Query fan out: The process by which answer engines translate a single human prompt into multiple smaller, search-optimized queries. The engine doesn't search your exact prompt - it dispatches a set of rewritten variations.
Fan out variation: A single query in the dispatched set. Each variation has a share percentage indicating how often that specific phrasing appears across executions.
Word transformations: The vocabulary changes between your original prompt and the fan out variations. Words are categorized as added, dropped, or preserved - revealing the language the engine actually uses when searching.
Freshness pressure: When engines inject date tokens (e.g., "2026") into fan out variations, signaling that outdated content may be filtered out.
The retrieval chain: Prompt → Fan out variations → Results per variation → Extracted clips
Platform behavior:
Accessing fan out analysis:
What to examine in the analysis drawer:
Prioritization principle: Win the highest-share fan out variations first. A variation appearing 21% of the time matters more than one appearing 3% of the time.
Persona impact: The same prompt produces different fan out variations depending on the selected persona. A content marketing persona generates variations about "cited content" and "SEO tools" while a brand/PR persona generates variations about "sentiment control" and "reputation management."
When to use fan out:
When NOT to use fan out:
For each high-share variation, ask:
