Query fan-out is the set of sub-queries an AI engine runs behind a single question before it answers. Fan-out analysis shows you those hidden queries, so you can write for what the engine actually searched rather than for what the buyer typed.
Most AEO work optimises for the buyer's question. The engine does not search the buyer's question. It breaks it into several narrower queries, retrieves sources for each, then composes an answer from what it found.
That means the citations are won at the sub-query level, one layer below where almost everyone is looking. If you have never seen the fan-out, you are optimising for a question the engine never ran.
When an engine answers a tracked prompt, we record the sub-queries it ran on the way there.
Fan-out is collected per prompt across every country, every engine and every capture date, so you see the queries that recur rather than a single sample.
Fan-out differs between engines. Narrow it down to see how a specific engine decomposes your category.
Export the list and write against the sub-queries that keep appearing.
Who it's for. Content and SEO leads who have run out of ideas from keyword tools, and want to see the questions the engines are asking themselves.
It is the set of narrower sub-queries an AI engine generates and searches behind a single user question, before composing its answer from the sources it retrieves.
Because citations are won at the sub-query level. A page that answers a recurring sub-query can be cited even when it does not target the original question.
No. Engines decompose the same question differently, which is why the analysis is filterable per engine.
Most do not. It is one of the clearer differences between Zadoosh and the tools we are usually compared against.
Start a free trial and track your first prompts across AI engines today.
Start free trial