Google AI Overviews cite brands, but recommend rivals in most cases
Google AI Overviews cited brands’ own listicles, then picked rivals in 69% of cases, exposing a sharp gap between inclusion and endorsement.

Google AI Overviews can quote a brand’s own content and still steer the user to a competitor, and that is the part marketers keep underestimating. In a study of 100 B2B “best [category] software” queries, Lily Ray found that Google cited self-promotional listicles but left the original brand out of the recommendation in 69% of cases.
Ray checked those queries on April 15, May 15 and June 8, a three-point snapshot that showed the same pattern again and again: being in the source set did not mean being chosen as the answer. For brands that have treated “we got cited” as proof that their AI search playbook is working, the finding is a warning shot. Google is happy to pull a brand’s own listicle into AI Overviews, but that same system can still decide that a rival deserves the actual nod.

That distinction matters because AI Overviews are not a simple list of links. Google’s own guidance says the feature gives users an AI-generated snapshot with links to dig deeper, which means the answer layer can both summarize and sort. Google launched AI Overviews to everyone in the United States on May 14, 2024, after saying the feature had already been used billions of times in Search Labs and would reach hundreds of millions of users that week. Google also said it expected more than 1 billion people to have access by the end of 2024.
The company has since said it is updating AI search so users can find original content and trusted sources more easily, using source preferences, labels and carousels to highlight original reporting and helpful discussions. That makes Ray’s finding even more pointed: the game is no longer just about getting included, but about earning enough credibility that the system trusts a brand as the recommendation, not merely the citation.
Ray, who is vice president of SEO Strategy and Research at Amsive and founder of Algorythmic, has repeatedly focused on the risks of GEO and AI-search manipulation tactics. Her latest finding fits a broader pattern around self-promotional “best of” pages and the possibility that Google is weighting trust and review quality more heavily than a brand’s own sales copy. For publishers and B2B teams, the lesson is blunt: citation visibility is not recommendation visibility, and the commercial prize sits with the second one.
This article was produced by Prism’s automated news system from verified source data, official records, and press releases, then run through automated quality and moderation checks before publishing. The system is built and supervised by the people who set the standards it runs under. Read our full AI policy.
Did this article answer your question?


