ChatGPT's local answers skip Google reviews, study finds

If you have spent years collecting Google reviews, here is a finding worth knowing about: when ChatGPT answers a local question, it may not be looking at them at all.

Near Media's podcast (episode 275) covers a summer study by Local SEO Guide, run with Local Falcon, across several million local queries. The finding reported is that ChatGPT's local sources were Bing, Foursquare, Yelp, TripAdvisor, BBB and editorial "best of" lists — with no Google review data among them.

This is a trade publication reporting a third-party study and a panel discussion around it. It is not a Google announcement, and nothing about how your Google Business Profile works has changed.

What the study reported

Two numbers came out of it.

First, the source list above: Google reviews were not among the things ChatGPT was pulling from for local answers.

Second, across eight verticals, more reviews correlated with more AI visibility on both Google and ChatGPT — but only up to roughly 200 reviews. Past that, the panel says visibility plateaus, though the geographic radius a business can rank in keeps widening.

One of the speakers, Mark Kabana, frames the wider problem like this: a model first filters for candidate businesses using the facts it actually has — parking, pricing, accessibility, opening details. If a fact is missing, you are excluded before any ranking happens. That is his argument rather than a measured finding, but it matches what we already know about how Google uses profile attributes.

What this actually means for a small business

Not that Google reviews stopped mattering. Google reviews still drive what happens on Google, which is where the overwhelming majority of local searches still take place. If you are still building yours up, getting your first Google reviews is the higher-value job by a distance.

What it does suggest is that if you have never touched your Yelp, Foursquare, TripAdvisor or BBB listing, those pages may be doing quiet work you cannot see. Worth a look if the name, address, phone number or opening hours on any of them are wrong.

The more useful takeaway is the attributes one. Every blank field on your profile — and on your website — is a question a machine cannot answer about you. Whether you have parking, whether you are step-free, what you charge, which services you actually offer. Filling those in costs nothing and helps a real customer as much as it helps a model. Our guide on what to put on your Google Business Profile covers the fields most owners leave empty.

One small check from the episode is worth doing: ask an AI tool what your business charges. The panel's point is that a competitor's page or an old forum thread may be quoting a price you abandoned years ago, and you will never find out unless you look. If the number is wrong, the fix is publishing the correct one somewhere clear on your own website.

The lead-source question

The panel's number-one recommendation was unglamorous: ask every enquiry how they found you, on your forms and on the phone.

That is the only way any small business will ever know whether AI answers are sending it work. Visibility reports from tools cannot be tied to an actual booking. A hundred people telling you "I saw you on Google" or "I asked ChatGPT" can be.

The episode also discusses falling calls despite rising rankings at a large franchise client, which the panel attributes mainly to paid ads taking more space on results pages rather than to AI. That is their reading of one account's data, not a measured conclusion, and there is nothing a single-location business can do about it either way.

No action is urgent here. Fill in your blanks, check your prices, and start asking people how they found you.