Why doesn't my page appear in Google's AI Overviews?

It used to be that if you ranked on page one, you were the raw material an AI Overview drew from. That link is weakening fast. In Ahrefs' Brand Radar analysis of 863,000 search results and four million AI Overview URLs, about 38% of the pages cited in AI Overviews also ranked in Google's top 10 for the same query. A year earlier the same measure was roughly 76%. In other words, Google is now selecting most of its cited sources from somewhere other than the front page of your exact search.

Part of what changed is the engine underneath. As of January 2026 AI Overviews run on Gemini 3, a newer model than the versions that powered them through 2025, and it appears to lean harder on the mechanism below.

Google has confirmed that when a search triggers AI, its system performs a "query fan-out." Instead of answering your exact words, it expands the query into a set of related sub-questions, runs searches for those, and assembles the answer from what it finds. The pages that recur across those sub-query results are the ones most likely to be cited.

That reframes the whole task. You are no longer optimizing a page to rank for one keyword; you are trying to be a source that shows up across the cluster of questions a topic implies - the definitions, the comparisons, the how-tos, the edge cases a buyer would ask on the way to a decision. A page that answers only the headline query, and none of the questions around it, can rank well and still never surface in the fanned-out set the Overview is built from.

The move is from thin and keyword-shaped to thorough and question-shaped. Cover the topic the way the fan-out does: answer the main question cleanly up top, then resolve the sub-questions that surround it on the same page or across a tight cluster, each in a passage an engine can lift on its own. That is the craft of answer engine optimization, and it is exactly what makes a page eligible across many sub-queries rather than one.

One specific lever is worth calling out, because it surprises people. Ahrefs reports that YouTube is the most-cited domain in AI Overviews, and that mentions on YouTube are among the strongest factors correlating with AI Overview visibility in their research. A video can be pulled into an Overview even when your own page is not ranking for the query, so for many businesses video is an underused way to enter the cited set. And none of this happens if the crawler cannot reach you in the first place: AI Overviews are built on Google's regular index, so Googlebot has to be allowed - see how to allow AI crawlers (and note that the Google-Extended token only affects training, not Overviews).

Even once you are cited, do not expect it to hold still. Ahrefs found that AI Overviews change roughly every two days, with the set of cited sources shifting each time. Appearing in one Overview today is not a ranking you have won. Missing one tomorrow means nothing either - each is a single draw from a moving target. The only honest read is your share of citations across many checks over time, which is the subject of measuring AI citations, and the practical how-to is in how to check if AI search mentions your brand. The deeper reason this is a footprint problem and not a single-page fix is covered in AI search visibility.

If you rank but are not cited, guessing which sub-queries you are missing is slow. The free 10-minute AI search audit gives you a structured first read, and the Free Snapshot checks what the engines say about your category today. The SXO Audit goes further: it runs 25 buyer prompts across six engines with three trials each over time, so instead of staring at one Overview you can see which topics already cite you, where competitors are selected instead, and how your citation share moves as you broaden your coverage.

Try the Free Snapshot   See the SXO Audit

By the Web Cited Editorial Research Team. Last updated 1 June 2026.