Contently: Your best-ranked page might be invisible to Google's AI
A page ranking in the top five on Google no longer guarantees that Google’s AI will surface it when generating an answer. That gap between rankings and citations has grown substantially since mid-2025, and Alex Soto’s June 2026 piece on Contently’s Content Strategist blog documents the scale of the shift and what content teams can do about it.
The article opens with data that should reframe how content teams think about SEO success. The overlap between top-10 Google rankings and AI Overview citations dropped from around 76% in July 2025 to roughly 38% by March 2026. Of the remaining citations, about 31% come from pages ranked between positions 11 and 100, and another 31% from pages beyond the first ten results entirely. A page can dominate traditional search for a query and still be absent from the AI answer that most users read first.
Why rankings and citations diverge
The mechanism behind this divergence is Google’s “query fan-out” technique: when a user enters a query, the AI system decomposes it into multiple related sub-queries before generating a response. A page that ranks well for the original query may not appear consistently across those sub-queries. Citations go to pages that surface across the widest range of sub-query variations—pages with depth and consistency on a topic rather than optimization for a single keyword.
Soto describes this as a two-gate structure. Traditional ranking is the first gate: it establishes enough authority for the content to be considered. But clearing that gate no longer means the content will be cited. The second gate requires content that is self-contained, credibly structured, and thorough enough to address the follow-up questions the AI anticipates.
What this requires from writers and editors
The practical changes Soto describes are not about gaming AI systems but about writing more thoroughly. Self-contained sections matter: an AI pulling one paragraph from an article needs that paragraph to stand alone as an answer, with context, claims, and sources intact. Thorough topic coverage matters: articles that address the obvious follow-up questions within the content itself fare better than articles that answer the headline question and stop.
E-E-A-T signals—Experience, Expertise, Authoritativeness, Trustworthiness—need to be made explicit rather than implied. Named credentials, citations to primary sources, and author bylines with verifiable track records all help AI systems identify content as quotable rather than merely findable.
The article also introduces Answer Engine Optimization (AEO) as a distinct practice from traditional SEO, noting that the two are compatible but not identical. Teams optimizing only for rankings are leaving the second gate unaddressed.
Who this is for
The piece is written for content strategists and marketing leaders managing publication programs at some scale—teams with enough output that systematic changes to structure and depth are tractable. Independent writers will find the framing useful even if the operational recommendations are aimed at larger teams.