What is Query fan-out?
Query fan-out is how AI search systems answer one question: they silently expand it into several related sub-queries, retrieve sources for each, and assemble a single answer from the results — so a page can be cited for a question it never literally contains.
How does query fan-out work in AI search?
When you ask Google's AI Mode or ChatGPT search a question like 'best accountant for a small café in Lyon', the system does not run that string once. It generates a fan of narrower queries — accountants in Lyon, accounting for hospitality, small-business bookkeeping fees — runs each against the web index, and synthesises one answer from the union of sources. Google described the mechanism publicly when it launched AI Mode in 2025; Perplexity and ChatGPT search behave the same way.
The consequence is that citation is decided per sub-query. A page that answers one narrow facet very plainly can be cited in the final answer even though it never mentions the original question.
What does query fan-out change about how to write?
It rewards pages that answer one specific question completely, under a heading phrased as that question, with a self-contained paragraph an engine can lift. A long page that covers ten things vaguely loses to ten short sections that each own one sub-query. It also rewards covering the adjacent questions a customer would ask next — those are the sub-queries the fan-out produces.
How do you know which sub-queries you are cited for?
You cannot see the fan-out directly, but you can measure its output: ask the assistants the questions your customers ask and record whether you are named and which pages are cited. SeeGeo's tracking does this on a schedule across ChatGPT, Claude and Gemini, and the audit's extractability score tells you whether a page has the shape a sub-query can lift.
Frequently asked questions
What does Query fan-out mean?
Query fan-out is how AI search systems answer one question: they silently expand it into several related sub-queries, retrieve sources for each, and assemble a single answer from the results — so a page can be cited for a question it never literally contains.
What is query fan-out in AI search?
Query fan-out is the step where an AI search engine turns one question into several related sub-queries, retrieves sources for each, and merges them into one answer. It is why a page can be cited for a question it never states word for word.
Does query fan-out mean I should write longer pages?
No — it means writing more precisely. Each sub-query is matched to a passage, so pages that answer one question plainly under a question-shaped heading are lifted more often than long pages that mention many things.
Which engines use query fan-out?
Google AI Mode and AI Overviews describe it explicitly; ChatGPT search, Perplexity and Claude with web search retrieve in the same multi-query way even where they do not use the term.