INP while secondary pages fell to 69%, an 11 point gap.

- Mobile INP improved to 77% of websites with good scores in 2025, up from 74% in 2024, while desktop held at 97%.
- The top 1,000 websites gained the most, rising from 53% to 63% good INP, yet they still sit 14 points below the overall mobile average.
- Mobile home pages reached 80% good INP, a 7 point gain, while secondary pages slipped to 69%, opening an 11 point gap that did not exist in 2024.
- Lab Total Blocking Time worsened by 58% even as field INP improved, so lab scores alone no longer predict what real users experience.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>llms.txt fetch counts: what AI crawlers requested on 83 sites</title>
      <link>https://seomadman.com/reports/llms-txt-fetch-counts-by-ai-crawler</link>
      <guid isPermaLink="true">https://seomadman.com/reports/llms-txt-fetch-counts-by-ai-crawler</guid>
      <pubDate>Fri, 18 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>EZY Research&apos;s 12-week server-log panel shows OpenAI and Anthropic crawlers fetching llms.txt in single digits while pulling robots.txt thousands of times.</description>
      <content:encoded><![CDATA[Server logs from 83 websites over 12 weeks show major AI crawlers barely touch llms.txt: OpenAI's crawler fetched it 7 times against 3,990 robots.txt fetches, Anthropic's 9 against 3,120, and PerplexityBot never. Meta-ExternalAgent is the exception at 193 llms.txt fetches, more than its 172 robots.txt fetches.

- OpenAI's crawler fetched llms.txt 7 times across 83 sites in 12 weeks, against 3,990 robots.txt fetches from the same crawler.
- Anthropic's ClaudeBot fetched llms.txt 9 times versus 3,120 robots.txt fetches, and PerplexityBot fetched it zero times versus 775.
- Googlebot requested llms.txt 67 times, roughly a hundredth of the 5,125 times it requested robots.txt.
- Meta-ExternalAgent is the one crawler that fetched llms.txt more than robots.txt, with 193 llms.txt requests against 172.
- The panel skews toward small-business sites whose owners sought an AI-visibility tool, so the counts describe that population and not the whole web.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>How much AI Overview citations overlap with the top 10</title>
      <link>https://seomadman.com/reports/ai-overview-citation-source-overlap</link>
      <guid isPermaLink="true">https://seomadman.com/reports/ai-overview-citation-source-overlap</guid>
      <pubDate>Wed, 16 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>BrightEdge&apos;s own year-long tracking: only 17% of AI Overview citations also rank in the organic top 10, flat since last year, with wide swings by industry.</description>
      <content:encoded><![CDATA[BrightEdge tracked AI Overview citations against organic rankings across nine industries from February 2025 to February 2026. Only about 17% of cited sources also rank in the organic top 10, a figure that has barely moved in a year. Healthcare and B2B tech sit near 23%, while restaurants and finance sit under 12%.

- BrightEdge found only about 17% of sources cited in AI Overviews also rank in the organic top 10, roughly unchanged from a year earlier.
- Five out of six AI Overview citations pull from content that does not appear on page 1 of traditional organic results, according to the same tracking.
- Top-10 overlap ranges from 24.0% in healthcare down to 9.3% in restaurants, a gap of more than 14 percentage points across the nine industries BrightEdge tracks.
- Entertainment, travel and e-commerce saw the largest gains in overlap over the year, rising from single digits to the mid-teens even as the overall average barely changed.
- AI Overview presence itself grew from roughly 30% to 48% of BrightEdge's tracked queries over the same twelve months, a 58% increase.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>AI referral traffic to publishers doubled in a year, Similarweb finds</title>
      <link>https://seomadman.com/reports/ai-referral-traffic-by-industry-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/ai-referral-traffic-by-industry-2026</guid>
      <pubDate>Wed, 16 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Similarweb measured 770.7 million monthly AI referral visits worldwide, up 117.4% year over year, led by marketplaces and news.</description>
      <content:encoded><![CDATA[Similarweb's traffic panel measured 770.7 million average monthly visits sent from AI platforms to websites worldwide between June 2025 and May 2026, up 117.4% year over year. ChatGPT sent over 80% of that referral traffic to the top 1,000 global domains. Marketplaces led every tracked industry at 46.8 million monthly visits, growing 237.3%, with news and travel tied just behind.

