Brands with active review profiles get cited by AI 75% of the time. Brands with none get 1%.

A brand with no claimed review profile gets cited by ChatGPT, Gemini, Perplexity, or Google AI Mode in about 1% of relevant answers. Claim that profile and add a handful of reviews, and the citation rate jumps to 53.5%. Build it out further, to a median of 81 reviews, and it climbs again, past 75% in several categories. Those figures come from a joint study by Trustpilot and Seer Interactive that analyzed 804,491 AI responses across 1,926 brands in eight industries during March 2026. A fifty-point jump from a single claimed profile is the kind of number that should reorder a brand team's GEO priorities this quarter, not sit in a slide deck nobody revisits.

What the Trustpilot and Seer Interactive study actually measured

Trustpilot and Seer Interactive split 1,926 brands into four profile tiers, from no verified profile to a fully optimized one, then measured how often each tier showed up across 804,491 AI responses to 15,783 prompts on ChatGPT, Gemini, Perplexity, and Google AI Mode in March 2026.

The tiers themselves matter more than the labels. A minimal profile could have as few as one to thirteen reviews, while an active one carried a median of 81. To isolate the profile effect from plain brand strength, researchers matched 206 brands per tier on Ahrefs' Domain Rating, so a company with a weak website couldn't simply borrow a strong one's citation rate. Eight verticals were covered, including Money & Insurance, Travel & Vacation, and Electronics & Technology, chosen because how much people trust strangers' reviews varies sharply between booking a flight and buying a laptop.

Travel & Vacation produced the largest single jump: a 43x increase in citation rate between brands with no profile and brands with a minimal one. That fits a category where most travelers have no personal history with the company an AI engine is about to recommend, and lean entirely on what strangers said. Review and trust sites rank as the second most-cited source category overall behind the open web, which lines up with this blog's earlier look at how little AI citation actually comes from a brand's own website.

Why a review profile moves an AI answer at all

AI engines rarely query Trustpilot on purpose. The study found 71.7% of citations arrived through ordinary background searches that never mentioned reviews at all, against just 0.5% from a prompt that explicitly asked about Trustpilot. The platform shows up because it already ranks well in the searches a model runs behind the scenes, not because anyone asked for it by name.

ChatGPT referenced a TrustScore or star rating in 59% of the more than 90,000 responses that cited Trustpilot, most often when the conversation touched on customer service, shipping, or whether a company was legitimate at all. Legitimacy questions are exactly where a model has the least training data to fall back on, so it reaches for whichever third party looks most structured and already carries a verifiable rating.

A system with no first-party way to verify a claim defaults to whichever third party looks most corroborated. That instinct isn't new. Search engines have rewarded this kind of third-party trust signal for years; AI chat answers just inherited it.

The dose-response curve, from no profile to an optimized one

The lift isn't a single switch that flips once a brand claims a profile. Citation rate climbed at every tier Trustpilot and Seer Interactive measured: roughly 1% with no profile, 53.5% with a minimal one, about 25 points higher again with an active profile, and a further 6 points once a domain-rating-matched cohort moved from active to fully optimized.

A brand with no active review profile gets cited by AI chatbots about 1% of the time. One with a mature, optimized profile gets cited closer to 75% of the time.

The same tiers also showed up in co-mentions, beyond standalone citations. Brands with optimized profiles were named alongside competitors roughly 9.5 times more often than brands with no profile at all, and the gap widened to 20x in Electronics & Technology specifically, a category where shoppers routinely ask an AI engine to compare several brands in one answer instead of asking about just one.

Domain authority doesn't erase the need for reviews

A strong website doesn't buy its way out of this. Among brands with weak Domain Rating, 0 to 15, a minimal review profile reached a 54.5% citation rate, almost identical to the 52.6% posted by brands with Domain Rating 16 to 35 at the same profile tier. Review profile strength and site authority appear to work as separate levers, not substitutes for each other.

Brands with Domain Rating 16 to 55 reached 75 to 81% citation rates at the higher profile tiers, among the strongest results in the whole study. That pattern lines up with what this blog's own research on LLM Brand Visibility keeps finding: domain authority was built to predict Google rankings, not whether a model decides to name a brand out loud, and review profiles are turning out to be one of the few off-site signals that moves the second number on its own.

The real risk: this is easy to fake, and the FTC is watching

The obvious temptation is to inflate a review profile instead of earning one, and that's now a specific, enforced violation. The FTC's Consumer Review Rule took effect on October 21, 2024, and bans buying reviews, writing fake ones, or presenting a company-run review site as independent.

Enforcement has been ramping up, not fading. The FTC sent warning letters to ten companies during the 2025 holiday shopping season, and the rule allows civil penalties of up to $53,088 per violation, a figure that adds up fast across a profile padded with dozens of fake entries. A platform caught hosting manipulated reviews is also a platform an AI model will eventually learn to discount, which erases the exact citation lift this study measured.

Trustpilot and Seer Interactive are careful to call their own findings correlational, not causal. Marketing spend and audience size could easily explain part of why bigger, more visible brands also have more complete review profiles, and the firms note the Travel & Vacation spike may partly reflect how many unfamiliar booking sites AI models already treat with suspicion.

What to actually do with this before your next GEO review

Start with the platform your buyers already check: Trustpilot for consumer brands, G2 or Capterra for B2B software, and Google or Yelp for anything with a physical location. Claim the profile if nobody has, respond to every review instead of letting a profile sit static, and prioritize real product and location reviews over chasing a star-rating total alone.

Trustpilot's own data shows adding product reviews lifts citation rates by about 12 percentage points, and collecting location reviews adds up to 7 more, with the effect building further once a location passes five to nine reviews. From there, track citation rate by engine rather than as one blended number, the same way this blog's guide to tracking brand mentions across ChatGPT, Gemini, and Perplexity recommends tracking AI Share of Voice, since the four platforms in this study didn't move in lockstep.

If an AI answer is already citing a review profile and getting the facts wrong, that's a correction problem rather than a reviews problem, and the fix looks more like what this blog has already laid out for fixing wrong AI answers about a brand than like collecting five more stars.