Query Fan-Out: The Questions Behind AI Answers
Google’s AI Mode and AI Overviews may split one question into several related searches, so a brand page has to answer those hidden sub-questions too. White Wood’s own monitoring engine, SOLEDAD, found whitewood.ai neither named nor cited in any of 172 answers for “cara muncul di hasil pencarian AI” on six AI engines from 18 September to 18 October 2026.
1. One question becomes many searches
Google states that AI Mode and AI Overviews may issue multiple related searches across subtopics and data sources to build a single response (Google Search Central). Google’s own launch post describes AI Mode “breaking down your question into subtopics and issuing a multitude of queries simultaneously” (Google, May 2025). So when a Jakarta buyer asks how brands appear in AI answers, the system may also look for a definition, a comparison and a how-to before it writes a word.
2. Ranking for the headline is not the whole answer
A page can rank for the main query and still leave a sub-question unanswered. If your page explains what GEO is but never compares it with SEO, the comparison has to come from another page. Google does not say how it picks that page, so the safe move is to answer the branch yourself.
3. Sub-questions are predictable, so map them
A practical way to map the hidden questions behind a buyer prompt is five types: definition, comparison, criteria, process and cost. You don’t need access to the engine to guess them. Write the buyer’s question at the top of a page, then list the follow-ups a careful buyer would ask next. That list is your fan-out map, and it doubles as your heading plan.
| Buyer question | Likely hidden sub-questions | Where to answer them |
|---|---|---|
| How do brands appear in AI answers? | What is GEO? How is it different from SEO? How long does it take? | One guide page with a comparison table |
| Which agency should we hire for GEO? | What should a report show? Which engines matter in Indonesia? | Criteria section plus FAQ |
| How do we measure AI visibility? | Which prompts? Which languages? Per engine or blended? | Method section with a sample table |
| Why is Gemini empty when ChatGPT names us? | Is the page extractable? Is the brand clearly identified? | Diagnosis checklist |
4. Cover close branches on the page, split distant ones
Answer the sub-questions that sit close to the main topic on the same page, and give a distant branch its own URL. A definition and a comparison belong together. A full pricing breakdown probably doesn’t. Twenty thin pages for every synonym is the wrong response; it splits your evidence.
5. Open each section with its answer
Each section should open with the answer to its sub-question, then explain it. A reader looking for “how long does GEO take” should not have to reach paragraph four, after a story, to find it. We write this way for readers; Google does not say it changes which page gets cited.
How we use this at White Wood
White Wood maps the fan-out for every prompt it tracks in SOLEDAD, White Wood’s AI-monitoring engine, before a page is briefed. That zero on our own tracked prompt “cara muncul di hasil pencarian AI” is why we are rebuilding our pages around their sub-questions rather than the headline alone. For the basics of AI search surfaces, start with how brands show up in AI search.
What this cannot tell you
Google says AI Mode and AI Overviews may use query fan-out. It does not say which sub-questions it runs for your prompt, or how often. Covering sub-questions is how we work, not a proven route to being cited.
Declared interest: White Wood sells this work and measures it with SOLEDAD, our own AI-monitoring engine.
FAQ
What is query fan-out in AI search?
It is how Google’s AI Mode and AI Overviews may split one question into several related searches across subtopics before writing a single answer. Mapping those sub-questions is a planning tool, not a guaranteed route into the answer.
Is query fan-out only a Google thing?
Google documents it by name for AI Mode and AI Overviews. Other assistants publish less detail on how they search, so treat the fan-out map as an editorial hypothesis and check it against real answers.
Do I need special markup for fan-out?
No. Google says there are no additional technical requirements or special schema for its AI features. What remains is a clear page with good answers.
Sources
- Google Search Central, AI features and your website: https://developers.google.com/search/docs/appearance/ai-features
- Google, AI Mode update (May 2025): https://blog.google/products-and-platforms/products/search/google-search-ai-mode-update/
- SOLEDAD, White Wood’s AI-monitoring engine: baseline for whitewood.ai on six engines (ChatGPT, Gemini, Perplexity, Copilot, Google AI Overviews, Google AI Mode), 18 September to 18 October 2026