- Similarweb's worldwide traffic panel measured 770.7 million average monthly AI referral visits between June 2025 and May 2026, more than double the prior year's figure.
- ChatGPT accounted for over 80% of AI referral traffic to the top 1,000 global domains, per Similarweb, with Gemini, Claude, Perplexity and Grok splitting the rest.
- Marketplaces received the most AI referral traffic of any tracked industry, at 46.8 million average monthly visits, up 237.3% year over year.
- News and travel sites tied for second, each receiving 44.5 million average monthly visits, both up 115.6% year over year.
- Beauty had the smallest tracked AI referral volume at 2.0 million monthly visits, despite the fastest relative growth of any industry at 312.5%.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>Core Web Vitals by CMS, 2025: Duda and Wix lead, WordPress lags</title>
      <link>https://seomadman.com/reports/core-web-vitals-by-cms-platform-2025</link>
      <guid isPermaLink="true">https://seomadman.com/reports/core-web-vitals-by-cms-platform-2025</guid>
      <pubDate>Wed, 16 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>The 2025 Web Almanac&apos;s CMS chapter breaks Core Web Vitals pass rates down by platform: Duda at 85%, Wix up 14 points to 74%, WordPress flat at 45%.</description>
      <content:encoded><![CDATA[HTTP Archive's 2025 Web Almanac measured Core Web Vitals pass rates for ten CMS platforms using July 2025 CrUX field data. Duda led at 85% of mobile sites passing, TYPO3 followed at 79%, and Wix jumped 14 points to 74%. WordPress stayed near the bottom at 45%, and Weebly was the only platform to decline, slipping to 47%.

- Duda had the highest mobile Core Web Vitals pass rate of any CMS platform in the 2025 Web Almanac, at 85%, followed by TYPO3 CMS at 79%.
- Wix posted the largest year-over-year gain of any platform, rising 14 percentage points from 55% to 74% of mobile sites passing all three metrics.
- WordPress passed Core Web Vitals on only 45% of mobile sites, among the lowest of the ten platforms measured, though it still improved roughly 4 points year over year.
- Weebly was the only platform in the study to lose ground, declining to a 47% pass rate.
- The gap is widest on INP: Squarespace led at 96% while 1C-Bitrix and Weebly tied for last at 71%, a 25-point spread the Almanac calls unsolved.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>WebSite leads schema.org type adoption at 12.73% of mobile pages</title>
      <link>https://seomadman.com/reports/schema-org-type-adoption-web-almanac-2024</link>
      <guid isPermaLink="true">https://seomadman.com/reports/schema-org-type-adoption-web-almanac-2024</guid>
      <pubDate>Wed, 16 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>The 2024 Web Almanac breaks JSON-LD down by schema.org type: WebSite tops out at 12.73% of mobile pages, Organization at 7.16%, Product at just 0.77%.</description>
      <content:encoded><![CDATA[HTTP Archive's 2024 Web Almanac structured-data chapter broke JSON-LD adoption down by individual schema.org type across its June 2024 crawl of nearly 17 million websites. WebSite led at 12.73% of mobile pages, followed by Organization at 7.16% and BreadcrumbList at 5.66%, while Product reached only 0.77% and Article just 0.18%.

- The 2024 Web Almanac's structured-data chapter found WebSite the most common JSON-LD type, appearing on 12.73% of mobile pages.
- Organization followed at 7.16%, BreadcrumbList at 5.66%, and LocalBusiness at 3.97%, all well ahead of content and commerce types.
- Product structured data reached only 0.77% of mobile pages and Article just 0.18%, despite both types being central to Google's rich-result eligibility.
- FAQPage adoption rose slightly on desktop, from 0.2% in 2022 to 0.6% in 2024, even after Google removed FAQ rich results from search in August 2023.
- These per-type figures come from HTTP Archive's June 2024 crawl of 16,935,953 websites, the same dataset behind the chapter's headline 41% overall JSON-LD adoption rate.]]></content:encoded>
      <category>reports</category>
      <category>structured-data</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>Google SERP volatility by category, mid-September 2026</title>
      <link>https://seomadman.com/reports/serp-volatility-by-category-september-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/serp-volatility-by-category-september-2026</guid>
      <pubDate>Wed, 16 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Semrush Sensor&apos;s live index reads 2.1 out of 10 for mid-September 2026, but News and Finance run far hotter than Science or Health.</description>
      <content:encoded><![CDATA[Semrush Sensor put Google's overall SERP volatility at 2.1 out of 10, inside its normal range, on September 16, 2026. That average hides a wide spread: News sat at 5.0, Finance at 4.1 and Food & Drink at 3.7, while Science and Health held at 0.8, showing a calm headline number can mask a rough day in one vertical.

- Semrush Sensor's all-categories score read 2.1 out of 10 on September 16, 2026, which the tool itself labels normal range.
- The News category scored 5.0 on the same day, more than double the all-categories figure and the highest of the 25 categories shown.
- Finance scored 4.1 and Food & Drink scored 3.7, both well above the sitewide average despite the calm headline number.
- Science and Health both scored 0.8, the lowest readings on the page, showing volatility is not evenly distributed across verticals.
- Semrush's own methodology note says the score comes from a proprietary formula over a fixed daily keyword set, not a disclosed weighting anyone can reproduce.]]></content:encoded>
      <category>reports</category>
      <category>core-updates</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>Every confirmed Google ranking update since December 2025</title>
      <link>https://seomadman.com/reports/google-ranking-updates-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/google-ranking-updates-2026</guid>
      <pubDate>Thu, 10 Sep 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Eight confirmed updates, with Google&apos;s own start dates and rollout durations, and what the pattern says about when to judge a decline in your data.</description>
      <content:encoded><![CDATA[Google has confirmed eight ranking updates since December 2025, six of them in 2026. Core updates in that set averaged 14 days to roll out and spam updates 8 days, so a judgement made in the first week of any of them was made before the rollout that caused the movement had finished.

