Why Your Page Ranks on Google But Isn't Cited in AI Search

A page can rank in Google's top 10 and still never get cited in AI search. AI assistants pick sources through a "query fan-out" process that rewards pages appearing across many related sub-queries, not just the original term. Strong rankings help, but they are no longer enough on their own.

Many UK businesses assume that once they reach page one of Google, AI search visibility follows automatically. The evidence from 2025 and 2026 points the other way. AI assistants and AI Overviews increasingly cite pages that do not rank well, and ignore pages that do. Understanding why is the first step to fixing it.

Does ranking higher on Google get you cited in AI search?

Not reliably. Ranking helps, but the correlation is far weaker than most people expect.

In an Ahrefs analysis of 15,000 prompts, only 12% of links cited by ChatGPT, Gemini and Copilot appeared in Google’s top 10 for the same prompt. Around 80% of those citations did not rank anywhere in Google for the original query at all. Perplexity was the outlier, with closer to one in three of its citations pointing to top-10 pages.

The picture inside Google’s own AI Overviews has shifted too. Ahrefs first found that 76% of AI Overview citations pulled from the top 10, then later reported that figure had dropped to around 38%, with the rest coming from pages ranking 11 to 100 or beyond. The conclusion is consistent across both studies: a top ranking improves your odds but is closer to a coin flip than a guarantee.

What is query fan-out and why does it change everything?

Query fan-out is the process where an AI system splits one search into several related sub-queries, runs them, and then cites the pages that show up most often across all of them.

Google has confirmed its AI systems perform this fan-out. Instead of citing whatever ranks for your exact target keyword, the AI assembles an answer from sources that recur across the broader topic. A page optimised narrowly for a single head term can rank well yet appear in almost none of the fan-out sub-queries, so it never enters the citation pool.

This is why broad topical coverage now matters more than a single well-ranked page. If your content only addresses the obvious query and ignores the adjacent questions, the fan-out process simply routes around you. Our guide to AI search optimisation covers how to map that wider topic cluster.

Why do AI engines cite pages that rank poorly?

Because AI citation selection runs on different signals to classic ranking. Relevance to a sub-query, clear entity definitions, and the presence of a directly quotable passage often outweigh raw position.

Ahrefs’ study of brand visibility across ChatGPT, AI Mode and AI Overviews found that branded web mentions and YouTube presence correlate strongly with AI visibility, with YouTube mentions showing the highest single correlation in their dataset. None of those signals are the classic backlink-and-on-page mix that drives Google rankings. A page can be weak on traditional SEO yet rich in the signals AI engines lean on, and get cited anyway.

The same study noted that Wikipedia is the most cited single source inside ChatGPT, ahead of Reddit, Forbes and G2. These are reference and discussion sources, not the commercial pages that usually win the SERP. If your strategy only targets the SERP, you are competing for the wrong real estate.

How do you get cited when ranking is not enough?

Treat AI visibility as a separate discipline that overlaps with SEO rather than a free by-product of it.

• Cover the full topic, not just the head term, so your pages surface across fan-out sub-queries.
• Write clear, self-contained passages that an AI can lift and attribute without ambiguity.
• Strengthen entity signals: who you are, where you operate, and what you are an authority on.
• Build the off-page signals AI engines actually weight, such as consistent brand mentions and video.

These are the levers that distinguish AI search work from traditional optimisation. If you want to see where your business currently stands, our free AI visibility audit checks how you appear across ChatGPT, Gemini, Perplexity and AI Overviews, and our comparison of AI search agencies shows which specialists focus on this work.

What this means for UK businesses in 2026

The decoupling of rankings and citations is now well documented rather than speculative. A page-one position remains valuable, but it is the start of AI visibility work, not the finish. Businesses that keep measuring success by Google rank alone will keep losing AI citations to pages they out-rank.

The practical takeaway is to audit your topic coverage and entity signals separately from your keyword rankings, and to accept that the two scoreboards no longer move together.