If a customer asks ChatGPT to recommend a dentist, plumber, or coffee shop nearby, does your business come up? A new industry study tracking more than 120,000 mentions across five major AI models suggests the answer depends on a set of signals that most small businesses have never optimized for.

What happened

A location marketing analytics firm analyzed how AI models — including ChatGPT and Gemini — surface and recommend businesses with multiple locations. The research pulled together 120,000 mentions to identify which factors most consistently correlated with a business being named in an AI-generated answer.

The findings point to four recurring signals: consistency of business information across the web, the volume and recency of customer reviews, the strength of structured data on a business's website (the behind-the-scenes code that tells search engines and AI crawlers what a page is about), and the presence of a business across third-party directories and platforms the AI models pull from.

None of these factors is new to search marketing. What's new is confirmation that AI answer engines appear to weigh them similarly to how Google has long weighed them for local search rankings — but not identically, and not consistently across all five models tested. A business that shows up reliably in Gemini's answers may be underrepresented in ChatGPT's, depending on which data sources each model draws from and how frequently that data gets refreshed.

For multi-location businesses specifically — franchises, regional service chains, multi-branch clinics — the stakes are higher because inconsistency multiplies. A single-location bakery only has one address to get right. A 40-location chain has 40 sets of hours, phone numbers, and review profiles that can drift out of sync.

Why it matters

This fits into a pattern that's been building since early 2024: search marketers scrambling to understand "answer engine" behavior the same way they once studied Google's algorithm. Several SEO platforms have launched AI-visibility tracking tools this year, essentially trying to reverse-engineer how chatbots decide what to cite. That reverse-engineering business is still young, and methodologies vary widely between vendors — a caveat worth keeping in mind, since some of this research comes from companies selling tools built around the same signals they're studying.

The broader trend, though, is real regardless of who's measuring it: a growing share of consumers now ask AI tools for recommendations instead of typing a search query and scrolling results. Local visibility is no longer a single-channel problem confined to Google's local pack.

What this means for small businesses

The practical work looks a lot like traditional local SEO, just extended to more places. Business name, address, and phone number need to match exactly across your website, Google Business Profile, Yelp, Bing Places, and any industry-specific directories — small formatting mismatches ("St." versus "Street") can matter more than they used to.

Review volume and freshness also carry weight. A business with 200 reviews from three years ago may be less visible than a newer competitor with 40 recent ones. Encouraging steady, ongoing reviews — rather than one-time campaigns — is now a visibility tactic, not just a reputation one.

For businesses with a developer or web platform that supports it, adding structured data markup (schema) to location pages is a low-cost, one-time technical fix that multi-location competitors are increasingly prioritizing.

What to watch

Watch whether Google, OpenAI, or Microsoft introduce any official business data feed or API for AI assistants — a formalized system would replace today's guesswork with clearer rules, the way Google's local pack eventually did for maps-based search. Also watch whether mainstream SEO tool providers (Semrush, Ahrefs, BrightLocal) roll out standardized AI-visibility scores, which would make this easier to track without hiring a specialist.

The bottom line

Business listing consistency, active review generation, and basic structured data are now doing double duty — supporting both traditional search rankings and AI-generated recommendations. Multi-location businesses have the most cleanup work to do, but the underlying fixes are inexpensive and mostly a matter of auditing existing listings rather than adopting new technology.