Query fan-out analysis: see the searches AI runs before it answers
Query Fan-Out is AnswerixAI's view of the sub-queries AI answer engines search while answering your tracked prompts. Each daily run records the searches engines such as Perplexity and Microsoft Copilot actually made. It counts how many answers ran each one, labels its search intent, and lets you track any sub-query as a prompt of its own.
Query fan-out
Observed in tracked answers · last 30 days
Tracked prompt
DailyBest espresso machines under $500?
- best espresso machines under $500
Best / Top×11 - espresso machine under $500 2026Best / Top×7
- best espresso machine reviewsvs. / Review×6
- affordable home espresso machineRecommendation×4
- semi-automatic vs super-automatic espresso machineComparison×3
- espresso machine with built-in grinder under $500Best / Top×3
Captured from the answers, never predicted+2 more
Built for marketers, executives, builders, and developers worldwide.
What is query fan-out?
Query fan-out is how an AI answer engine, also called a generative engine, turns one prompt into several narrower web searches before it writes a reply. Each of these searches is a sub-query. It is the search step of retrieval-augmented generation, where a model looks up sources before it answers. The results of those sub-queries decide which pages the engine reads, which brands it names and which sources it cites. If your pages don't answer the sub-queries, they are less likely to be found for the original prompt at all.
- 01
User prompt
Someone asks a broad question, such as which espresso machine to buy.
- 02
Sub-queries
The engine splits it into narrower web searches: prices, reviews, comparisons.
- 03
Source retrieval
Each search returns pages: reviews, forums, guides, shops and brand sites.
- 04
Answer synthesis
The engine weighs what it found and merges it into one response.
- 05
AI answer
The reply names brands and cites sources found through those searches.
Which answer engines show their sub-queries?
Perplexity and Microsoft Copilot report most of the sub-queries AnswerixAI captures, and only some of their answers run a live search at all. Answers that report no search still count toward a prompt's coverage, so you can see how often the engines research it, but they add no sub-queries.
Are the sub-queries real or predicted?
Real. We built AnswerixAI to never generate sub-queries itself. It lists only the searches the engines reported in the answers it tracked for you. A prompt whose answers never searched shows zero sub-queries instead of a guessed list.
How does AnswerixAI show query fan-out?
The Query Fan-out tab on the Prompts page groups the searches by prompt or ranks them by frequency. It covers the last 24 hours, 7, 30 or 90 days, or a custom range of up to 90 days. Every tracked answer also lists its own sub-queries.
Every tracked prompt with a run in the date range gets a row showing how many of its answers ran a live search, such as 29 of 113 (26%), and how many distinct sub-queries they produced. Expand a prompt to see each sub-query with the engine that ran it, the number of answers that searched it, its intent and a button to track it.
Query fan-out
By prompt · last 30 days
- Best monthly coffee bean subscription for offices50%10 / 20 · 50%10
- Best espresso machines under $500?26%29 / 113 · 26%8
- best espresso machines under $500Best / Top×11
- espresso machine under $500 2026Best / Top×7
- best espresso machine reviewsvs. / Review×6
- affordable home espresso machineRecommendation×4
- best espresso machines under $500
- What are the best coffee subscription services in 2026?23%26 / 114 · 23%6
- Manual vs automatic espresso machine for home use39%15 / 38 · 39%6
- Which specialty coffee subscriptions ship internationally?19%8 / 42 · 19%6
- Most sustainable coffee brands0%0 / 36 · 0%0
What can you do with the sub-queries?
A recurring sub-query is a question the engines keep researching around your market. Our team built three ways to act on it: measure it, write for it and share it.
Track a sub-query
From the High frequency view
RecommendationTrackedTracked prompts
- office coffee subscription serviceNew
Same platforms as your latest prompt - Best coffee subscription for beginnersSubscriptions
- Most sustainable coffee brandsSustainability
- Best espresso machines under $500Equipment
Press + next to any sub-query and it becomes a tracked prompt on the same AI platforms as your most recent prompt. From its first run it gets its own visibility, mentions and citations. A new prompt counts toward your plan’s prompt limit, and a sub-query you already track shows a Tracked badge instead of the button.
A content opportunity built from a tracked prompt lists that prompt’s observed fan-out queries and how often each was searched. Next to them sit its estimated AI search volume, your visibility and the competitors cited. Cover those sub-queries on the page you write or update, so it answers the searches the engines actually run.
Add the Query Fan-out section to any report; the Monthly Executive Report includes it by default. It lists the period’s ten most-searched sub-queries. For each engine it shows how many distinct sub-queries it ran across how many answers, so one engine’s repeated wording isn’t mistaken for the whole picture.