- Google's Search Status Dashboard lists eight confirmed ranking updates between 11 December 2025 and 18 August 2026.
- Core updates in that set took an average of 14 days to complete, ranging from 11 days 21 hours to 18 days 2 hours.
- Spam updates were faster but far less consistent, from 19 hours 30 minutes in March 2026 to 26 days 15 hours in August 2025.
- The February 2026 Discover update ran 21 days 17 hours, the longest rollout of the period, and it targeted Discover rather than web results.
- No core update has been confirmed since 21 May 2026, the longest gap between core updates in this record.]]></content:encoded>
      <category>reports</category>
      <category>core-updates</category>
      <category>search-console</category>
    </item>
    <item>
      <title>Why four 2026 &apos;ChatGPT market share&apos; numbers do not agree</title>
      <link>https://seomadman.com/reports/ai-chatbot-market-share-trackers-compared</link>
      <guid isPermaLink="true">https://seomadman.com/reports/ai-chatbot-market-share-trackers-compared</guid>
      <pubDate>Sun, 19 Jul 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>StatCounter, Similarweb, Sensor Tower and OpenAI itself all published a 2026 ChatGPT usage figure. None of the four is measuring the same thing.</description>
      <content:encoded><![CDATA[Four sources published a ChatGPT usage number in 2026: StatCounter's 76.85% is chatbot referral share, Similarweb's sub-60% figure is web-visit share, Sensor Tower's 1 billion is app installs, and OpenAI's own 900 million is self-disclosed weekly active users. Each is real and correctly reported. None can be averaged against the others, because none measures the same population.

- StatCounter puts ChatGPT at 76.85% of worldwide AI chatbot referral traffic in April 2026, down from 84.21% a year earlier, based on referral logs from over 1 million websites.
- Similarweb's web-traffic panel shows ChatGPT's share of generative AI site visits falling from 73.3% in August 2025 to as low as 52.7% in May 2026 as Gemini rose past 27%.
- Sensor Tower estimates ChatGPT's app crossed 1 billion monthly active users in May 2026, the fastest app ever to reach that mark, but this is a third-party estimate.
- OpenAI itself disclosed 900 million weekly active users in a February 2026 blog post, the only number here that comes directly from the company rather than a tracker.
- None of these four figures describes the same population, so none can be added, averaged, or substituted for another in a single "AI market share" claim.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>How three CTR trackers reported three different position-1 rates</title>
      <link>https://seomadman.com/reports/organic-ctr-by-position-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/organic-ctr-by-position-2026</guid>
      <pubDate>Sat, 11 Jul 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Three trackers measured 2026 organic CTR by position and reported position 1 anywhere from 27.6% to 39.8%, because none used the same method.</description>
      <content:encoded><![CDATA[Three independent CTR trackers published 2026 position-1 organic click-through rates that disagree by 12 points: Backlinko found 27.6% from 4 million search results, First Page Sage's meta-analysis put it at 39.8%, and Advanced Web Ranking's continuous tracker reports quarter-over-quarter shifts rather than a fixed number at all. The gap is method, not disagreement about reality.

- Backlinko's analysis of 4 million Google search results found a 27.6% click-through rate at position 1, with the top three results capturing 54.4% of all clicks.
- First Page Sage's 2026 meta-analysis of multiple CTR datasets reports position 1 at 39.8%, falling to 18.7% at position 2 and 1.6% by position 10.
- Advanced Web Ranking's continuously updated tracker reports a quarter-over-quarter percentage-point shift across 22 industries, not one fixed baseline curve.
- Position-1 estimates from the three trackers span 27.6% to 39.8%, a 12-point gap driven by different data and formats, not contradictory measurements.
- Ahrefs' own by-position breakdown shows AI Overviews cut click-through rate by 58% at position 1, 50.8% at position 2, and 46.4% at position 3.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>How five llms.txt trackers reported five different adoption rates</title>
      <link>https://seomadman.com/reports/llms-txt-adoption-tracker-comparison</link>
      <guid isPermaLink="true">https://seomadman.com/reports/llms-txt-adoption-tracker-comparison</guid>
      <pubDate>Wed, 08 Jul 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Five independent trackers measured llms.txt adoption in 2026 and reported numbers from 7.4% to 28%, because none counted the same thing.</description>
      <content:encoded><![CDATA[Five trackers published llms.txt adoption numbers in 2026, and none of them agree: Rankability found 8.7% of the top 1,000 sites, Ahrefs found 28% of its customer base, SE Ranking found 10.13% across 300,000 domains, and a Majestic Million scan found 7.4%. The gap is method, not disagreement about reality.

