Query Fan-Out Tool

QueryBurst's Query Fan-Out tool simulates how Google's AI Mode and AI Overviews decompose a user query into thematic sub-queries for research. Based on Google's patented "Thematic Search" methodology, it generates the same type of fan-out queries that AI systems dispatch internally — plus the reasoning behind each one, authority signals to look for, and potential sub-themes for deeper exploration.

Why Simulated Fan-Outs Are Just As Good

• They're LLM-generated — just like the real ones. Google's fan-out queries are generated by a language model. They're probabilistic, not deterministic. Same training data, same patterns, same thematic clusters.
• Real fan-outs change every time. Run the same query twice and you'll get different sub-queries. Tracking the "exact" ones is tracking noise. The themes converge — the specific wording doesn't matter.
• Any experienced SEO can predict them. Fan-out queries for "best CRM software" will include pricing comparisons, integration lists, and reviews. This tool systematises what you already know and adds the layers you don't.
• ChatGPT just removed public fan-out data. Prompt trackers that relied on exposed fan-out queries lost that signal overnight. This tool makes that irrelevant — because you never needed the "exact" ones in the first place.

Why This Is Different

• Based on Google's patents. Not a guess. Google's "Thematic Search" patent (US12158907B1) describes how search results are decomposed into thematic clusters using passage summarisation and a clustering engine. Their "Search With Stateful Chat" patent describes the broader AI Mode pipeline. Our tool simulates this methodology.
• Reasoning, not just queries. Every fan-out query comes with an explanation of why it was generated and why it's strategically important. Tracking tools never gave you this.
• Authority signals included. For each theme, you get the quality indicators, credibility markers, and content types that the system would prioritise when evaluating results.
• Free to use. Available in our SEO app with no subscription required. Also available as part of the Verify platform for site-level analysis.

Stop tracking fan-outs. Start understanding them.

The AI search industry has spent the last year building tools to scrape, track, and chart query fan-outs — the sub-queries that AI systems like Google AI Mode and ChatGPT generate behind the scenes when constructing an answer.

Then ChatGPT removed the data from their API. Overnight, the trackers lost their signal. Panic ensued.

But here's the thing: it doesn't matter. Fan-out queries are generated by a language model. They're probabilistic. They change every time. The "exact" fan-outs for a given query at a given moment are a snapshot of noise, not a strategic asset.

What matters is the themes. And the themes are predictable — because the model learned them from the same internet we all read. Google's own patent describes a two-phase process: start with a broad search, then break into thematic sub-queries based on expected content patterns. That's what our tool simulates.

The difference is that we don't just give you a list of queries. We give you the reasoning behind each one, the authority signals the system would look for, and the sub-themes it would drill into. That's the intelligence that tracking tools never provided — because they were too busy monitoring the output to understand the input.

What You Get (Beyond Just The Queries)

Prompt trackers gave you a list of sub-queries. Our tool gives you the complete research strategy — the same intelligence that Google's AI uses to plan its answer.

Initial Broad Query

The foundational search the system would start with — the first query in the two-phase process described in Google's Thematic Search patent.

Hypothetical Content Summary

What types of information and sources the system expects to find from the initial search. This tells you what the model considers "standard" for your topic.

Thematic Sub-Queries (Priority Scored)

3–7 specific fan-out queries, each scored by priority (critical, important, or supplementary). These are the research threads the AI would pursue.

Generation Reasoning

For each theme: why this specific sub-query would emerge from the initial results. This shows the logical chain from broad search to focused research.

Strategic Reasoning

Beyond generation reasoning: why each sub-query is strategically necessary for comprehensive coverage. This is the "why it matters" layer.

Authority Signals

The quality indicators, credibility markers, and specific criteria the system would use to evaluate results for each theme.

Cite Examples

Types of content, specific publications, and source formats the system would prioritise when selecting citations for each theme.

Synthesis Instructions

How the AI should combine all research findings into a coherent, user-focused response. This reveals the answer structure the model is building toward.

DIY Version — Copy This Prompt

Want a quick version? Copy this prompt into any LLM. It'll generate the fan-out queries. For the full patent-based analysis with reasoning, authority signals, and sub-themes — use our free tool.

// Paste this into ChatGPT, Claude, or Gemini

You are simulating ChatGPT's query fan-out process when web search is enabled.

Given a user query, generate the set of sub-queries that ChatGPT would likely dispatch to build a comprehensive answer.

For each sub-query, explain why it's needed.

Output as JSON: { "original_query": "string - the user's query", "fan_out_queries": [ { "query": "string - the sub-query", "purpose": "string - what this adds to the answer", "priority": "high | medium | low" } ] }

Frequently Asked Questions

What is query fan-out?

Query fan-out is the process by which AI search systems like Google AI Mode, ChatGPT, and Perplexity break a single user query into multiple sub-queries to gather comprehensive information. For example, "best personal injury lawyer in New York" might fan out into sub-queries about specialisations, fee structures, case results, client reviews, and jurisdiction experience. Google confirmed this technique at Google I/O 2025, and the mechanism is described in their "Thematic Search" patent (US12158907B1) and "Search With Stateful Chat" patent.

How is this different from tracking real fan-outs?

Real fan-outs are generated by a language model and are probabilistic — they change every time. Tracking the "exact" fan-outs for a given query at a given moment gives you a snapshot of noise, not a strategic asset. Our tool generates the same type of thematic sub-queries using the same type of model, and adds layers that tracking tools never provided: reasoning, authority signals, and sub-themes. You get the intelligence without the noise.

Is this based on the actual Google patent?

Yes. The tool is primarily based on Google's "Thematic Search" patent (US12158907B1, granted December 2024), which describes how Google takes the top search results for a query, generates summary descriptions for every passage using a language model, clusters those summaries into thematic groups, and ranks the themes by prominence — including how many distinct documents mention each theme. This is the fan-out mechanism: broad query in, thematic sub-queries out. We also reference the "Search With Stateful Chat" patent, which describes the broader AI Mode pipeline including snippet scoring and citation verification.