Between October 23 and 30, 2026, the clearest signal was that search demand, search referrals and search revenue are now moving separately. Google reported its fastest query growth in five years while Reddit reported losing more than a million Google referrals a week, and Adobe found that AI-referred retail visits earned 83% more revenue per visit. Google also expanded ads in AI Mode to ten more countries, tightened its guidance on indexing user-generated pages, and confirmed no ranking update, while rank trackers admitted to bad data.
Published · Covers to · Sources checked · By Asfand Yar Junejo, publisher asfandjunejo.com
This brief only includes items that could be traced to a primary source or to the company that made the claim. Each one carries a label: Confirmed for official statements and filings, Strongly supported for published research with a stated method, and Directional for observations that point somewhere but prove little on their own.
What changed in SEO and AI search this week?
Ten developments were verified for October 23 to 30, 2026; the five with the most direct effect on SEO work are listed first.
| Development | Date | Evidence |
|---|---|---|
| Google updates its guide on keeping user-generated pages out of Search | Oct 26 | Confirmed Google documentation |
| Ads in AI Mode expand to ten more countries; multiturn ads launch in the US | Oct 28 | Confirmed Google announcement |
| Alphabet: Search revenue up 12%, fastest query growth in five years | Oct 28 | Confirmed SEC filing |
| Reddit: Google referrals down more than 1 million a week in October | Oct 29 | Confirmed earnings call, company-reported |
| Semrush confirms "poisoned" Google results in automated rank tracking | Oct 27 | Confirmed vendor statement |
| Adobe: AI-referred retail visits earn 83% more revenue per visit | Oct 29 | Strongly supported first-party data, US retail only |
| Kevin Indig: organic visits down 13.2% (US) and 24.4% (France) over two years | Oct 28 | Strongly supported modelled traffic estimates |
| Findrix: 57% of AI Mode product-answer links point to Google's own pages | Oct 29 | Strongly supported links shown, not clicks |
| Ranking volatility reported on Oct 24 and 27; no Google update confirmed | Oct 24–27 | Directional tracker chatter only |
| Gemini 4 Carbon, GPT-6.1 Terra and Microsoft's Copilot harness released | Oct 26–28 | Confirmed vendor announcements |
What changed in Google Search this week?
One documentation change with practical weight, on user-generated pages, and no confirmed ranking update.
Which user-generated pages should stay out of Google's index?
Pages made for direct sharing, collaboration or utility, such as shared chats, generated reports or user-built mini apps, should be kept out of Search by default, according to Google's updated guide.
Google revised Prevent user-generated spam on your site on October 26 and added a section titled "Only index content meant for Search". The guide now also covers tools that let users publish generated pages and web apps, not only classic forums and hosting platforms, and it points site owners to Search Console to watch for unexpected jumps in indexed pages, impressions or unrelated queries [1].
Use noindex (a robots meta tag or an X-Robots-Tag header) when a page must stay reachable by link but out of Search. A robots.txt block alone does not guarantee that: Google can still index a blocked URL without reading it if other pages link to it [2].
Confirmed Google Search Central page, last updated October 26, 2026, read on October 31
Was there a Google ranking update?
No update was confirmed. Rank trackers showed volatility on October 24 and again around October 27, but Google's Search Status Dashboard lists no incident between October 23 and 30.
Search Engine Roundtable reported heated tracker readings and forum chatter on both dates [3]. The most recent incident on the dashboard is the September 2026 spam update, which ended around October 8 [4]. Given the rank-tracker data problems described below, check any drop against your own Search Console clicks and impressions before calling it an update.
Directional Volatility reports come from third-party trackers, not from Google
What happened in AI Mode, and why are referrals falling while searches grow?
Google is putting ads into more AI Mode conversations, Alphabet reported growing search demand, and Reddit reported shrinking Google referrals: together they show that more searching no longer means more clicks to websites.
Where do ads in AI Mode run now?
On October 28 Google started bringing ads in AI Mode to Australia, Canada, India, Kenya, Malaysia, New Zealand, Nigeria, Pakistan, the Philippines and Singapore, with Brazil, Mexico and the United Kingdom planned for the fourth quarter. It also launched multiturn ads in the US, which can appear in follow-up questions inside a conversation and are labelled "Sponsored". Google says AI Mode queries are on average three times longer than traditional searches, citing its own US study from May 2026 [5].
For advertisers, comparison and decision questions are becoming ad inventory. For organic visibility, the effect on cited links and click-through is not yet measured.
How fast is Google Search growing?
