Query fan-out
Query fan-out is the process by which an AI search engine decomposes a single user prompt into multiple synthetic sub-queries, retrieves results for each independently, and then synthesises one answer from the combined results. It means a page can be retrieved for questions the user never typed, because the engine generates and searches those questions on the user’s behalf.
Why query fan-out matters for an existing site
One prompt fans out into 5–10 sub-queries you never see. The engine writes its own questions, retrieves chunks for each, and stitches the answer together — which means it’s shopping for sections, not pages. So the practical move is boring and specific: every H2 answers one sub-question completely, on its own, without leaning on the paragraph above it. Write for the fan-out, not the headline query. A page built as one long argument gets skipped for chunks; a page built as a set of self-contained answers gets mined repeatedly. Here’s what a single prompt actually turns into behind the glass:
Prompt: “How do I get my website cited by ChatGPT?” Fans out to (our worked example):
- What is generative engine optimization?
- Which on-page factors make a page citable by AI answer engines?
- How do answer engines choose which sources to cite?
- Do brand mentions affect AI citations?
- How do you measure a site’s AI citation share?
- Is GEO different from SEO?
Six retrievals, one answer. Own six chunks and you’re in it six ways.
Related: AEO · GEO Read up: Query Fan-Out for Existing Sites · How AI Answer Engines Choose Citations · Information Gain