The engine searched for six things. You've written about two.
When an engine answers a question, it quietly runs several searches of its own first. RankSage records the ones your stored answers actually triggered, and checks whether your content covers each of them.
The problem
Everyone guesses what an answer engine looks for. The engine already tells you — in the sub-searches it runs — and that trail is thrown away by every tool that only reads the final answer.
Observed Fan-out on one screen.
What it measures, and what comes out.
Outcome and evidence at each step, never the recipe. Every number here can be traced to something stored.
- 01
Observed, not simulated
These are the sub-searches recorded from real answer runs — what the engine's own retrieval did, not a model's guess at what it might do. That difference is the whole feature.
- 02
Attached to the answer they built
Each fan-out trail stays linked to the stored answer it produced, so you can read a citation result and the searches behind it side by side.
- 03
Checked against your content
Every observed sub-search is matched against your crawled pages. A sub-search you have nothing for is a named, evidenced gap — not a hunch.
- 04
It compounds with question mining
Mined questions tell you what people ask; observed fan-out tells you what engines retrieve when they answer. Covering both is what makes a page the source an answer gets built from.
What it does not do.
Disqualifying fast is part of the product. These are the edges, written down.
Not every engine exposes its retrieval trail, and none promises to. Where a run carries no observable fan-out, that is reported as not observed — never invented.
The trail reflects the runs you have stored. More tracked prompts mean more observed fan-out.
Coverage matching tells you a sub-search is unaddressed. Writing the page that addresses it is still the work.
The capabilities it joins.
Read the longer version.
Questions about Observed Fan-out, answered.
What it measures, what it needs, and where its edges are.
When an answer engine handles a question, it typically runs several narrower searches of its own — the fan-out — and builds the answer from what they return. If your content doesn't cover those sub-searches, it rarely makes it into the answer.
It isn't a guess. RankSage records the sub-searches that real answer runs actually triggered, so the list you optimise against is what the engine did, not what a tool predicted it might do.
Treat it as a content gap with unusually strong evidence: an engine went looking for this while answering a question you care about, and found nothing of yours. It can flow into a brief like any other gap.
AI is already answering for you.
Sign in to Google Analytics and Search Console. Read-only, about five minutes, no code on your site for that step.
No payment details. One email when access opens.