Welcome to the 1st edition of the AI SEO Newsletter.
Every week, we curate the latest AI SEO news, studies, tactics, and tools that can shape your strategy.
Search & AI Updates
-
Third-party rank trackers are struggling as Google steps up blocking with IP bans, more CAPTCHAs and "poisoned" results: Nozzle found 8% of SERPs came back with swapped rankings. Tools including Ahrefs, Semrush, Sistrix and SE Ranking were affected, so double-check any sudden ranking drop before reacting.
Read More → -
Google rolled out its fourth spam update of 2026 on September 24 — global, all languages, up to a two-week rollout window. No detail yet on exactly what it targets.
Read More → -
Google is testing paying publishers for content that shapes Gemini, AI Overviews and AI Mode answers — an invite-only pilot with earnings tracked in a new Search Console widget.
Read More → -
Google's redesigned "intelligent" Search box, powered by Gemini 3.5 Flash, is now live for everyone — it takes longer queries plus images, files and video, and expands dynamically as you type.
Read More → -
Google is starting to show AI Overviews for a ton of branded queries. Reddit, Salesforce, Amazon, Adobe + many more. 93% of the brands tested have AI Overviews appearing.
Read More → -
OpenAI rolled out GPT-6 to every ChatGPT user on October 7, together with "Intelligent UI": answers can now include buttons, forms, charts and interactive tools, not just text. Paid plans get GPT-6 Sol; Free and Go users get GPT-6 Luna.
Read More → -
Google launched a UGC Fresh Data Program on October 8, giving approved forums and social platforms a direct pipeline to get fresh posts into Search faster than crawling. One more sign that community content is getting priority in Google's results.
Read More → -
Google quietly added UTM parameters to outgoing links so that site owners and SEO will be able to more accurately attribute referrals from Gemini. The move comes amid increasing frustration at Google for collapsing the amount of traffic sent by Search because of AI Search and Gemini.
Read More → -
Firefox has rolled out Perplexity as a built-in default search engine option globally, putting an AI answer engine alongside Google and Bing in one of the world's most-used browsers.
Read More → -
Advertisers can now bring more of their existing first-party audience data into ChatGPT Ads through an expanded partnership between LiveRamp and OpenAI.
Read More →
Worth Reading
-
How does AI search work?
AI search engines do not simply retrieve a webpage and repeat its contents. They can rewrite a user's question into multiple searches, gather a pool of webpages, assess potential sources, and cite only a subset in the final answer.
Read More → -
When Google stops sending clicks, what still works?
Rand Fishkin and SparkToro's research: 68% of US Google searches now end without a click, up 7.5 points in 24 months — the fastest acceleration in a decade. His answer: stop optimizing for traffic, start building influence where buyers already spend time.
Read More → -
The State of AI Search Optimization 2026
Kevin Indig's Growth Memo breakdown of what's actually moving the needle in AI search right now, from one of the most-read newsletters at the intersection of SEO and AI.
Read More → -
11% of AI Overview claims aren't backed by their own cited sources
A Washington University study decomposed 98,020 AI Overview claims across 55,393 queries and found 1 in 9 unsupported by the pages Google itself cites for them — mostly omission, not outright contradiction.
Read More → -
ChatGPT cut its Reddit citations by 88% in six months
Ahrefs studied 66.6 million ChatGPT responses: only 14.4% triggered a web search, and when it did search, it increasingly went straight to official brand sites (75.9% of its most common site: searches). Original sources are winning over aggregators.
Read More → -
Why Gemini cites someone else for your query
Suganthan Mohanadasan ran the same query through Gemini five times: it drew on 22 different pages, and only 4 appeared every time. His page covered just 6 of the 40 claims Gemini made, which is the gap that decides who gets cited.
Read More → -
How to plan 2027 when traffic is down and AI won't attribute
Kevin Indig and Amanda Johnson on budgeting when around 68% of Google searches end without a click: split spend across proven work, capped experiments with kill dates, and brand bets, then judge them on AI citations, branded search and revenue together.
Read More →
Our Latest Insights
We ran 278 B2B SaaS buyer prompts through ChatGPT across six categories, then analyzed what made it cite and recommend one brand over another. Here's what stood out:
Read the Full Study →-
66% of ChatGPT recommendations happen without a citation
ChatGPT named a brand in its answer without ever citing that brand's own website two-thirds of the time — citations and recommendations are separate games.
-
Listicle presence beats having your own site cited
Brands featured in 40%+ of their category's listicles were named in 47% of ChatGPT answers, versus just 6% for brands absent from them — a stronger correlation (+0.64) than being cited directly (+0.46).
-
Domain authority still predicts ChatGPT citations
42.6% of ChatGPT citations go to pages with an 80+ Domain Rating, and DR 80+ domains get cited 2.1x more often than domains under DR 40 — classic SEO fundamentals still matter for AI visibility.
-
Fresh content wins, but refreshed old content works too
Among cited pages with a visible date, the median is 3.9 months old and 69.7% were published within the last year — but of citations to pages 2+ years old, 38.6% went to ones refreshed in the last 12 months.
-
Cited pages name more entities, not just more words
Cited pages carry about 10 named entities per 1,000 words versus 8 for pages that rank but aren't cited. Pages naming 6+ brands averaged 2.13 citations versus 1.21 for pages naming none.
Tools Worth Testing
Have any questions or suggestions? Email us at hello@position.digital.
See you next week.
Cheers,
The Position Digital team