Alphabet's third-quarter release, filed on October 28, shows Google Search & other revenue up 12% to $63.35 billion, and the company said its year-over-year growth in search queries was the highest in five years [6]. The quarter ended September 30, so these are new disclosures about July to September, not measurements of the brief's week.
What did Reddit say about Google traffic?
On its October 29 earnings call, Reddit said Google search referrals fell by more than 1 million a week on average during October, and its chief executive linked the decline to the shift towards AI Mode and AI Overviews [7][8]. That attribution is the company's own view; Reddit did not publish a method that separates AI features from other causes.
Confirmed Figures as disclosed by Google, Alphabet and Reddit. Directional The cause of Reddit's decline
What did new SEO and AI search research find?
Three studies landed this week: AI visitors to US retailers spent more, organic losses varied widely by site type and country, and AI Mode product answers mostly linked to Google's own pages.
Are AI-referred visitors worth more?
In US retail in September, yes. Adobe's report of October 29 found that, compared with other visits, AI-referred visits converted 90% better, earned 83% more revenue per visit, stayed 73% longer and viewed 25% more pages, while AI traffic to retail sites grew 71% year over year. Adobe's data covers more than 1 trillion visits to US retail sites [9][10].
The figures are relative and limited to US retail. They do not show that AI referrals outperform other channels in B2B or services, so measure your own conversions and lead quality rather than borrowing Adobe's ratio.
How much organic traffic have sites lost?
Kevin Indig's analysis of 17,863 domains (663,861 country-domain-month records) found organic visits down 13.2% in the US and 24.4% in France over two years, while paid search visits rose. Publishers lost far more (about 41.5% in the US and 42.5% in France); without them, the decline was 7.8% in the US and 22.5% in France [11].
The data is modelled traffic estimates, and the study does not show that AI caused the losses. Its practical point is that an average decline figure says little about your site: compare against peers with the same business model.
Where do AI Mode product answers send people?
Findrix asked 18,400 buyer questions and collected 55,198 answers from ChatGPT, Gemini and Google AI Mode. In AI Mode product answers, 57% of the links pointed to Google's own pages, and 46.5% of US product answers linked only to Google pages [12]. The study counted links shown, not clicks or purchases.
Strongly supported Published research with stated samples; each covers a specific market and none measures your site
Can you trust your rank tracker this month?
Not without checking it. On October 27 Semrush confirmed that some Google results collected by rank trackers are wrong, which can show ranking changes that never happened.
Semrush called them "poisoned" results: pages listing results that do not match the query, sometimes on page one. It said it added checks to detect and drop invalid results; other trackers have reported similar problems since September. Google has not explained the cause, so claims that Google serves altered results to tools on purpose remain guesses [13].
- Compare any large tracker movement with Search Console clicks and impressions for the same query and page
- Wait for the movement to persist over several days before changing a page
- Flag one-day swings to unrelated results as data errors, not ranking changes
Confirmed The problem exists. Not established Its cause
Which AI model and agent releases matter for SEO work?
Three releases matter for how SEO work gets done, and none of them is a confirmed change to how AI search engines pick sources.
- Gemini 4 Carbon (Google, Oct 28): aimed at cybersecurity, coding and long-running agent tasks; access is limited to selected cyber defenders, and the benchmark scores are Google's own [14]
- GPT-6.1 Terra (OpenAI, Oct 26): available in the API, ChatGPT Work and Codex, with computer use, at $1 per million input tokens and $5 per million output tokens [15]
- Copilot harness (Microsoft, Oct 28): the orchestration layer behind Copilot, opened to developers in Microsoft Foundry as a public preview, free until December 1 [16]
For audits and reporting agents, the useful lesson is about controls rather than models: keep agents read-only by default, and require a person to approve anything that publishes, edits a site or changes an account setting.
On Bing, the most recent relevant change is still the Topics Portfolio beta in Bing Webmaster Tools, announced on October 13, which groups the AI queries that cite your site into themes and lists queries where competitors are cited instead [17].
What did I find on my own site this week?
That my own analytics had been recording nothing, which is the same measurement lesson as the rank-tracker story, closer to home.
On October 31 I checked this site's Google Analytics 4 property and found it had never received data. Site Kit reported its tag as inserted, but the theme's script allowlist removed it before pages reached visitors. The fix was one line, and Realtime showed the first visits within minutes. The point for any team: check that a tag actually fires on the live page before trusting a dashboard built on it.
On November 1 I also published the AI-Crawler Access Audit 2026, a register of how 500 websites treat AI crawlers in robots.txt, for anyone who wants measured data rather than assumptions about crawler access.
Confirmed First-hand, this site only
What was left out of this brief, and why?
