A search query used to be just that: one query, one set of results. Now, one complex AI search query can trigger an estimated eight to 20 or more background searches (Google hasn't published exact counts), each producing its own results. These results are scored for relevance, with the strongest ones across all the sub-searches being synthesized into one comprehensive answer.
This is query fan-out in a nutshell, and it’s how AI search systems improve their answers’ usefulness. It’s also a key factor in deciding which pages these systems use to generate answers: pages that address more background subqueries are likely more relevant to the prompt, making AI systems more likely to reference and cite them. Even if they don’t rank #1 for any keywords.
Below, we'll cover how query fan-out works, how to find the subqueries relevant to your brand, and how to optimize for them.
What is query fan-out?
Query fan-out is an AI search system process that splits a user query into multiple subqueries (also known as “fan-out queries”), collects information for each subquery, and then merges relevant information into a single response.
Google popularized the term when introducing Google AI Mode, a conversational AI interface available within Google Search.
In her Google I/O 2025 keynote speech, Head of Search Elizabeth Reid said: “AI Mode isn’t just giving you information — it’s bringing a whole new level of intelligence to search. What makes this possible is something we call our query fan-out technique.”
“Now, under the hood, Search recognizes when a question needs advanced reasoning. It calls on our custom version of Gemini to break the question into different subtopics, and it issues a multitude of queries simultaneously on your behalf.”
AI search systems use query fan-out to enhance their responses. In fact, Google’s official search documentation mentions query fan-out as a core technique for retrieving content from its search index to generate AI answers.
How query fan-out works
Query fan-out works by identifying a prompt’s intents, retrieving content that fulfills them, and then synthesizing an answer from the most relevant results.
Broadly, this model comprises five components:
- Analysis of the prompt to understand its intents, or what the user is looking for. For example, the prompt “What’s the best laptop for a college student who needs long battery life and does some video editing?” indicates that the user is looking for a laptop that’s suitable for college use, has a long battery life, and can support video editing work. AI thus needs to generate an answer that fulfills all three intents.
- Decomposition, or breaking down, of the prompt into multiple subqueries that address the identified intents. AI might split our example prompt above into subqueries like “best laptops for college students,” “laptops with long battery life,” and “video editing laptop requirements.”
- Retrieval, where the AI system searches sources like proprietary and search engine indexes for content that addresses any of the prompt’s subqueries. The system may also subsequently conduct more searches based on what it discovers from initial results.
- Scoring of all retrieved content based on how well it addresses each subquery, to identify and rank the results most relevant to the prompt.
- Synthesis, where the AI system uses the highest-scoring content to generate its answer
The implication? Pages that address the intents behind multiple subqueries are more likely to be referenced and cited by AI. Addressing just one overall intent, corresponding to one target keyword, is no longer enough.
Types of fan-out queries
Types of fan-out queries include reformulation, implicit, comparative, recency, contextual variation, and next-step. Here’s what they mean and query fan-out examples for each.
| Fan-out query type | Definition | Original prompt | Fan-out query example |
| Reformulation | Rephrasing of the original prompt to account for how users can express the same intent in many ways | “set up a Google Business Profile” | “create a Google Business listing” |
| Implicit | Identifying underlying needs that users don’t expressly mention in their prompt | “wheelchair-friendly tourist attractions” | “tourist attractions with elevator access” |
| Comparative | Assessing two or more things against each other | “standing desk options” | “electric vs manual standing desks” |
| Recency | Queries where having the latest, most-updated information is key | “f1 race dates” | “f1 race dates 2026” |
| Contextual variation | Modification of the prompt to account for users’ personal characteristics | “gyms with childcare” | “gyms with childcare in [user’s city]” |
| Next-step | Addressing users’ subsequent needs after they get the answer to their initial query | “how to register a trademark” | “trademark registration lawyer” |
One caveat about these queries: they're synthetic and probabilistic. AI systems generate them on the fly, they shift from one run to the next, and most carry little or no search volume of their own. Treat them as intent signals to cover, not exact phrases to target.
All these fan-out query types point to specific user intents, like wanting the latest information (recency queries) or information tailored to their situation (contextual variation queries). Being aware of these intents and fulfilling them in your content are key for optimizing for query fan-out in AI.
