Query fan-out

Query fan-out is when an AI search tool splits one question into many smaller searches, runs them at once, and writes its answer from all the results.

AI searchUpdated 2 min read

Also called fan-out queries · sub-queries

42%

of the pages AI cited on one day were cited again the next. Different searches bring back different pages.From our own research →

What it means

When you ask an AI search tool a question, it often doesn't search for your exact words. It works out the smaller questions behind it and searches for each one.

Ask "best CRM for a small real estate team", and the tool might also search for "CRM for real estate agents", "simple CRM pricing", "CRM with email templates" and "HubSpot vs Pipedrive for small teams". Then it reads the top results from all of them and writes one answer.

Google uses the name "query fan-out" for how AI Mode works. AI Overviews and ChatGPT do the same thing on a smaller scale.

Why it matters

Fan-out means the pages behind an answer don't all rank for the question you typed. Some rank for the related searches instead. A page that ranks for several of them gets more chances to be read.

A Surfer study of 10,000 keywords and 33,000 of the fan-out searches behind their AI Overviews found that pages ranking for the main search plus at least one fan-out search were 161% more likely to be cited. That's a link, not proof of cause, but it fits how the system works.

Fan-out is also one reason answers change. Different runs pick different related searches, which bring back different pages. In our study, only 42% of cited pages were the same from one day to the next.

What to do about it

  • Cover the questions around your main one: pricing, comparisons, use cases and "for X" versions.
  • Write pages that answer one clear question each, so they match a specific search.
  • Get onto pages that rank for the related searches, not only the main one.

Example

Ask ChatGPT "best payroll software for a small restaurant". Before it answers, it might run searches like "payroll software restaurants 2026", "restaurant payroll tips", "Gusto vs ADP small business" and "payroll software pricing comparison".

In our study, ChatGPT added the year to its search for every question. After that, the words it added most were "comparison", "pricing" and "reviews".

Free tool: Query Fan-Out GeneratorType a buyer's question and see the searches ChatGPT and AI Mode run behind it, and the pages that win them.Try it →

Frequently asked questions

Can I see the fan-out searches?

Sometimes. ChatGPT shows the searches it ran in some answers, and some tools record them. Google doesn't show AI Mode's searches.

Does fan-out mean I need a page for every keyword?

No. It means your topic should be covered from the angles buyers care about: pricing, comparisons, use cases. One strong page per angle is enough.

Why do AI answers cite pages that don't rank for my question?

Because they rank for one of the related searches the AI ran. That's fan-out.

Go deeper

All posts →
Research

AI product recommendations are mostly a summary of listicles

We fetched every 'best X' page ChatGPT, Gemini and Google AI cited in 839 answers. 90% of the brands they recommended were on one of those lists. Brands on none were recommended 2.5% of the time. Here's how it works and what to do about it.

Make AI search your next revenue channel

Start free trial