- Five independent trackers measured llms.txt adoption in 2026 and produced results between 7.4% and 28%, because none of them measured the same population.
- Rankability found 8.7% of the top 1,000 Tranco-ranked sites serve a valid file, rising to 15.8% among the 549 domains it could actually reach.
- Ahrefs found 28% adoption across 137,210 customer domains, but its own report cautions that Ahrefs Web Analytics customers skew more technical than the web at large.
- SE Ranking's scan of nearly 300,000 domains found 10.13% adoption and no measurable correlation between the file and AI citation frequency.
- Originality.ai's monitoring of over 3 million sites recorded an 8.8x rise in llms.txt files, from 4,088 in June 2025 to 36,120 by May 2026.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>How many confirmed Google updates ship a year, 2015 to 2026</title>
      <link>https://seomadman.com/reports/confirmed-update-frequency-2015-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/confirmed-update-frequency-2015-2026</guid>
      <pubDate>Tue, 07 Jul 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>87 confirmed major Google updates since 2015, counted year by year, next to Google&apos;s own figure for total changes, and why the two must stay separate.</description>
      <content:encoded><![CDATA[Ahrefs' tracker counts 87 confirmed major Google updates between 2015 and year-to-date 2026, averaging six to nine a year with a 2021 spike of fourteen. Google itself reports over 4,000 total Search improvements in 2023 alone - a figure for undisclosed daily tweaks, not confirmed major updates, and the two must never be quoted as the same thing.

- Ahrefs' Google Algorithm Updates History lists 87 confirmed major updates from 2015 through year-to-date 2026.
- The busiest year on record is 2021 with 14 confirmed updates, driven by spam, page experience and product review updates layered on top of two core updates.
- Google's own "How Search Works" page reports 700,000-plus experiments and 4,000-plus improvements to Search in 2023 - a total-changes figure, not a confirmed-update count.
- The confirmed count for 2023 was 9 major updates, meaning Google's own 4,000-plus figure for that year describes roughly 450 times more changes than the confirmed-update count.
- 2026 has recorded 6 confirmed updates through 18 August, on pace with the 2015-2020 average after the higher-volume 2021-2022 period.]]></content:encoded>
      <category>reports</category>
      <category>core-updates</category>
    </item>
    <item>
      <title>Mobile-first indexing, from 2016 test to 2023 completion</title>
      <link>https://seomadman.com/reports/mobile-first-indexing-completion-status</link>
      <guid isPermaLink="true">https://seomadman.com/reports/mobile-first-indexing-completion-status</guid>
      <pubDate>Mon, 06 Jul 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Tracing Google&apos;s mobile-first indexing announcements from 2016 to the 2023 completion notice, and its documentation on mobile-desktop content parity.</description>
      <content:encoded><![CDATA[Google tested mobile-first indexing from November 2016 and announced the rollout complete on 31 October 2023. Its current documentation says most Googlebot crawl requests now use the smartphone crawler, and warns that a mobile page with less content than its desktop equivalent loses that content from the index entirely.

- Google first announced it was testing mobile-first indexing on 4 November 2016, on the Google Webmaster Central Blog.
- The site-by-site rollout ran publicly from March 2018 through repeated deadline extensions in 2019 and 2020.
- Google announced the transition was complete on 31 October 2023, in a post titled "Mobile-first indexing has landed".
- Google's current Googlebot documentation states the majority of crawl requests now use the smartphone crawler, and a minority use the desktop crawler.
- Google warns that a mobile page with less content than its desktop equivalent has that content excluded from indexing, not merely deprioritized.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
      <category>indexing</category>
    </item>
    <item>
      <title>How often AI Overviews actually appear, January 2025 to April 2026</title>
      <link>https://seomadman.com/reports/ai-overviews-serp-prevalence-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/ai-overviews-serp-prevalence-2026</guid>
      <pubDate>Sat, 04 Jul 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Semrush&apos;s own tracking: 6.49% of keywords in January 2025, a July peak near 24.61%, then 15.69% in November - the real trend, not a snapshot.</description>
      <content:encoded><![CDATA[Semrush found AI Overviews rose from 6.49% of tracked keywords in January 2025 to a peak near 24.61% in July, then eased to 15.69% by November. A separate Semrush study of commercial-intent keywords found AI Overview appearance grew a further 71% on average through April 2026, with jobs and education queries exceeding 60%.