Why optimizing for query fan-out matters for search engine and AI visibility
Optimizing for query fan-out matters for search engine and AI visibility because brands whose pages address users’ intents more thoroughly may gain higher search engine visibility and more AI mentions and citations.
Brands that cover relevant subqueries stand to feature in Google's AI Overviews, which may appear at the top of the search engine's results pages. Our AI search traffic study found this visibility increasingly shapes where discovery starts. Just like how this Google AI Overview prominently recommends Samsung and Apple's phones in response to the search query "best phone for concert videos":
The AI Overview recommends these brands, citing websites like Digital Camera World in support, even though none of their pages rank #1 on the search results.
AI search platforms may also reference and cite a brand’s pages more often when their content consistently appears relevant to fan-out queries.
Our experience updating four blog articles to target relevant fan-out queries points in this direction, though the effect was directional rather than guaranteed. Within a month, these articles' citation counts for our tracked prompts more than doubled, from two to around five. The climb was volatile, spiking as high as nine before settling, and brand mentions actually dipped over the same period. We treat this as encouraging early evidence, not a settled result.
How to find fan-out queries
To find fan-out queries your content doesn’t cover yet, use tools that run your target prompts in AI systems to surface these prompts’ possible subqueries. We share two of these tools below.
Once you have a list of subqueries, strike out those your content already addresses, so you can focus on optimizing for the rest.
AI systems generate responses probabilistically. So, the same prompt can produce different fan-out queries each time you run it. We suggest identifying trends in subquery intents rather than trying to nail down specific subquery wording.
1. Semrush
Semrush is a search visibility platform that helps brands track and improve their prominence in search engine results and AI answers. The platform’s AI Visibility Toolkit approximates the subqueries that AI systems may generate for your target prompts, while Enterprise AIO shows Google’s actual fan-out queries at scale.
To find fan-out queries with AI Visibility Toolkit, enter your target topic into the Prompt Research report and click “Analyze.”
The “Topics” tab of the report’s “Related topics” section shows topics related to your prompt. Click any topic to view its relevant sub-prompts, which AI systems may also use as fan-out queries.
To view Google’s actual fan-out queries for prompts, use Enterprise AIO’s Query Fan-Out Analysis AI Automations.
2. Query Fan-Out Tool
Queryfanout.io is a free tool that identifies subqueries by replicating Google's query fan-out process for your target prompt.
Other options for finding fan-out queries include ChatGPT Query Fan-Out Tool, a free query fan-out generator for Chrome browsers.
How to optimize for query fan-out
Once you’ve identified relevant fan-out queries, use these steps to optimize your pages for them:
1. Identify core topics
Identify core topics to build your AI visibility around, so you focus your query fan-out optimization efforts more effectively.
Start with topics directly related to your brand and what you offer. Doing this helps you:
- Control how AI systems portray your brand in their answers
- Show up during customer journey stages where visibility and influence matter most
- Leverage your authority since you’re an expert in these topics
Identify your brand’s most important topics with the Questions report in Semrush’s AI Visibility Toolkit. To do this, enter your domain into the report. Then, view the topics in the “Topic Distribution” section.
Next, view the report’s “Intent by Topic” section to learn the dominant intent for each topic. This information is helpful for mapping the topics to customer journey stages.
After identifying brand-related topics, expand your research to cover related topics that align with your brand’s expertise. Prioritize these related topics based on your business goals and audience interests.
For example, at Semrush, we publish content that covers not just our search visibility tools, but also broader digital marketing topics.
2. Plan topic clusters
Plan the topic clusters you’ll publish content around to cover your identified fan-out queries.
Topic clusters are groups of interlinked webpages that collectively cover a core topic in depth. They're made up of a central pillar page, which provides a broad overview of the core topic, and several cluster pages, which cover relevant subtopics in multiple subsections, including ones that address fan-out queries.
Topic clustering can encourage AI systems to prioritize mentioning and citing your content. It builds topical authority while directly addressing the subqueries these systems may generate during query fan-out.
You can create a mind map like this to plan your topic clusters:
If you need help identifying subtopics, use Semrush's Keyword Strategy Builder. Enter your core topic (like a topic you found with the AI Visibility Toolkit), and the tool groups the related queries into topic clusters, with a suggested page for each one.
