“Get ChatGPT To Name Us” — Okay, Then What?
Lately every marketing conversation lands in the same spot: “how do we get ChatGPT to mention us?” Search was the last battle, AI is the next one, and whoever the model recommends wins. I get it. The instinct makes total sense.
I had the same reaction the first time I got an AI to name the thing I work on. It popped up on screen and I nearly did a little fist-pump. Getting named really is a big step.
But when all your attention goes to getting named, the next question quietly drops out. How do you actually become the brand the AI is likely to cite and recommend? That’s where the real leverage is — and it turns out the gap between brands shows up sharply in the numbers.
The hint is in what shoppers do. Idea Grove, a US PR firm, asked 1,000 American consumers about this, and only 2% said they’d buy a brand they didn’t already know on an AI’s say-so alone (source). Idea Grove sells “trust-building,” so read that figure as a bit of a position — but the other 98% going off to check first, through search, the official site, or the AI’s , isn’t a stretch when you picture your own buying.
So even when the AI names you, people go looking for third-party backing. Which means if you build the state of being cited — being backed by someone other than yourself — before they go looking, the recommendation actually turns into something. The obvious next question is what backing works best. There’s a big dataset on exactly that, and that’s what I want to walk through.
When Third-Party Reviews Show Up, The AI Starts Citing You
The thing that does the work is third-party reviews, and there’s a sizable analysis on it: Seer Interactive ran it, commissioned by Trustpilot (source). They looked at over 800,000 AI answers pulled from ChatGPT, Gemini, Perplexity, and Google’s AI Mode.
The method is clean. They sorted brands by how they handle reviews into four groups, then compared how often the AI cited each one. And the gap between groups is almost startlingly tidy.
- A brand with no review profile gets cited about 1% of the time
- Just having a profile takes it to 53.5%
- A brand that actively collects reviews and replies to them hits 75.3%
For what it’s worth, review and trust sites made up 14% of all AI citation sources — the second-biggest source type. The model leans on what third parties say about a brand a lot harder than you’d guess.
One caveat: this study was commissioned by Trustpilot, a review platform, so it tilts toward playing up the value of collecting reviews — discount it a notch. But the scale is concrete, and “third-party voices move the AI” lines up neatly with the way shoppers go off to verify.
Why It Jumps This Hard: Profile, Collect, Reply
The fun part is that 1% → 53.5% → 75.3% isn’t just three numbers. It’s a three-rung ladder of things you can actually do.
The first rung is having a review profile. Just existing on one takes you from 1% to 53.5%. From the AI’s side, a brand with no third-party voices to point at is a brand it can’t really talk about — there’s nothing to cite. Get the profile and you finally stand in the doorway.
Rungs two and three are collecting reviews and replying to the ones that come in. Do that and you’re at 75.3%. New voices keep landing, and there’s a visible trail of the business actually responding. So the AI picks up evidence of freshness and good faith — “this brand is alive and pays attention to its users” — and folds it into what it reads.
Think about hunting for an apartment. A place with zero reviews feels riskier than one where the landlord replies to recent ones, even if you can’t say why. What the AI’s doing is maybe not so different.
Conclusion: You Build AI Visibility In Three Rungs
So here’s the whole thing in one breath: whether the AI cites and recommends you isn’t down to luck or name recognition — you can move it a lot by how you handle reviews. No profile, 1%. A profile, 53.5%. Collect and reply, 75.3%. Visibility is, literally, something you build.
Which makes the to-do list short, and it’s three rungs. First, have a third-party review profile. Then actively collect reviews. Then reply to the ones you get. You don’t need all three at once — just standing up a profile moves you from 1% to 53.5%, so the very first rung pays.
The flip side: pouring your whole budget into only getting named is like a gorgeous sign over an empty store. People go check what’s behind the recommendation, and whether the third-party voices are there is the thing that decides it. So design getting named and building the backing as one piece — not two separate projects.
Sources
- Idea Grove, “Only 2% of Consumers Buy From an AI-Recommended Brand Without Checking It First, Idea Grove Survey Finds”, BusinessWire, 2026-04-16, link
- Seer Interactive (commissioned by Trustpilot), “Brands that build trust through reviews increase AI citations from 1% to 75%”, PR Newswire, 2026-03, link