Items that circulated this week but could not be traced to a primary source were left out rather than repeated.
That includes a reported Google change to URL-separator guidance (the latest revision found dates from June 2025), an AI-search opt-out case study whose dates and traffic were not published, and a claimed government certification for an AI search product that the public register did not show at the stated level. If a primary source appears, they will be added with a correction note.
What should you do this week?
Check your measurement before acting on it, keep shareable pages out of the index, and report AI referrals by outcome rather than volume.
| Priority | Action | Why |
|---|---|---|
| Now | Check big ranking moves against Search Console before changing pages | Tracker data is unreliable this month |
| Now | List every template that creates shareable or user-generated URLs and set noindex where they are not meant for Search | Google's October 26 guidance |
| Now | Confirm your analytics tag fires on the live site | A dashboard is only as good as its tag |
| Next | Report AI referrals by conversions and lead quality, separately from citations and sessions | Adobe and Reddit show volume and value moving apart |
| Next | Review commercial comparison questions in AI Mode for the markets you sell to | Ads now appear in those answers in more countries |
| Later | Test new models on one controlled workflow before adopting them | Vendor benchmarks are self-reported |
For a structured way to separate these measurements, see ranking, retrieval and citation and AI visibility tracking from citations to qualified leads. Shorter source-only headlines are collected on the industry news page.
Frequently asked questions
Was there a Google core update in late October 2026?
No. Google's Search Status Dashboard listed no ranking incident between October 23 and 30, 2026; the volatility came from third-party trackers, some of which had data problems.
Should shared or AI-generated pages be blocked in robots.txt?
Use noindex for pages that must stay reachable by link but out of Search; a robots.txt block alone can still leave the URL indexed if other pages link to it.
Do AI referrals convert better than other traffic?
In Adobe's US retail data for September 2026 they did, by 90%; that does not transfer automatically to B2B or services, so measure it on your own site.
Are ads in AI Mode available in my country?
As of October 28, 2026, Google listed the US plus Australia, Canada, India, Kenya, Malaysia, New Zealand, Nigeria, Pakistan, the Philippines and Singapore, with Brazil, Mexico and the UK planned for the fourth quarter.
Sources
Each item links to the original statement or to the coverage that reported it; all were opened or checked on October 31, 2026.
- Google Search Central. Prevent user-generated spam on your site. Updated October 26, 2026. developers.google.com (opens in a new tab)
- Google Search Central. Introduction to robots.txt. developers.google.com (opens in a new tab)
- Search Engine Roundtable. Google search ranking volatility gets heated October 24th and Unconfirmed Google search ranking update. seroundtable.com (opens in a new tab), seroundtable.com (opens in a new tab)
- Google Search Status Dashboard. status.search.google.com (opens in a new tab)
- Google. Ads in AI Mode expand to new markets. October 28, 2026. blog.google (opens in a new tab)
- Alphabet Inc. Q3 2026 earnings release (Form 8-K, exhibit 99.1). sec.gov (opens in a new tab)
- CNBC. Reddit Q3 2026 earnings report. October 29, 2026. cnbc.com (opens in a new tab)
- Benzinga. Reddit Q3 2026 earnings call transcript. benzinga.com (opens in a new tab)
- Adobe. AI traffic drives 83% higher revenue per visit. October 29, 2026. business.adobe.com (opens in a new tab)
- Search Engine Land. AI traffic drives 83% more revenue per visit for retailers: Adobe. searchengineland.com (opens in a new tab)
- Kevin Indig, Growth Memo. Organic traffic is down, but it's not all AI. growth-memo.com (opens in a new tab)
- Search Engine Land. Google AI Mode sends 57% of product links to Google Shopping: Study. searchengineland.com (opens in a new tab)
- Search Engine Land. Semrush confirms "poisoned" Google results are impacting SEO tools. searchengineland.com (opens in a new tab)
- Android Authority. Google announces Gemini 4 Carbon. October 28, 2026. androidauthority.com (opens in a new tab)
- BleepingComputer. OpenAI launches GPT-6.1 Terra. October 26, 2026. bleepingcomputer.com (opens in a new tab)
- Visual Studio Magazine. Microsoft opens GitHub-built Copilot harness to developers through Foundry. October 28, 2026. visualstudiomagazine.com (opens in a new tab)
- Search Engine Roundtable. Topics Portfolio in Bing Webmaster Tools AI reports. seroundtable.com (opens in a new tab)
Confirmed Google, Alphabet and Reddit figures come from their own statements or filings. Directional Model and agent details come from launch coverage; vendor benchmarks were not independently tested.