Where relevant, use those clusters as your cluster pages' topics, and the queries grouped under each one as subsections within those pages.
Then, map your identified fan-out queries as new or existing cluster page subsections. If your fan-out queries don’t fit within any existing cluster page, start new ones for them.
3. Write NLP-friendly content
Write natural language processing (NLP)-friendly content for your topic cluster pages so that AI systems can better understand how their content fulfills relevant fan-out query intents.
NLP is a branch of artificial intelligence that helps computers process, understand, and generate text in human language. Writing in ways like these can improve how AI systems understand your content:
- Write in chunks. Chunks are self-contained, meaningful sections of content that can stand on their own, letting AI systems process, retrieve, and summarize them more easily. Write in full sentences, and restate context where helpful.
- Provide definitions. When you introduce a new concept, provide a clear and direct definition of it. Doing this helps AI systems understand what you’re talking about, especially when they use a fan-out query like “[topic] definition” to get a definition.
- Structure content effectively. Add descriptive subheadings to break your content into sections, and use heading tags to show their hierarchy. Structuring your content like this helps AI systems identify content related to highly specific subqueries. You can also use tables and lists to create easily parsable information. Our guide to optimizing content for AI search engines goes deeper on each of these.
- Use clear language. Avoid jargon, overly complex sentence structures, and unnecessary fluff. Writing clearly makes it easier for AI systems to understand your content and extract valuable information.
4. Apply schema markup
Apply schema markup to your pages to help AI systems parse and extract their content for relevant subqueries.
Schema markup is a method of formatting data in your pages' HTML to signal to machines, like search engines and AI systems, the types of content these pages contain.
For example, you can use Product schema markup to add machine-readable labels to a product’s name and size:
<script type="application/ld+json">
{
"@context": "https://schema.org/",
"@type": "Product",
"name": "Nike Air Force 1 '07 Men's Shoes",
"size": "US 8"
}
</script>With these labels, AI systems can better identify your product’s name and size for addressing subqueries about either attribute.
Check out our guide to schema markup to learn more about the available schema markup types and how to implement them on your pages.
5. Build your off-site presence
Build your brand's presence on third-party sites, because AI systems search the wider web for answers to fan-out queries, not just your site. A larger off-site footprint gives these systems more surfaces to discover your brand's relevance to subqueries, and more chances to mention you in responses.
Increase your off-site presence by taking steps like these:
- Set up directory listings: Create a profile for your brand on directories relevant to your industry. Examples of directories are G2 for software businesses and Yelp for restaurants. If you run a local business, claim your Google Business Profile as well, since it’s a key information source for AI systems.
- Run digital public relations campaigns: Secure coverage of your latest achievements, products, or findings, on reputable industry publications. This coverage feeds AI systems fresh answers to subqueries about your brand’s credibility and latest news.
- Participate in community forums: Participate in community forums: Provide helpful responses on Reddit, Quora, and other forums, especially where discussions relate to common subquery intents, to encourage AI systems to reference your input in answers.
6. Optimize for commercial queries
Optimize for commercial queries by addressing the fan-out queries AI systems generate about your offerings' characteristics, like their color, model, price, free plan availability, or operating hours, as relevant. Doing this helps AI systems recommend your offerings in response to relevant prompts more often.
Tactics for optimizing for commercial fan-out queries include:
- Address potential fan-out queries directly by adding relevant content to your product and service pages. For example, if you’ve identified “[product name] refund policy” as a potential fan-out query, you could add a question like “What is the refund policy for [product name]?” to your product listing’s FAQs section, followed by your answer to it.
- Publish comparison content to shape the AI narrative about how your offerings stack up against competitors’ in response to fan-out queries like “[your product] vs [competing product] review”
- Get more user reviews, as they help corroborate information about your offerings, like whether they’re true to size. Positive review language can also influence AI systems to describe your offerings more favorably. These effects matter most when users are actively researching whether your offering meets their needs.
7. Check your subquery coverage
Check your subquery coverage to identify how well your content addresses your target fan-out queries’ intents, so you can take steps to improve coverage.
