GEO Glossary
Query Fan-Out
With query fan-out, the AI splits a question into several sub-searches, finds sources for each and merges the results into one answer.
At a glance
PROMPT "Best domain extension for an AI startup?" │FAN-OUT ├─ ai domain cost ├─ .ai or .com for startups └─ startups with ai domain │ANSWER combines all sub-searches, with sources
Example
Google itself describes the approach for AI Mode: the question is broken into subtopics and many searches run at the same time (source: Google, 20 May 2025).
Why it matters for GEO
You get found not only for the original question but for every sub-search. Content that answers each sub-question in a clear section is more likely to end up in the answer.
In practice
Derive the sub-questions for your most important Prompts , for example with fan-out simulators or by looking at the search steps in Perplexity. Then check whether your page has a passage for each sub-question.
How it fits together
Via RAG the results of the sub-searches land in the model’s context, and via Grounding the answer is based on those sources. That’s why query fan-out helps decide which pages end up with a Citation .