# What Is Query Fan Out and What It Means for Your Pages _Published: 2026-09-17_ [Resources](/resources/)  /  AI search # What query fan out is, and what it does to your service page By [Chris Johns](/about/)Published September 17, 2026Updated September 17, 20268 minute read On this page What it isWhat it does to your pageHow to structure for itHow to measure itWhat nobody can promiseThe stepsQuestions people ask The short answer Query fan out is when a search system takes one question, silently splits it into several sub-questions, runs them separately, and assembles an answer from the results. The practical consequence is that a thin service page which only answers the headline question stops being retrievable, because the pieces of the answer are being gathered from pages that answer the parts. ## One question becomes eight When someone asks an assistant who they should hire for a furnace repair in Boise, the system does not run that phrase as a single search. It decomposes it. What furnace repair companies operate in Boise. What do people say about them. What does the work typically cost. What qualifications matter. Which of these companies publishes evidence of those qualifications. Each of those becomes its own retrieval, and the final answer is assembled from whichever sources answered each part best. That is fan out. > You are no longer competing to be the best page for a question. You are competing to be the best source for each fragment of it. ## What this does to a typical service page A typical local service page introduces the company, lists the services, and asks for a call. It answers one question: what do you do. Under fan out, that page can be retrieved for exactly one of the eight sub-questions and is invisible for the other seven. The pages that get cited answer the fragments. What it costs. How long it takes. What qualifications to look for. What goes wrong. Who it is not right for. Those are not blog topics bolted onto a service page, they are the substance of what a buyer is actually asking. A worked example For a Boise plumbing query, the sub-questions a system is likely to generate include cost range, emergency availability, licensing, typical timeline, and what distinguishes one provider from another. Count how many of those your service page answers in a form a machine could lift cleanly. For most sites the answer is one. ## How to structure a page for fan out - Answer each sub-question under its own heading, phrased the way a person would ask it rather than as a keyword. - Put the answer in the first sentence under the heading. A machine lifting an answer takes the top of the section, not the payoff at the bottom. - Be specific enough to be worth quoting. Ranges, timelines, and named conditions are quotable. Adjectives are not. - Keep each answer self contained, so it makes sense lifted out of the page and set beside three other sources. - State the limits. The conditions under which your answer does not apply are frequently the most citable thing on the page. This is the same discipline as writing a good FAQ, applied to the body of the page rather than bolted on at the end. ## How to tell whether it is working You cannot see fan out directly. What you can do is run a fixed set of prompts against the assistants your buyers use, record the answers verbatim, and note which sources get cited for which part of the answer. 1. Write ten questions a real buyer would ask, in their words. 2. Run each one and save the full answer and the citation list. 3. Note which sub-answers came from which domains. 4. Publish or rewrite the pages that answer the fragments you are absent from. 5. Rerun the same ten prompts a quarter later and compare. The comparison is the whole point. Without the first run recorded, the second run is an anecdote. ## What nobody can promise you No one can guarantee an assistant names your business. The systems are not deterministic, they change without notice, and anyone selling a guaranteed placement is selling something they do not control. What you can control is whether your facts are easy to find, internally consistent, and specific enough to be worth quoting. That is the whole job, and it happens to be the same work that makes a page useful to a human reading it. ## How to measure whether AI assistants cite your business 1. Write ten questions a real buyer would ask, in their words. 2. Run each one and save the full answer and the citation list. 3. Note which sub-answers came from which domains. 4. Publish or rewrite the pages that answer the fragments you are absent from. 5. Rerun the same ten prompts a quarter later and compare. ## Questions people ask Is query fan out the same as semantic search?+Related but not the same. Semantic search is about understanding meaning rather than matching strings. Fan out is about decomposing one request into several retrievals. A system can do one without the other, though modern assistants generally do both. Does fan out affect normal Google results too?+Increasingly yes, particularly where an AI overview appears. The overview is assembled from multiple retrievals in the same way, which is why pages that answer a specific sub-question sometimes get cited in an overview while ranking modestly in the blue links. Should I write more pages or better pages?+Better pages first, more pages second. A page that answers five sub-questions properly is worth more than five thin pages, and it avoids the problem of your own pages competing with each other. How is this different from writing FAQs?+An FAQ section is a good start but it is usually treated as an afterthought at the bottom of a page. Fan out rewards putting real answers in the body under real headings, with the FAQ picking up the remainder. Do I need schema markup for this?+FAQ schema helps a machine parse what is already there, but it does not create substance. Structure and specificity in the visible content matter far more than the markup wrapped around them. How long before this matters for a local business?+It already does, quietly. Assistants are answering local service questions now, and the sources they cite were written before anyone was thinking about fan out. Being early here is unusually cheap because so few local sites are contesting it. Next step If you want to see which questions the assistants are already answering about your category, and who they are citing instead of you, that is where the AI visibility work starts. [See how AI search visibility works →](/ai-search-visibility/) ![Chris Johns, founder of Big Brain Digital](/wp-content/uploads/2026/08/bbd26/img/photo-founder.jpg)Chris Johns Founder of Big Brain Digital. More than 15 years across SEO, content operations, ecommerce, and AI search. Based in Boise. [More about Chris](/about/). ## Keep reading [AI search visibilityLLM SEO: how to be cited](/resources/llm-seo/)[AI search visibilityHow to rank in ChatGPT](/resources/how-to-rank-in-chatgpt/)[ServiceAI Search Visibility](/ai-search-visibility/)