List your pages, and then map them against the topic clusters and fan-out queries you’ve identified earlier. Flag fan-out queries that your pages currently don’t address.
Next, check the quality of content on pages that do address fan-out queries. Your content’s quality affects whether AI systems use your pages to generate answers, so flag pages whose content appears thin or irrelevant to your target subqueries.
Finally, plug your identified gaps. Create new pages for fan-out queries you lack coverage for. Also, update existing pages to provide more direct, relevant answers to subqueries that need stronger coverage.
Repeat this coverage audit, and search for new fan-out queries, every quarter. Regular maintenance will help you pick up and fill new content gaps as your audience’s preferences evolve.
Bonus: Mini case study
As a bonus, here’s a mini case study of how Stripe’s marketing efforts demonstrate many principles of query fan-out optimization.
The brand’s site has solutions pages tailored to different business stages, business models, and use cases. In turn, these pages have subsections that provide direct, detailed information on relevant subtopics.
This detailed and varied information likely helps AI systems recognize Stripe’s relevance to various intents and extract useful information for fan-out queries.
The Stripe website also covers relevant topics through resources like its blog, customer stories, support center, and newsroom.
In the guide below, Stripe uses clear structuring to break down a complex topic. And provides direct, easy-to-understand explanations throughout.
Beyond its website, Stripe often features in press coverage from reputable industry publications and news sites, which helps expand its visibility on the wider web.
Stripe’s AI search visibility significantly outperforms competitors’, according to data from Semrush’s AI Visibility Toolkit. Several factors likely explain this, but the breadth and depth of the brand's on-site content, plus its extensive off-site presence, could have played an especially important role.
Measure your fan-out coverage
Measure your fan-out coverage with a tool like Semrush’s AI Visibility Toolkit, which reports on your brand’s visibility and narrative in AI systems.
Enter your domain into the tool, and its Narrative Drivers report will show your share of voice for non-branded queries across AI search platforms. In other words, how often these platforms mention you as opposed to (or alongside) your competitors for queries that don’t contain names of brands.
You can even see if your brand is mentioned first, second, or further down in various AI systems’ responses to specific prompts.
The “Key Sentiment Drivers” section of the tool’s Perception report provides insight into your brand’s portrayal in AI responses, too.
Work to emphasize strengths and mitigate weaknesses, so you generate more positive coverage in AI responses. And ultimately attract more customers.
FAQs
Does query fan-out apply to ChatGPT and other LLMs, or just Google AI Mode?
Query fan-out applies to ChatGPT and many other LLMs, not just Google AI Mode. Google coined the term to describe its information retrieval technique for AI Mode, and the industry has since adopted "query fan-out" as a general term for the equivalent process in LLMs like ChatGPT and Claude. None of the companies that develop these LLMs have shared their own name for it.
How many subqueries does AI run per prompt?
The number of subqueries AI runs per prompt depends on the prompt’s complexity and how powerful the AI’s model is. An AI system might run just one subquery for a simple prompt, but fire off anywhere from 8 to 20+ subqueries for a complex one, especially if you enable the system’s deep research mode.
Are fan-out queries the same as long-tail keywords?
No, fan-out queries aren’t the same as long-tail keywords. While both may look similar, long-tail keywords are the product of real human search behavior, being phrases that people have typed into a search bar. In contrast, fan-out queries are synthetic. AI systems generate them on the fly to retrieve a larger number of relevant results for building fuller answers.
How do I find the fan-out queries for a topic?
To find fan-out queries for a topic, use tools that can run prompts related to it, and then surface their possible subqueries. For example, Semrush’s AI Visibility Toolkit approximates the subqueries that AI systems may generate for your target prompts, while Enterprise AIO shows Google’s actual fan-out queries at scale.
Does schema markup help with query fan-out?
Schema markup supports query fan-out by adding machine-readable labels to data on your pages, which helps AI systems parse and extract it accurately. It isn't a citation lever on its own. Types of data you can add schema markup to include product and service information, and frequently asked questions and answers.
How does query fan-out affect ecommerce and product pages?
Query fan-out affects ecommerce and product pages by determining the pages AI systems surface when they break commercial queries into subqueries about specifics like color, model, and price. These systems are more likely to reference, and generate answers using, pages that address these specifics clearly.
