A page can rank #1 and still go uncited because AI citation rewards a different set of signals than classic ranking: a clearly stated, cleanly extractable answer plus visible source trust. A page that earns the top spot by being comprehensive, well-linked, and authoritative overall can still bury its actual answer inside long prose, never state a clean claim a model can lift, and so get passed over for a page that does. Rank and citation are related but separate games, and because AI systems are changing fast, this is worth re-checking as their behavior evolves.

The core of it is extractability. Classic ranking can reward a page for the sum of its qualities, even when the specific answer is spread across several paragraphs or implied rather than stated. An AI answer, by contrast, needs a discrete, self-contained chunk it can pull and attribute. If your #1 page makes the reader infer the answer, or wraps it in qualifications and narrative, there may be no clean sentence for the system to grab, so it reaches for a source that hands the answer over directly.

Trust is the other half. AI systems lean on signals that the source is credible and the claim is verifiable, and those signals are not identical to the link-and-relevance picture that produced your ranking. A page can be the best overall result and still not present the kind of specific, checkable, plainly-attributed claim that makes a model comfortable citing it.

So the divergence is not a fault in your page or a bug in the system, it is two mechanisms optimizing for different things. Ranking asks “is this the best page for the query.” Citation asks “does this page contain a trustworthy, liftable answer to the specific thing being asked.”

To close the gap, find the exact question your #1 page is meant to answer and make sure it states the answer in one clean, self-contained, verifiable sentence near where the question is raised, rather than assuming your ranking will carry it into the AI answer on its own.