A brand can win an AI Overview citation and still hand the sale to a competitor. That is not a hypothetical — it is happening on a majority of “best software” queries right now, and the mechanism behind it explains why so many comparison pages are quietly working against the brands that publish them.
For years, software companies built “best tools in category X” listicles that ranked their own product first. It was cheap, scalable, and it worked in classic search. In AI search, that same tactic can backfire: Google’s AI Overview pulls the listicle as a source, then recommends a different product from inside that brand’s own ranked list.
The data: cited does not mean recommended
SEO researcher Lily Ray tested this directly. In research published in June 2026, she analyzed 100 B2B “best software” queries in Google’s AI Overviews, checking each one three times between April and June.
Across the 80 queries that produced an AI Overview, self-ranked listicles were cited 323 times. In 224 of those instances, Google named the brand’s own page as a source, then recommended a competitor that was ranked somewhere inside it. Put plainly: when a brand’s own listicle earned the citation, that brand lost the recommendation 69% of the time.
Citation and recommendation are not the same outcome
AI search produces two distinct results, and only one of them drives revenue. A citation means the engine named your page as a source behind its answer. A recommendation means the answer told the reader which product to actually choose. Buyers act on the recommendation, not the citation — and a citation is easy to mistake for progress because the brand name still shows up on screen.
What gets cited depends largely on the content of the page itself. What gets recommended depends on what the rest of the internet says about the brand: how many independent sites mention it, link to it, and review it. Ray’s data shows that recommended brands consistently had far more referring domains and far more mentions across AI Overviews and ChatGPT than brands that were cited and then passed over. On-page edits do not close that gap, because the gap is not on the page — it lives in how much of the web covers the brand independently.
One category example makes the pattern concrete: for “best LMS for selling courses,” Google repeatedly cited a listicle from Oasis LMS, which ranked itself first. Google then recommended Kajabi, Thinkific, LearnWorlds, and Teachable instead — all four ranked somewhere inside the Oasis piece it had just cited. Ray found the same split across CRM, help desk, and SEO software categories.
How to audit your own AI search position
You do not need specialized tools to run this check for your own category. The goal is to stop treating citations and recommendations as one metric and start measuring them separately.
- Build a query list. Use the phrases a buyer would actually type: “best project management software,” “[competitor] alternatives,” “best [category] software.”
- Record citations and recommendations separately. For each query, log the pages Google cites as sources and, separately, the products the answer actually recommends.
- Repeat every query. Run each one more than once — AI answers shift between sessions, so a single snapshot is unreliable.
- Score share of recommendations, not share of citations. The number that predicts revenue is how often you’re recommended, not how often you’re named as a source.
- Extend past Google. Ray’s research documents the pattern in Google’s AI Overviews specifically, so start there, then run the same query list through ChatGPT and Perplexity to see which publishers those engines lean on for your category.
Recommendations come from coverage you don’t publish
Ray’s data also shows where AI recommendations actually originate. Google leans heavily on third-party and user-generated sites — Reddit, Forbes, and YouTube are among the most-cited domains for these queries. The content that earns a recommendation is independent: reviews, comparisons, and walkthroughs published by someone other than the vendor, a pattern that lines up with why calling yourself the best can end up helping your competitors win instead.
That means the fix isn’t more self-published content — it’s more independently published content about your product, on domains you don’t control: reviews, side-by-side comparisons, and walkthroughs, published as ongoing output rather than one-off placements.
Why an affiliate program is the structural fix
The fastest way to generate that kind of coverage is to give independent creators a financial reason to write it. When a creator earns money each time their coverage converts a customer, they keep producing reviews, updating comparisons, and publishing walkthroughs without you commissioning each individual piece.
A handful of creators on a revenue-share deal will produce coverage, but not consistency — every new placement still needs manual outreach. An affiliate program is what turns that into an always-on channel: recruiting partners, tracking what each one produces, rewarding the ones who perform, and paying them reliably. Affiliates in this context are third parties — niche site owners, YouTube reviewers, newsletter writers, media publishers — who earn a commission for referred customers and, in the process, produce exactly the third-party content AI Overviews draw from when answering “best software” queries.
The brands that currently dominate AI answers tend to have exactly this kind of network behind them: third-party sites reviewing and comparing their products for a commission, continuously adding new coverage. Programs built for editorial output outperform programs built purely for referral volume — chasing raw clicks tends to attract coupon and deal sites that rarely produce the kind of editorial content AI Overviews actually cite.
Telling a strong partner from a weak one takes judgment. Tautvydas Vasiliauskas of affiliate platform FirstPromoter put it this way: “Affiliates are one of the biggest sources of AI citations right now, and yet most brands don’t even think about it. A citation in an AI Overview today doesn’t mean much on its own, because we’re seeing AI-generated sites get cited for a few weeks and then disappear once Google catches up to them. So check the organic history behind it first, look at the fluctuations, scroll through the content. And do that for every partner type, not just websites. A YouTube channel or an influencer can end up in an AI answer too, and they need the same check.”
That vetting matters because a referring domain earned this quarter doesn’t keep earning on its own. Low-quality affiliates and self-referrals dilute the same pool of mentions that AI systems draw recommendations from, so keeping that pool clean — detecting fraud, vetting partners, blocking self-referrals — has to be continuous, not a one-time cleanup. Managing that by hand consumes exactly the hours an affiliate program is supposed to save, which is why most brands running programs at scale use dedicated software to track partner performance against revenue and automate payouts, tiers, and bonuses rather than doing it manually.
The takeaway
The self-ranked listicle had a long run in classic SEO, and AI search is ending it. Google now routes recommendations to the brands the wider web already trusts, built from independent content it doesn’t control. Building that trust means shifting budget from content that cites you toward partnerships that get you recommended — a distinction worth checking before your next content sprint, and one that ties directly into what your content needs to look like to earn AI citations in the first place.
Frequently asked questions
What’s the difference between an AI citation and an AI recommendation?
A citation means an AI answer named your page as a source. A recommendation means the answer told the reader which product to actually choose. Buyers act on recommendations; citations alone don’t move sales.
Why do self-ranked “best software” listicles backfire in AI search?
AI Overviews often cite the listicle as a source, then recommend a different product ranked inside it — usually one with stronger independent coverage across the rest of the web.
How do I check whether AI search recommends my brand?
Build a list of buyer-intent queries, run each one multiple times, and log citations and recommendations as two separate figures. Track your share of recommendations, not citations.
Does an affiliate program actually help with AI search visibility?
It can, because AI systems draw recommendations heavily from independent, third-party content — reviews, comparisons, and walkthroughs. A well-vetted affiliate program is one of the more reliable ways to generate that content consistently.