- Semrush found AI Overviews appeared on 6.49% of its 10 million-plus tracked keywords in January 2025.
- That share peaked near 24.61% in July 2025 before easing back to 15.69% by November 2025, according to Semrush.
- A follow-up Semrush study of 600,000-plus keywords found commercial-intent AI Overview appearance grew 71% on average between November 2025 and April 2026.
- By April 2026, more than 60% of high-CPC jobs and education keywords in that Semrush panel triggered an AI Overview.
- Conductor measured a separate 21.9 million-query sample at 25.11% prevalence for a four-week window in September and October 2025, well above Semrush's nearby figure.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>How Google&apos;s scaled-content spam enforcement moved through 2026</title>
      <link>https://seomadman.com/reports/spam-policy-enforcement-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/spam-policy-enforcement-2026</guid>
      <pubDate>Thu, 02 Jul 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>From the May 2026 spam-policy rewrite to the August spam update, tracing enforcement against mass-produced, low-value content across 2026.</description>
      <content:encoded><![CDATA[Google's spam policy defines scaled content abuse by outcome, not production method: mass pages that fail to help users, however they were made. Google revised the policy on 15 May 2026 to cover manipulation of AI Overviews and AI Mode, then ran spam updates in June and August that analyst Glenn Gabe documented hitting programmatic and AI-generated content sites.

- Google's spam-policies page defines scaled content abuse as pages generated "for the primary purpose of manipulating search rankings and not helping users," whatever produced them.
- Google revised its spam policy on 15 May 2026 to cover manipulating generative AI responses, extending scaled content abuse to AI Overviews and AI Mode.
- Google confirmed two spam updates in 2026, one in June and one in August, the second running roughly two and a half days per the Search Status Dashboard.
- Analyst Glenn Gabe of GSQi documented four blinded case studies from the August update, including a 1.5-million-URL site and a site that lost over 200,000 queries.
- Google publishes no official count of pages or sites actioned for spam; every traffic figure here comes from third-party analysis of visible ranking movement.]]></content:encoded>
      <category>reports</category>
      <category>core-updates</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>What Google&apos;s own Discover documentation actually requires</title>
      <link>https://seomadman.com/reports/discover-technical-eligibility-requirements</link>
      <guid isPermaLink="true">https://seomadman.com/reports/discover-technical-eligibility-requirements</guid>
      <pubDate>Tue, 30 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Google&apos;s Discover eligibility page sets an exact image-size floor and a named list of content policies. Here is what it says, word for word.</description>
      <content:encoded><![CDATA[Google's own Discover documentation requires images at least 1200 pixels wide with more than 300,000 total pixels, paired with the max-image-preview:large meta setting, to be eligible for large-image treatment. It also names clickbait, sensationalism, and withheld context as reasons Discover will not recommend a page, and states no special structured data is required.

- Google's Discover documentation sets an image floor of at least 1200 pixels wide with more than 300,000 total pixels, shown alongside a 16x9 example.
- That image size only matters combined with the max-image-preview:large setting; without it, Discover will not use a large image treatment at all.
- Google names clickbait and sensationalism specifically, describing tactics that use misleading or exaggerated details or that cater to morbid curiosity, titillation, or outrage.
- Discover will withhold recommendations from content that hides crucial information needed to understand what it is about, even if the page is otherwise eligible.
- Google's page states no special structured data is required for eligibility; schema.org markup or an og:image tag only controls which image gets picked.]]></content:encoded>
      <category>reports</category>
      <category>content-strategy</category>
    </item>
    <item>
      <title>What Google has actually said about Google Lens search volume</title>
      <link>https://seomadman.com/reports/google-lens-visual-search-growth</link>
      <guid isPermaLink="true">https://seomadman.com/reports/google-lens-visual-search-growth</guid>
      <pubDate>Tue, 30 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Google&apos;s own dated numbers: 12 billion Lens searches a month in June 2023, nearly 20 billion by October 2024 - and why the 22%/26% claims don&apos;t check out.</description>
      <content:encoded><![CDATA[Google has disclosed Google Lens's own monthly search volume several times since 2021: more than 12 billion by June 2023, nearly 20 billion by October 2024, and a further 65% year-over-year jump reported at Google I/O 2025. Widely repeated claims that "22%" or "26%" of Google queries are image-based trace to no Google source and are excluded here.

- Google's own blog said Lens handled more than 12 billion visual searches a month in June 2023, up roughly 4x from about 3 billion in 2021.
- By October 2024, Google's blog put the figure at nearly 20 billion visual searches every month, a further 67% climb in about sixteen months.
- At I/O 2025, Pichai said Lens grew 65% year over year with 100 billion-plus visual searches that year, and Google's blog separately cited 1.5 billion monthly Lens users.
- No blog.google post or named Google statement supports the widely repeated claims that Google Images drives 22% of all web searches or that 26% of Google queries are image-based.
- Google discloses these figures itself at events and in blog posts without ever publishing the measurement methodology or panel behind any of them.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>How much link equity does a 301 redirect actually pass?</title>
      <link>https://seomadman.com/reports/redirect-link-equity-retention-estimates</link>
      <guid isPermaLink="true">https://seomadman.com/reports/redirect-link-equity-retention-estimates</guid>
      <pubDate>Mon, 29 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Google has never published a percentage for how much link equity a 301 redirect passes. The 90-99% figure in circulation comes from somewhere else.</description>
      <content:encoded><![CDATA[No source, including Google, has published a measured figure for how much link equity a 301 redirect passes. Google's statements, from Matt Cutts in 2013 to Gary Illyes in 2016, say redirects lose "no" or "identical" equity to a link, without a number. Moz's widely quoted 90-99% range is an estimate, not a study, and Semrush avoids a figure entirely.

