Google Cited Your "Best Of" Page — Then Recommended Your Competitors

Picture this: you've spent months building an authoritative "10 Best CRM Tools" page. You ranked your own product at #1, backed it up with genuine comparisons, and optimized it for AI Overviews. Then one day you check Google and see something that makes your stomach drop — your page is right there in the AI Overview citations. Your hard work paid off.

But the AI Overview itself? It's recommending three of your competitors. Your product doesn't appear anywhere in the actual answer.

This isn't a hypothetical. It's exactly what Lily Ray found in a study circulating this week, and it has real implications for anyone who's been building "best of" listicles as a GEO (Generative Engine Optimization) tactic.

What Lily Ray's Study Actually Found

Ray — VP of SEO Strategy at Amsive and one of the most data-driven analysts in the industry — noticed something odd: Google had changed how it handles "best [product category]" queries. Brands that ranked their own product at the top of their own listicles were still getting their pages cited in AI Overviews. But Google was ignoring those self-serving rankings and generating its own recommendations instead.

The headline stat: in 69% of cases, when Google cited a brand's own "best of" listicle in an AI Overview, the brand itself was not included in the AI's actual product recommendations. Google used the page as a source while effectively overruling its top pick.

That last stat is the context that makes this sting. When an AI Overview shows up, organic clicks drop roughly 40%. So you don't just lose the AI recommendation — you also lose clicks you would have gotten from ranking organically. You get cited and skipped simultaneously.

Why Google Is Doing This — And Why It Makes Sense

Once you understand how Google's AI processes these queries, this behavior is completely logical. It's just not what the SEO community was hoping for.

Google doesn't use your listicle as a direct answer. It uses your listicle as one signal among many to understand the landscape of a topic. Your "best CRM" page tells Google which products exist in that category, what criteria matter, and what the competitive set looks like. Google then synthesizes that with dozens of other sources — review sites, user forums, third-party comparisons, independent publications — to generate its own opinion.

If your page ranks "your own product #1" but most other sources rank it #4 or #5, Google is going to give more weight to the consensus. It cites your page because it found value in your category breakdown. But it's not going to repeat your self-promotional ranking.

This also lines up with what Google has been saying about AI Overviews for years: they're designed to synthesize information, not amplify any single source's perspective. The problem is that a lot of brands were operating under the assumption that "if my page gets cited, my product gets recommended." That was always an optimistic read of how this works.

The GEO Tactic That Quietly Backfired

For roughly 18 months, one of the most popular GEO strategies was publishing self-serving "best of" lists. The logic was clean: if AI systems learn from web content, then a page explicitly saying "Product X is the best" should increase Product X's recommendation probability. And for a while, with some AI tools, it seemed to work.

The problem is that Google's AI Overviews have gotten considerably better at detecting self-promotional bias. Lily Ray believes Google made a specific algorithmic change to how it handles these queries sometime in early-to-mid 2026 — essentially adding a filter that says: "if this page is published by the company whose product appears at the top of this list, discount that self-ranking significantly."

The tactic worked early on because AI systems were naive about brand ownership. Now they're not. And the SEO community is left holding a bunch of "best [our product]" pages that actively funnel traffic toward competitors in AI results.

What to Do Instead — A Practical Response

There's no perfect fix here. Google is going to recommend what Google thinks is best, and that's ultimately the right call even if it's frustrating for brands. But there are concrete things you can change.

Stop building lists where you rank yourself #1

If you're going to publish comparison content, make it genuinely unbiased. Include your product honestly, give it fair treatment, and give competitors credit for things they actually do well. This isn't just moral high ground — it's a signal to Google that your content is trustworthy rather than promotional. Pages that acknowledge real tradeoffs get cited and trusted differently than pages that look like product brochures dressed as comparisons.

Shift focus to third-party citation signals

Instead of trying to rank yourself in your own content, focus on getting third parties to rank you in theirs. That means PR strategy, working with independent reviewers, building strong product profiles on G2, Capterra, and Trustpilot, and making sure your entity information is consistent across every platform Google can find. When independent sources put you at #1, Google will cite them — not you — and that works in your favor.

Build pointed comparison content

There's still real value in comparison content — just not self-promotional broad lists. Build pages like "[Your product] vs. [Competitor]" where you can be honest and detailed about actual differences. These pages tend to attract high-intent searchers and get cited in AI Overviews specifically because they address a pointed comparison question, not a general "best of" query.

What "Good" AI Overview Presence Actually Looks Like

The sites getting cited most consistently in Google AI Overviews share a few patterns that have nothing to do with writing lists that promote themselves.

• They answer specific questions directly. AI Overviews are triggered by questions, and the sources cited tend to be pages that answer clearly and quickly — not pages that bury the answer in promotional copy.
• They have clear entity markup. Schema.org Organization and Product markup helps Google understand exactly what your brand is, what you do, and how you relate to products being discussed.
• They're cited by other sources, not just themselves. The most reliably cited pages are ones that third-party publishers reference. A well-placed mention in Search Engine Land does more for AI citation probability than ten self-promotional listicles.
• They publish use-case specific content. "Best CRM for solo consultants" gets AI-cited more reliably than "Best CRM" — the specificity matches the actual question Google is trying to answer. Broad lists invite Google to synthesize; specific answers invite Google to cite directly.
• They update content regularly. Google's AI has a strong recency preference for competitive categories. Pages with visible update dates and genuinely refreshed content get favored over static lists.

How to Actually Track Whether This Is Affecting You

The most frustrating part of this situation is that most brands don't even know it's happening. You'd have to manually search dozens of queries, note which sources Google is citing, and cross-reference them against whether your brand appears in the actual recommendations. That's not a sustainable workflow.

The smarter approach is to audit your AI visibility systematically:

The Bigger Lesson Here

This whole situation is a clean illustration of something worth saying plainly: GEO is not just SEO applied to AI. The mechanics are different enough that tactics that worked in one context actively backfire in the other.

In traditional SEO, you could publish a self-promotional page, optimize it well, build some links, and it would rank and convert. The ranking was the signal. In AI Overviews, the citation is separate from the recommendation. Google can mine your page for context while completely disregarding your preferred answer. You don't control the output — you only influence the inputs.

That's a genuinely different strategic reality, and the brands adjusting fastest will have a real advantage. Not because there's a clever hack, but because building content that earns third-party trust at scale is now the clearest path to consistent AI recommendations.

The brands still publishing lists that put themselves at #1 in 2026 aren't just wasting effort. They're handing Google a free category reference that routes customers to their competitors. Stop helping Google build that resource for free.