TLDR
SEO analyst Lily Ray tracked 100 B2B software queries and found Google AI Overviews cited 38 per cent fewer self-promotional vendor listicles in September 2026 than in June. Brands that rank themselves first in their own listicle get excluded from AI recommendations 83 per cent of the time.
What Ray measured
SEO consultant Lily Ray, founder of Algorythmic and VP of SEO and AI Search at Amsive, pulled 100 "best B2B software" queries through Google AI Overviews in September 2026 and compared citation patterns against data she had collected in June.[1] The methodology is hers alone; Google has not commented, and no independent body has replicated the sample.
AI Overviews now cite 38 per cent fewer self-promotional "best of" listicles than they did in June 2026.[1] That is a material shift across four months, and it tracks a sequence of Google policy updates that tightened what qualifies as genuinely helpful content.
The self-ranking backfire
When a brand places itself first in its own vendor listicle, AI Overviews leave that brand out of the recommendation 83 per cent of the time, up from 69 per cent in June.[1] That exclusion rate climbed 14 percentage points in a single quarter, and it is the number worth auditing against your own content inventory.
The competitive consequence is sharper still. In 70 per cent of cases, a self-promotional listicle actively helps competitors get recommended while the publishing brand is omitted entirely.[1] A brand that builds a listicle to own a category ends up routing AI-generated traffic directly to rivals. Ray said that based on recent communications, something significant may be on the horizon.[1]
Google's policy moves
Three Google policy updates in 2026 form the backdrop. On 15 May, Google clarified that its spam policies apply to AI Overviews and AI Mode in Search, bringing AI-surfaced content under the same rules that govern organic results.[4] On 1 October, Google updated its generative AI content guidance to require that publishers manually fact-check and review all AI-generated content for accuracy and trustworthiness before publishing.[2]
On 2 October, Google revised its helpful content guidance to define quality factors as effort, originality, talent or skill, and accuracy, and named fake author profiles a form of deception.[3] Self-promotional listicles, by construction, tend to fail most of those criteria. Ray said nobody outside Google knows when the next major update will arrive or what it will target.[1]
Ray's sample of 100 queries is not large enough to treat these figures as definitive, and no third party has audited the data. Read the numbers as directional indicators from a limited sample that require independent verification before they can support a broader definitive conclusion. The directional shift across four months is consistent with Google's documented policy moves. The competitive inversion, where a brand's own listicle routes readers to rivals in AI summaries, is the pattern B2B marketers should now check against their existing content.
KEY TAKEAWAYS
SOURCES & CITATIONS
FREQUENTLY ASKED QUESTIONS
What did Lily Ray actually measure?
Does this mean vendor listicles are dead for SEO?
What Google policy changes are relevant here?
Jonas Valenti writes about search and how businesses get discovered. He has spent years watching what makes a company visible online, and is unsentimental about tactics that no longer work.