- Google has never published a specific percentage for how much link equity a 301 redirect passes, in any statement traced for this report.
- Matt Cutts said in 2013 that PageRank lost through a 301 is "currently identical" to what dissipates through a normal link, without quantifying either figure.
- Gary Illyes tweeted in July 2016 that "30x redirects don't lose PageRank anymore," a stronger claim than Cutts' 2013 statement, still with no number attached.
- Moz's redirection guide states a 301 "passes between 90-99% of link equity," the most specific figure in circulation, but does not publish a study or crawl behind it.
- Semrush's own 301 redirect guide says search engines transfer "almost all" ranking power, choosing the same unquantified language Google itself uses.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>How often Google rewrites a meta description: three studies compared</title>
      <link>https://seomadman.com/reports/meta-description-rewrite-rate-compared</link>
      <guid isPermaLink="true">https://seomadman.com/reports/meta-description-rewrite-rate-compared</guid>
      <pubDate>Fri, 19 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Portent, Ahrefs and Search Engine Journal each measured how often Google rewrites a meta description in 2020: 71%, 62.78% and over 70%, weeks apart.</description>
      <content:encoded><![CDATA[Two studies measured this six weeks apart in 2020. Portent found Google rewrites meta descriptions 71% of the time on mobile and 68% on desktop, across 30,000 keywords. Ahrefs found 62.78%, across 20,000 keywords and 192,656 pages. Search Engine Journal's "over 70%" headline is coverage of Portent's figure. The gap is method and device split, not a real disagreement.

- Portent's September 2020 study of 30,000 keywords found Google rewrites meta descriptions 71% of the time on mobile and 68% on desktop, via STAT Search Analytics.
- Ahrefs' October 2020 study of 20,000 keywords and 192,656 unique pages found a 62.78% rewrite rate, splitting to 59.65% for fat-head keywords and 65.62% for long-tail keywords.
- Search Engine Journal's "over 70% of the time" headline is coverage of Portent's own mobile figure, not a third independent dataset.
- Ahrefs found Google displays a page's own meta description in the snippet only 37.22% of the time overall, even before counting partial rewrites.
- All three figures come from within six weeks of each other in 2020, so the 62.78% to 71% spread reflects different sample composition and device handling, not a trend over time.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>AI crawler block rate: how often a 403 actually stops the bot</title>
      <link>https://seomadman.com/reports/ai-crawler-blocking-rate-trend-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/ai-crawler-blocking-rate-trend-2026</guid>
      <pubDate>Wed, 17 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Live HTTP data from Cloudflare&apos;s network shows AI crawler 403 rates rising through 2026, distinct from counting robots.txt disallow rules.</description>
      <content:encoded><![CDATA[Across Cloudflare's network, AI crawler requests returning HTTP 403 Forbidden ran at 7.9% in the first half of 2026, with overall 4xx responses to crawlers climbing from 10.2% in Q2 2025 to 35.8% by Q2 2026. That is a live response-code measurement, distinct from counting how many robots.txt files merely name a crawler in a disallow rule.

- In H1 2026, 7.9% of verified AI crawler requests on Cloudflare's network returned HTTP 403 Forbidden, plus 0.7% rate-limited with 429.
- Overall 4xx responses to crawler requests rose from 10.2% in Q2 2025 to 35.8% in Q2 2026, while clean 200 OK responses fell from 80.5% to 49.2% over the same span.
- A disallow rule and a 403 measure different things: a Q1 2026 scan of 4,047 robots.txt files found GPTBot named in 5.52% of disallow rules, ahead of CCBot and ClaudeBot.
- Cloudflare made blocking AI crawlers the default for new domains from 1 July 2025, and says customers blocked over 416 billion AI bot requests in the policy's first five months.
- A robots.txt disallow is a request a crawler can ignore, while a 403 is enforced at the edge regardless of whether the bot ever reads the rules file.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>AI citation rate by content format: what checks out</title>
      <link>https://seomadman.com/reports/ai-citation-rate-by-content-format</link>
      <guid isPermaLink="true">https://seomadman.com/reports/ai-citation-rate-by-content-format</guid>
      <pubDate>Tue, 16 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Five precise stats on AI citation rates by content format circulate online. Two trace to real studies. Three do not survive verification.</description>
      <content:encoded><![CDATA[Vendor blogs circulate exact-sounding AI citation rates by content format, but most do not trace to a real source. A real Wix study of 1,056,727 citations found listicles at 21.9%, articles at 16.7%, product pages at 13.7%. A real DeltaV Digital panel found listicles at 61% in B2B tech specifically. Three other widely repeated figures could not be verified anywhere.

