Fan-out retrieves subtopics, not your best page: planning for coverage

Google documents that both AI Overviews and AI Mode may use a query fan-out technique, issuing multiple related searches across subtopics and data sources to develop a response. If that is how an answer gets assembled, the unit of competition is not the one page you optimised for the headline question. It is whether anything you publish is eligible to be retrieved for the smaller questions sitting underneath it.

Fan-out splits one question into many

A fan-out does not grade your best page against a rival's best page. The system issues several related searches and gathers supporting links from whatever comes back across those subtopics. A single thorough article can satisfy one branch and miss the others entirely. So the quantity that matters is how much indexed, snippet-eligible material you have spread across the subtopics a question decomposes into.

The eligibility gate has not moved

Google states that to be eligible to be shown as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to be shown in Google Search with a snippet, fulfilling the Search technical requirements. Google also states you don't need to create new machine-readable files, AI text files, or markup to appear in these features. The gate is the familiar one, and failing it removes a page from every branch at once.

Coverage is a publishing-rate question

Breadth costs operations, not cleverness. Our Tatano Energy platform, a multilingual e-commerce build for a biomass boiler manufacturer, reports country domains indexed separately: 4, languages served: 7, SEO articles published every day: 8, and manual intervention required: 0. The daily SEO autoblog is driven by search trends. That is the shape of coverage work: many pages across subtopics and markets, maintained without a person steering each one.

The same pattern shows up at personal scale. Our Matthieu Pesesse platform runs an autonomous publishing engine with AI-written articles, resilient multi-model fallback at 3 levels, full SEO and GEO optimisation, publishing daily without supervision at roughly 0 h of weekly content operations. Its GEO audit score moved from 66 to 90+ after the overhaul, which is the part worth copying: breadth that is graded, not breadth that is assumed.

Crawler access is the lever you set yourself

OpenAI documents that it uses OAI-SearchBot and GPTBot robots.txt tags to let webmasters manage how their sites and content work with AI. Sites opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though they can still appear as navigational links. Disallowing GPTBot indicates content should not be used in training generative AI foundation models. Those are two separate decisions, so read what your file actually says.

Treat robots.txt as configuration with a lag rather than a switch. OpenAI notes that for search results it can take around 24 hours from a site's robots.txt update for their systems to adjust. If an opt-out was added during a migration and never reviewed, the effect persists long after the reason for it did, and nothing in your analytics will point at the file.

Measurement folds into one row

Google states that AI Overviews and AI Mode are reported on in the Performance report, within the Web search type. There is no separate column. You can watch clicks and impressions for Web, but you cannot split out the portion that arrived through an AI feature. Any dashboard presenting a clean AI Overviews figure is inferring it, so build reporting around the aggregate and around the inputs you control.

Structured data still earns its place

Google reports that Rotten Tomatoes added structured data to 100,000 unique pages and measured a 25% higher click-through rate for pages enhanced with structured data, compared to pages without structured data. Google also warns against blank pages built to hold markup, and states it is more important to supply fewer but complete and accurate recommended properties than every possible property with badly formed data. Markup makes a retrieved page more attractive; it does not make an unindexed page eligible.

Treat fan-out as a coverage problem with a dirty scoreboard. Widen the set of indexed, snippet-eligible pages across the subtopics a question splits into, check robots.txt for opt-outs nobody meant to keep, and keep markup complete on pages people can actually read. Then accept that the Web row in the Performance report is the only honest number available, and judge the work by publishing rate and by audit scores you set yourself.

Sources

Google Search Central — AI features and your website — https://developers.google.com/search/docs/appearance/ai-features

Google Search Central — Introduction to structured data markup in Google Search — https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data

OpenAI — Overview of OpenAI Crawlers — https://developers.openai.com/api/docs/bots

Neurolinks case study — Four markets, one codebase — https://neurolinks.be/work/tatano-energy

Neurolinks case study — A personal brand that publishes itself — https://neurolinks.be/work/matthieu-pesesse-media