What is query fan-out?
Query fan-out is the technique Google uses in AI Mode and AI Overviews to answer a question by breaking it into subtopics and running many related searches at once. The system then reads the results of all those searches and writes one answer, citing pages found along the way.
Google has described the technique publicly. Its practical consequence: the pages cited in an AI answer do not have to rank for the words the user typed. They only have to rank well for one of the hidden sub-queries.
How it works
Take the question "is a burr grinder worth it for a French press if I only brew on weekends". A fan-out system might issue searches such as:
- burr vs blade grinder French press
- best grind size French press
- budget burr grinders
- how long ground coffee stays fresh
- manual burr grinder for occasional use
Each sub-query returns its own results. The model pulls relevant passages from across them and composes an answer that covers cost, grind consistency, and freshness, with links to the sources used. For harder questions, Google's more extensive research modes run many more searches.
What it means for SEO
- Rankings for sub-topics count. A page that ranks well for "how long ground coffee stays fresh" can be cited for a much broader question.
- Passage clarity matters. The system needs a section it can lift. A heading that names the sub-question, followed by a direct answer, is easier to use than an answer spread across paragraphs.
- Coverage beats repetition. A page, or a group of pages, that answers the obvious follow-up questions matches more of the fan-out. This is the logic behind topic clusters and topical authority.
- Exact-match keyword work matters less for these features. Matching the meaning of related questions matters more.
How to write for it
- List the sub-questions a person would need answered to decide. People Also Ask boxes and Search Console queries are good sources.
- Answer each in its own section, with the answer in the first sentence.
- Link related pages so each sub-topic has a clear home. See internal linking.
- Include specifics: numbers, conditions, trade-offs, and first-hand testing. Generic summaries give a model nothing distinct to cite.
None of this is special markup. It is clear writing organized around the questions people ask. The AI Overviews guide goes into page-level changes.
Common misunderstandings
- You cannot see the sub-queries. Google does not publish them per search, and Search Console does not list them. You can only infer them.
- Fan-out is not a new ranking factor. It uses Google's normal index and ranking systems many times over.
- It is not only Google's idea. Other AI search tools also run several searches per question, though each works differently.
Common questions
Can I see which fan-out queries Google ran?
No. Google does not expose the sub-queries for a given answer. Related searches, People Also Ask, and your own Search Console query data are the closest proxies.
Does query fan-out mean long pages win?
Not by length alone. It favors content that clearly answers the specific sub-questions, which can be one well-organized page or several linked pages.
Is query fan-out used in regular search results?
Google describes it as a technique behind AI Mode and AI Overviews. Regular ranked results work from the query the user typed.