- Wix's AI Search Lab analyzed 1,056,727 citations, finding listicles at 21.9%, articles at 16.7%, and product pages at 13.7% of all citations.
- DeltaV Digital tracked 25,337 citations across eight industries in 2026, finding listicles at 61% of citations in B2B tech specifically, not as a universal figure.
- A real academic paper found a 17.3% citation-rate gain from structural optimization across six engines, but an industry blog presents that finding as its own research.
- A widely repeated claim that original-data pages get a 71% citation rate, comparison pages 64%, and listicles 61% does not appear on any source it is attributed to.
- A '156% higher selection rate' figure turned up attached to three different, unrelated claims across three vendor sites, each citing a different, unfetchable source.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>The page experience update: the real 2021-2022 rollout timeline</title>
      <link>https://seomadman.com/reports/page-experience-update-timeline</link>
      <guid isPermaLink="true">https://seomadman.com/reports/page-experience-update-timeline</guid>
      <pubDate>Sat, 13 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Google&apos;s own dated posts trace the page experience rollout from 2020 to its 2022 desktop finish, and what the 2026 docs now drop.</description>
      <content:encoded><![CDATA[Google previewed page experience ranking in May 2020, set a May 2021 launch date that November, then delayed the mobile rollout to mid-June 2021, completing it by the end of August 2021. Desktop followed from February to the end of March 2022. Google's current documentation now treats Core Web Vitals as the only page experience component that directly affects ranking.

- Google first previewed the page experience ranking change on 28 May 2020, then set a May 2021 launch date in a follow-up post that November.
- A 19 April 2021 post delayed the mobile rollout to "mid-June this year," with a later update confirming it was complete by the end of August 2021.
- Desktop rollout followed a separate timeline - begin February 2022, complete by the end of March 2022 - per Google's 4 November 2021 announcement.
- An August 2021 update clarified that Safe Browsing was never used as a page experience ranking signal, despite appearing in a similarly named Search Console report.
- Google's current page experience documentation states that beyond Core Web Vitals, the other original components no longer "directly" affect ranking, only user satisfaction.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>AI Overview traffic decline: seven claims, four hold up</title>
      <link>https://seomadman.com/reports/ai-overview-traffic-decline-estimates-compared</link>
      <guid isPermaLink="true">https://seomadman.com/reports/ai-overview-traffic-decline-estimates-compared</guid>
      <pubDate>Tue, 09 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Studies put the AI Overview traffic hit from 24% to 80%. We traced seven quoted figures to source; four hold up, and they measure different things.</description>
      <content:encoded><![CDATA[Independent measurements of how much AI Overviews cut publisher traffic disagree sharply: a randomized field experiment found outbound clicks fell 39.8%, an agency's client-data analysis found organic click-through rate down 61%, and disclosed company figures range from Chegg's 24% revenue decline to a disputed HubSpot estimate of 70 to 80%. None of these figures measure the same thing.

- The first randomized field experiment on AI Overviews, posted to SSRN in April 2026, found outbound organic clicks fell 39.8% when an AI Overview appeared.
- Seer Interactive's client-data analysis, updated in September 2025, found organic click-through rate on AI-Overview queries fell 61%, from 1.76% in June 2024 to 0.61%.
- A Reuters Institute report published in January 2026, using Chartbeat data from over 2,500 publisher sites, found global Google referral traffic down about a third year over year.
- Chegg disclosed a 24% year-over-year revenue decline in its own February 2025 earnings report and has since sued Google, naming AI Overviews as a cause.
- A widely repeated claim that HubSpot lost 70 to 80% of its blog traffic traces to third-party SEO-tool estimates that HubSpot's own blog post disputes as overstated.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
    <item>
      <title>AMP&apos;s traffic record: five publishers, five different outcomes</title>
      <link>https://seomadman.com/reports/amp-traffic-impact-cases-compared</link>
      <guid isPermaLink="true">https://seomadman.com/reports/amp-traffic-impact-cases-compared</guid>
      <pubDate>Fri, 05 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Independent.ie gained 55% mobile traffic on AMP. Kinsta lost 59% of its leads. Here is what really happened across five documented, named case studies.</description>
      <content:encoded><![CDATA[Five named AMP case studies disagree on traffic. Independent.ie gained roughly 55% mobile search traffic; Belfast Telegraph, under the same publisher, saw none. Chartbeat and the Daily Beast found only 1 in 3 of 159 publishers had a clear gain. Kinsta lost 59% of mobile leads. Search Engine Land and Future plc later dropped AMP with no traffic loss.