How do you use query fan-out analysis?
Five steps take an observed search from the Query Fan-out tab to a change in AI visibility you can measure on the same dashboard.
- 1
Observe
Each daily run records the sub-queries engines report while answering your prompts.
- 2
Prioritize
Rank sub-queries by how many answers searched them and read their intent.
- 3
Track
Press + to follow an important sub-query as a prompt of its own.
- 4
Optimize
Cover the recurring sub-queries on the pages you brief and update.
- 5
Measure
Watch visibility, mentions and citations for the prompt and its sub-queries.
Explore Related Features
Put query fan-out to work
Track the prompts behind the sub-queries, turn them into content opportunities and briefs, and see which sources AI engines cite in your answers.
Prompt Monitoring & Volumes
Track the prompts buyers ask ChatGPT, Perplexity and Gemini every day, group them by topic, and estimate each prompt's monthly AI search volume and intent.
Content Intelligence
Find the prompts where AI answers favor competitors, rank each gap by AI volume, visibility and intent, then brief, draft and send the fix to your workflow.
AI Content Briefs & Outlines
Turn prompts where ChatGPT, Perplexity and Google AI Overviews cite competitors into briefs with a title, outline, keywords, CTA and competitor insights.
Citations Monitoring
Find the domains and URLs ChatGPT, Perplexity, Gemini and Google AI Overviews cite for your prompts, split into your site, competitors and third parties.
Customer stories
How do content teams use query fan-out?
“The query fan-out insights have been especially valuable for our content team. They help us understand the supporting questions behind a broad topic and create pages that answer the subject much more comprehensively.”
FAQ
Query fan-out FAQ
Query fan-out is how an AI answer engine splits one prompt into several narrower web searches before it writes a reply. The results of those sub-queries decide which pages the engine reads, which brands it names and which sources it cites. A page that answers them is more likely to be used for the original prompt.
When AnswerixAI runs your tracked prompts each day, it stores the sub-queries each AI engine reports alongside its answer. It never generates sub-queries itself. The Query Fan-out tab on the Prompts page then groups them by prompt or ranks them by how many answers searched them. Every tracked answer also lists its own sub-queries with the engine that ran them.
Perplexity and Microsoft Copilot report most of the sub-queries AnswerixAI captures, and only some of their answers run a live search. Starter tracks ChatGPT and Perplexity; Growth and Business add Gemini, Microsoft Copilot, Google AI Mode, Grok and Google AI Overviews. Answers that report no search still count toward each prompt's fan-out coverage.
By prompt lists every tracked prompt with how many of its answers ran a live search and how many sub-queries they produced; expand one to see its sub-queries. High frequency ranks all sub-queries across your prompts by how many answers searched them, with the engines that ran each one, the prompts that triggered it and its search intent.
It is the share of a prompt's tracked answers in the date range that ran at least one live search, shown as a count and a percentage, such as 29 of 113, or 26%. Without it, eight sub-queries would look the same whether they came from ten answers or five hundred, so it shows how often the engines research each prompt.
An AI model classifies each sub-query into the same intent types AnswerixAI uses for prompt volumes: Comparison, How-to, What is, Best / Top, vs. / Review, Recommendation or Problem Solving, with Other for anything else. Results are cached by query text, so a sub-query keeps the same intent everywhere it appears, and the badges fill in as the table loads.
Yes. Press + next to any sub-query and AnswerixAI adds it as a tracked prompt on the same AI platforms as your most recent prompt, so it gets its own visibility, mentions and citations. New prompts count toward your plan's limit: 50 prompts on Starter and Growth, 200 on Business. Sub-queries you already track show a Tracked badge instead.
Sub-queries show the narrower questions AI engines research around a prompt, such as price limits, comparisons and reviews. Covering them on the pages you write or update makes those pages useful for the searches the engines actually run. Content opportunities built from a tracked prompt list its observed fan-out queries, so briefs can target them directly.
Yes. Query fan-out is part of Starter, Growth and Business, and every plan runs your prompts daily and starts with a 14-day free trial. The sub-queries you see depend on the engines your plan tracks: Starter includes Perplexity, while Growth and Business add Microsoft Copilot, the other main source of fan-out data.
The Query Fan-out tab covers the last 24 hours, 7, 30 or 90 days, or a custom range of up to 90 days, matching the rest of the Prompts page. Reports can include a Query Fan-out section for their own period, with the ten most-searched sub-queries and each engine's distinct sub-queries and answers with fan-out.
See the Searches Behind Your AI Answers
Use AnswerixAI to grow AI Search visibility with AEO and GEO.