If you've been optimizing for search rank as your primary path to AI-driven discovery, new data suggests that strategy has a significant gap. A study published this week by Marketplace Pulse, based on research from AI-optimization firm Autopilotbrand.com, analyzed 12,810 Alexa for Shopping recommendations across 1,963 non-branded search queries in May and June 2026.
The finding: 63.9% of the products Alexa for Shopping recommended fell outside the organic top-10 for the matched search term. Not a few. Nearly two-thirds.
The number that's harder to dismiss: 40.9% of Alexa for Shopping's recommended products never appeared on the visible search results page at all. They weren't ranking poorly. They weren't buried on page three. They didn't show up in normal search in any visible position, and the AI recommended them anyway.
What the Study Measured
The researchers posed two types of queries to Alexa for Shopping: a best-recommendation question (something like "what's the best queen mattress?") and a bare category search ("queen mattress"). The gap between what the assistant recommended for the former versus what showed up in regular search results for the latter is what the study captures. Asked to recommend rather than list, Alexa for Shopping surfaced a materially different, and often deeper, set of products than the standard search page most brands optimize for.
On paid advertising, the data is equally striking. Only 14.3% of Alexa for Shopping's recommendations were products running a sponsored listing on the matching search page, and 83% of those already ranked organically. The sponsored-only product, the one paying for visibility without an organic rank to match, was largely absent from Alexa's recommendations.
The implication: buying your way onto the search page does not buy your way into the AI's answer.
Why This Matters and What It Doesn't Mean
It's important not to overcorrect from this data. Search rank and paid advertising still matter for the majority of Amazon shopping sessions that don't go through Alexa for Shopping's recommendation engine. The study is measuring a specific surface: what happens when a shopper asks Alexa for a recommendation rather than browsing search results.
That surface is growing. It's not the whole game.
What the data does establish is that the inputs Alexa for Shopping uses to rank and recommend products are meaningfully different from the inputs that drive organic search rank. The traditional Amazon algorithm rewards sales velocity, conversion rate, and keyword relevance built through advertising and listing optimization. Alexa for Shopping, according to this data, is drawing from a different pool.
The exact inputs it weights, Amazon hasn't disclosed. But a product that scores well on listing quality, reviews, and Q&A completeness but hasn't accumulated the sales history to break into organic top-10 can apparently appear in Alexa's recommendations regardless.
That's genuinely new territory. On traditional search, category incumbents have a compounding advantage because sales velocity feeds rank which drives more sales. On the AI recommendation surface, as Marketplace Pulse put it, "rank incumbency is a weaker moat." A challenger brand with a thorough listing and strong review quality can appear where a search-page leader does not. For brands that have been locked out of top-10 organic positions by well-funded incumbents, this is the most significant structural change in Amazon discovery in years.
What to Do With This Information
The practical response isn't to abandon keyword optimization or pull back on advertising. Both still drive real results on the traditional search surface. The response is to build a second content strategy aimed at the AI surface, in parallel with your existing approach.
Based on what we know about how Alexa for Shopping evaluates products, which includes reading listing text semantically rather than matching exact keywords, the elements that matter most are: complete and specific product attributes in Seller Central, A+ Content modules with real text that answers shopper questions, a Q&A section seeded with the questions buyers genuinely ask before purchasing, and review volume that includes detailed use-case language. None of these are new tactics. What's new is the evidence that they drive a different kind of visibility than traditional keyword ranking, and that visibility now has quantified scale behind it. You can read more about how we approach AI-ready listing strategy as part of Amazon brand management on our services page.
One note on scope: this is an early study from a single source, covering a defined query set in May and June. It's a signal, not a verdict. Amazon hasn't confirmed what factors drive Alexa for Shopping recommendations, and the system is actively evolving.
But the directional finding, that search rank has limited bearing on AI recommendations, is consistent with what practitioners have been observing qualitatively for months. Now there are numbers attached to it.
If you want to think through what an AI-optimized listing strategy looks like for your catalog alongside your existing search and ads approach, schedule a call and we'll start with your top-selling ASINs.