- Independent.ie reported roughly 55% higher mobile search traffic from AMP, but its own Head of SEO credited a first-mover advantage, not AMP itself.
- Chartbeat and the Daily Beast analyzed 159 publishers and found only 34% (1 in 3) showed a statistically clear traffic increase from AMP, averaging +22% with wide variance.
- Kinsta, a lead-generation site rather than a publisher, saw mobile leads fall 59% after two months on AMP and reversed the rollout.
- Belfast Telegraph, launching AMP in 2019 under the same publisher group as Independent.ie, measured no traffic change but a 49% drop in mobile digital ad revenue.
- Search Engine Land and Future plc removed AMP in 2021 and 2022 and reported no traffic loss; Future's audience director said monetization improved.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>Schema.org&apos;s vocabulary hit 823 types with version 30.0 in 2026</title>
      <link>https://seomadman.com/reports/schema-org-vocabulary-growth-2026</link>
      <guid isPermaLink="true">https://seomadman.com/reports/schema-org-vocabulary-growth-2026</guid>
      <pubDate>Thu, 04 Jun 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Schema.org&apos;s own release notes put the vocabulary at 823 types and 1,529 properties as of version 30.0, alongside a new Google-backed usage dataset.</description>
      <content:encoded><![CDATA[Schema.org's own documentation lists the vocabulary at 823 Types, 1,529 Properties, 19 Datatypes, 96 Enumerations and 535 Enumeration members as of version 30.0, released 19 March 2026. In June 2026, Google and the Schema.org community published a companion usage-statistics dataset covering term adoption across millions of domains, updated monthly on GitHub.

- Schema.org's schemas.html page states the vocabulary currently consists of 823 Types, 1529 Properties, 19 Datatypes, 96 Enumerations and 535 Enumeration members.
- Version 30.0, released 19 March 2026, added equivalence annotations and EU Digital Product Passport examples on top of the existing vocabulary.
- On 4 June 2026, Schema.org's blog announced a usage-statistics dataset built with Google, published in CSV and JSON on the project's GitHub repository and updated monthly.
- The dataset reports term usage across millions of domains at a high level, not a count of any single domain's markup, and schema.org's post names no specific domain total.
- This measures vocabulary size and term-level usage tracking, a different question from what share of pages carry structured data at all.]]></content:encoded>
      <category>reports</category>
      <category>structured-data</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>Google&apos;s review-fraud scale, and its April 2026 policy tightening</title>
      <link>https://seomadman.com/reports/google-review-fraud-enforcement-scale</link>
      <guid isPermaLink="true">https://seomadman.com/reports/google-review-fraud-enforcement-scale</guid>
      <pubDate>Fri, 17 Apr 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Google&apos;s own Maps blog states 2025 review-fraud enforcement scale, and its April 2026 update bans staff review quotas and naming employees in reviews.</description>
      <content:encoded><![CDATA[Google's own Maps trust and safety report states it blocked or removed over 292 million policy-violating reviews and 13 million fake Business Profiles in 2025. On 16 and 17 April 2026 Google published an automated spam-response protocol and a Rating Manipulation policy update banning staff review quotas and soliciting reviews that name an employee.

- Google's own blog post states it blocked or removed over 292 million policy-violating reviews and over 13 million fake Business Profiles in 2025.
- These are Google's self-reported totals, with no independent audit behind them; treat them as an attributed claim, not a verified count.
- On 16 April 2026, Google documented its automated response to a review-spam spike: remove fake content, pause new reviews, alert the owner, and show a notification banner.
- Google's Rating Manipulation policy now bans merchants directing staff to solicit a set number of reviews, or reviews naming a specific staff member.
- Google still permits a business to ask a genuine customer for a review; what changed is the ban on quotas and on coaching staff to seed employee names into the request.]]></content:encoded>
      <category>reports</category>
      <category>technical-seo</category>
    </item>
    <item>
      <title>Penske Media v. Google: the antitrust timeline so far</title>
      <link>https://seomadman.com/reports/penske-media-google-antitrust-timeline</link>
      <guid isPermaLink="true">https://seomadman.com/reports/penske-media-google-antitrust-timeline</guid>
      <pubDate>Thu, 12 Feb 2026 00:00:00 GMT</pubDate>
      <dc:creator>Adam Hafez</dc:creator>
      <description>Penske Media sued Google over AI Overviews in 2025. The dated record: the complaint, Google&apos;s dismissal motion, Penske&apos;s opposition, and an EU complaint.</description>
      <content:encoded><![CDATA[Penske Media sued Google for antitrust violations on 12 September 2025, alleging Google coerces publishers into supplying content for AI Overviews without fair compensation. Penske's 12 February 2026 opposition to Google's motion to dismiss cites click-through declines of up to 34.5 percent. The European Publishers Council filed a related EU complaint on 10 February 2026.

- Penske Media, publisher of Rolling Stone, Variety, Deadline and Billboard, sued Google in US federal court on 12 September 2025.
- The complaint frames Google's conduct as coercing publishers into supplying content that Google republishes in AI-generated answers without permission.
- Google moved to dismiss in January 2026, calling Penske's claims legally defective in every way.
- Penske's 12 February 2026 opposition cites a click-through decline of up to 34.5 percent when an AI Overview appears, and a third of its affiliate revenue lost since 2024.
- The European Publishers Council filed a separate EU complaint on 10 February 2026, alleging an Article 102 TFEU abuse of dominance over AI Overviews and AI Mode.]]></content:encoded>
      <category>reports</category>
      <category>ai-search</category>
    </item>
  </channel>
</rss>
