Every Saturday I pull the week's best SEO thinking from the people I rate, and add my take. This week: ranking volatility ran hot with no confirmed core update, Alphabet took AI Mode ads global while conceding they earn less than search ads, AI Overviews passed 80 per cent of searches, and a ChatGPT tracking bug quietly buried a week of referral data under Direct. Seven items worth your time.
Google and the Algorithm
October volatility, decoded (seovendor.co). October closed with tracker spikes on 24 and 27 October and heavy forum chatter, but still no confirmed core update. The month review separates the signals: the September 2026 spam update did not finish until 8 October, and Google's 1 October AI content guidance revision quietly added Search Quality Rater language. My take: when a spam tail overlaps a guidance change, it is easy to misread one noisy month as one algorithm event. Annotate both dates before you diagnose a client dip.
Alphabet takes AI Mode ads global (finance.yahoo.com). On the Q3 call, Alphabet announced ads in AI Mode will expand to international markets, in a quarter that brought USD 127.6 billion in revenue, up 25 per cent year on year, while search ad growth kept slowing. Philipp Schindler also conceded AI Mode ads carry lower monetisation than traditional search ads. My take: scaling a worse-monetising format internationally says the traffic shift is real enough that Google cannot wait for parity. If AI Mode matters to your funnel, expect the ad load to rise before the economics do.
AI Search
AI Overviews pass 80 per cent of searches (airankingskool.com). Data from Semrush, SE Ranking and seoClarity puts AI Overviews on more than 80 per cent of Google searches, up from about 50 per cent in June 2025 and roughly 15 per cent the year before. The same write-up cautions the three trackers sit about 20 points apart because each counts a different keyword set. My take: the direction is settled and the decimal point is not. Pull your own query set before you quote 80 per cent in a board deck; your share could sit well above or below it.
Cyrus Shepard's AI opt-out data (seroundtable.com). Cyrus used Google's generative AI opt-out controls in Search Console to pull his site out of AI Overviews and AI Mode, shared the numbers, and the headline figure is a 0.9 per cent click-through rate on AI surfaces. My take: the interesting part is not the rate, it is that site-level AI controls now exist and publishers are running real experiments with them. Watch what happens to his traffic over a full quarter before you copy the move.
Technical
The ChatGPT tracking bug that ate referral data (searchengineland.com). When GPT-6 rolled out in October, links in ChatGPT answers stopped carrying the utm_source=chatgpt.com tag, so analytics tools filed those visits under Direct. OpenAI confirmed it was a bug and the fix rolled out in stages from 15 October; most cited answers carried the tag again within days, but the lost attribution is not coming back. My take: annotate 9 to 16 October in GA4 before anyone labels AI referrals a failure, then check your Direct landing pages for that window. That is where the traffic went.
Mueller on why cheap AI tactics age badly (seroundtable.com). John Mueller, interviewed by David Iwanow, was asked about listicle strategies aimed at AI search and said the systems eventually catch up: people find a trick, everyone copies it, then it becomes easy to recognise and the effort gets thrown out. The interview also covers crawl stats, redirect chains and ecommerce product data. My take: nothing here is new, which is the point. It is keyword stuffing with a fresh coat of paint, and the fix is the one that has always worked: build something search engines want to show for years, not weeks.
Strategy and Data
Adobe put a number on the ChatGPT dip (b2the7.com). Adobe's Cedric Huesler checked about 500 pages where ChatGPT normally drives at least 5 per cent of visits and compared 9 to 13 October with the week before: ChatGPT-tagged traffic fell while Direct rose by a matching amount. My take: this is how you audit a platform change. Isolate the pages that depend on the channel, compare the exact window, publish the method. Steal the approach for your next AI traffic scare.
Source Links
https://seovendor.co/google-algorithm-updates-october-2026/
https://finance.yahoo.com/technology/ai/articles/google-expands-ai-search-ads-214837473.html
https://airankingskool.com/post/ai-overview-tracking/
https://www.seroundtable.com/google-ai-overviews-mode-ctr-42421.html
https://searchengineland.com/chatgpt-appears-to-drop-utm-tracking-making-ai-referral-traffic-harder-to-measure-496552
https://www.seroundtable.com/john-mueller-google-systems-ai-spam-42420.html
https://www.b2the7.com/news-blog/marketing-trends-october-26-2026