Fanout Queries Dashboard
The Challenge
You're optimizing for the question — not the search
AI assistants don't answer prompts directly. They fan each one out into narrower searches and ground their answer in what those searches return. If you can't see those sub-questions, you can't see why you're cited — or why you're not.
You optimize for prompts, not real searches
AI models rarely answer your prompt directly. They fan it out into dozens of narrower sub-questions and search those instead. If you don't know what they're actually looking up, you're optimizing blind — tuning content for a question the model never really asked.
Content gaps stay invisible
A model might search 'brand return policy' 24 times across a month and never surface your page. That gap never appears in a rankings report — but it's the exact reason you aren't being cited. Without the fanout, the miss is silent.
You can't tell must-own topics from noise
Some sub-questions are searched by every model; others are one-off tangents. Without seeing which queries overlap across ChatGPT, Gemini, and Perplexity, there's no way to know where to spend content effort first.
The signal is buried in raw responses
The sub-queries live inside each model's grounding and related-question metadata, scattered across hundreds of runs and providers. Reading them by hand — and deduping them into something actionable — is effectively impossible.
Capabilities
The searches behind every AI answer
Fanout Queries turns raw model grounding into a ranked, filterable map of what AI is really looking up about your brand.
Behind-the-answer visibility
See the actual sub-questions each AI model searched while answering prompts about your brand — aggregated across every run and provider into one clean list.
Frequency ranking
Every fanout query is ranked by how often it was looked up, so the topics models care about most rise to the top instead of drowning in the long tail.
Per-model breakdown
For any query, see exactly which models searched it and how many times — ChatGPT, Gemini, Perplexity, Google AI Overview, and more.
Shared across models
Filter to the sub-questions two or more models agree on. These consensus topics are the ones to own first — every major assistant is looking for them.
Prompt drill-down
Click any fanout query to see which of your tracked prompts triggered it, so you can trace a topic straight back to the questions your buyers ask.
One-click export
Export the full, filtered list as CSV and hand your content team a ready-made brief of the exact topics AI is searching for.
Topic Map
Turn model search behavior into a content roadmap
Fanout Queries reverse-engineers how AI answers work. Instead of guessing which pages to write, you see the literal searches models run behind your prompts — ranked by frequency and mapped to the prompts that triggered them. It's a content backlog written by the models themselves.
- Every sub-question ranked by lookup volume
- Relative-frequency bars for instant scanning
- Trace each topic back to its source prompts
- Export to CSV for briefs and planning
Topic Map
Consensus Signal
Own what every model agrees matters
Not all sub-questions are equal. The 'Shared across models' filter isolates the queries multiple assistants look up for the same prompt — the consensus core. Win those pages and you're answering the questions ChatGPT, Gemini, and Perplexity all agree are essential, not chasing one model's quirk.
- Only queries searched by 2+ models
- Provider breakdown per query
- Prioritize consensus topics over one-offs
- Filter and search across the full set
Consensus Signal
How It Works
From answer to roadmap in three steps
Run your prompts
Maya runs your tracked prompts across every major AI model, just as your buyers would ask them.
Maya captures the fanout
Every sub-question each model searched is extracted, normalized, and aggregated into a ranked, deduped list.
Own the topics
Prioritize the shared, high-frequency queries, brief your team, and turn AI search behavior into pages that get cited.
Who It Helps
A content roadmap written by the models
GEO Specialists
Stop guessing at target queries. Build your GEO plan around the exact sub-questions models search — and measure coverage on the topics that actually drive citations.
Content Strategists
Get a demand-backed content backlog straight from AI search behavior. Every fanout query is a page idea validated by the models your buyers already ask.
SEO Teams
Extend keyword research into the AI era. See the semantic sub-questions behind head terms and close the content gaps that keep you out of AI answers.
Agencies
Show clients precisely what AI is looking for — and deliver content roadmaps backed by real model search data instead of intuition.
Frequently asked questions
What exactly is a fanout query?+
When an AI model answers a question, it usually breaks that question into several narrower searches — 'sub-questions' it looks up to ground its response. Those searches are fanout queries. Maya captures them from each model's grounding and related-question metadata and aggregates them across all your prompts.
Which models does this cover?+
Fanout queries are collected from every provider Maya runs that exposes them — including ChatGPT, Gemini, Perplexity, and Google AI Overview. Provider names are normalized, so the same engine gathered different ways is shown as one model.
How is this different from keyword research?+
Keyword research tells you what people type into a search box. Fanout queries tell you what AI models search on your behalf when they answer — the machine-generated sub-questions behind a response. It's the demand signal for the AI-answer era, not the traditional search era.
What does 'shared across models' mean?+
It filters the list to sub-questions that two or more AI models looked up for the same topic. These consensus queries are the highest-priority topics to own, because every major assistant is trying to answer them — win those pages and you show up across the board.
Can I export the data?+
Yes. Any filtered view — by provider, search term, or the shared-across-models filter — can be exported as a CSV. It becomes a ready-to-use content brief of the exact topics AI is searching for about your brand.