A lot of what's being sold as AEO or GEO at the moment is really just good SEO with a new name on it, as you're still optimising a site for what people are searching for and they're just searching in a different way. Query fan-out is one of the few bits that I'd say is genuinely AI-specific and falls outside traditional SEO. When you ask an AI search tool like Google's AI Mode a question, it breaks your prompt into several related sub-questions, searches for each one and then puts what it finds together into one answer.
I've written separately about where SEO, AEO and GEO overlap, so in this post I'm sticking to fan-out.
What Is Query Fan-Out?
Query fan-out is how AI search tools like Google's AI Mode and AI Overviews break one question down into several related sub-questions. They run a search for each of those at the same time and then pull what they find together into one answer.
It's Google's own term, so I like to go straight to the source. Its guide to AI features and your website says both AI Overviews and AI Mode may use a "query fan-out" technique, issuing multiple related searches across subtopics and data sources to develop a response. When it announced its AI Mode updates in May 2025, Google described this as breaking down your question into subtopics and issuing a multitude of queries simultaneously.
AI Overviews are the AI summaries that show up on some normal results pages (I cover those in my post on AI Overviews), while AI Mode is the more conversational version that arrived in the UK on 28 July 2025 as a tab on the Search results page and in the Google app.
To give you an example, say someone types "can I run a small bakery from my home kitchen in the UK?" into AI Mode. Google doesn't show you the searches it runs, so this is just my own illustration, but the sub-questions could look something like this:
- registering a home food business with the local council
- food hygiene ratings for home bakers
- insurance for a home baking business
- selling and delivering cakes you've made at home
- whether a mortgage or tenancy lets you run a business from home
It then picks out the passages that best answer each of those from the pages it finds and builds one answer out of the lot. So if you're a home baker with a page about how you got your hygiene rating, that page could be pulled in for one sub-question even though it would never show up for the main search.
What Query Fan-Out Changes
With a traditional search the main thing I'd look at is whether your page ranks for the phrase someone typed in. Fan-out adds another layer on top, as the AI is also looking for pages that answer the smaller questions around the main one, so how well your content covers those sub-questions now matters as well as the phrase itself.
That's also why a page can be cited without ranking well for the main search, like the baker's page above, as each sub-query retrieves its own set of results. Google's AI features guide says its models identify more supporting web pages while a response is being generated, which lets it show "a wider and more diverse set of helpful links" than a classic web search.
The clearest numbers on this come from a study by a tool vendor, so keep that in mind. Ahrefs analysed 863,000 keywords and around 4 million AI Overview URLs and found only 38% of cited pages ranked in the top 10 for the same query, down from 76% in its own July 2025 study. Ahrefs has said its parsing improved between the two studies and now picks up citations it used to miss, so the figures can't be compared directly and I'd treat the drop as a rough direction.
My post on what AI SEO is covers the wider picture.
What Query Fan-Out Does Not Change
The obvious reaction is to write a page for every sub-question you can think of, and Google has warned against exactly that. Its guidance on optimising for generative AI features, last updated in July 2026, accepts it might be tempting to create separate content for every variation of how people search, including fan-out queries, but says doing so "primarily to manipulate rankings or generative AI responses in Google Search violates Google's scaled content abuse spam policy". It also says, fairly bluntly, that a high quantity of pages doesn't make a website higher quality or more relevant to users.
Everything else in SEO carries on as normal. Google's AI features guide says there are "no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary", and any page an AI answer draws on still has to be crawled and indexed first. That's why I'd still start with technical SEO, along with genuinely useful content and links from sites people trust.
There's also advice going around among SEOs to treat fan-out queries as a fresh batch of keyword targets, each with its own new page or section built around the phrase. I tend to go about it the other way round. When I look at a page with fan-out in mind I try to work out which sub-questions someone reading it would genuinely want answering next, and those go on the existing page as sections or FAQ answers. If one turns out to be a separate topic with enough demand of its own it might deserve its own page, which is the same call I'd make in normal SEO anyway.
What Query Fan-Out Does to Your Search Console Data
Something I've noticed in Search Console is that AI Mode's fan-out logs impressions for searches that never turn into a click. I don't often see it mentioned in posts about fan-out, and if you're used to judging a page on its click-through rate the chances are you'll misread it.
Search Console's help page on how impressions, position and clicks are counted says clicking a link to an external page in AI Mode counts as a click and that standard impression rules apply, and every link in an AI Overview gets the same single position. Google's AI features guide also confirms this data sits in the normal Performance report under the Web search type. So if your page is one of the sources in an AI Mode answer it can log an impression, and if the answer on screen was enough for the person searching there's no click to go with it.
The independent figures on skipping the click are from the US, so bear that in mind for a UK site. Pew Research Center tracked the browsing of 900 US adults in March 2025 and found they clicked a traditional search result in 8% of visits when an AI summary appeared, compared with 15% when there wasn't one, and they clicked a link inside the AI summary itself in just 1% of visits. That study covered AI summaries on normal results pages, so it doesn't include AI Mode.
So impressions can climb in Search Console while clicks stay flat, which pulls your click-through rate down. That's down to how the data is counted, so before deciding a page has lost ground I'd look at clicks, enquiries and sales over the same period, as impressions and CTR on their own won't tell you much.
Google has also added a generative AI performance report to Search Console, which Google says it had rolled out to all websites worldwide by 31 August 2026. It only shows impressions from AI Overviews and AI Mode and doesn't list the fan-out queries behind them, but if your site has enough AI impressions to appear there you can compare them with your overall Web impressions to see how much of a rise is down to AI features.
I always pull Search Console data into my monthly client reports alongside Google Analytics, as having impressions sat next to enquiries and sales makes it much easier to tell whether it's the reporting or the performance that's changed. If you'd like a hand making sense of your Google Analytics and search data, I'm always happy to have a look.
Frequently Asked Questions
What Does "Query Fan-Out" Mean?
Query fan-out is when an AI search system breaks one prompt into several related sub-searches, runs them at the same time and combines the results into a single answer. Google uses the term to describe how AI Overviews and AI Mode put their responses together.
Is SEO Still Relevant With AI?
Yes. Technical SEO, genuinely useful content and authority still decide which pages AI search tools can find and trust, and fan-out is an extra layer on top of those. The tools that use fan-out are running searches when they answer you, so the SEO groundwork you've already done still feeds into what gets cited.
Should Every Fan-Out Query Get Its Own Page?
No. Google's guidance says creating separate content for fan-out queries primarily to manipulate rankings or AI responses goes against its scaled content abuse spam policy. Generally speaking, most sub-questions sit better as sections on one well-built page, and it's only a genuinely separate topic with its own demand that needs its own URL.
How Can I See the Fan-Out Queries for My Own Content?
Not directly. Search Console doesn't list the sub-queries AI Mode runs and Google doesn't show them in AI Mode either. Third-party tools can estimate likely sub-queries, but an estimate is a different kind of evidence from a query a platform actually ran, so I'd keep the two separate.
If you'd like some help working out which sub-questions belong on your pages, or what your Search Console data is really telling you, that's part of my AI SEO work, which builds on the ongoing SEO work that gets your pages found and trusted. Just drop me a message to set up a free discovery call.
