A homeowner in your city opens ChatGPT and types: "Who's the best HVAC company near me that's available this week?" They get a direct answer — one or two businesses named, with a short explanation of why. They call the first one. They never opened Google. They never scrolled a list.
This is how a growing share of local customers now find businesses. Not by searching and clicking through results, but by asking AI a question and acting on the answer. According to a 2025 BrightLocal survey, 58% of consumers reported using AI tools to search for local businesses — up from 19% the year before. That number has only grown since.
The problem: AI search is dramatically more selective than traditional search. Google's local 3-pack shows businesses 35.9% of the time. ChatGPT recommends a local business just 1.2% of the time it's asked. If you're not in that 1.2%, you don't exist to those customers at all.
The good news: the factors that determine AI visibility are knowable, and for local businesses they cluster around a handful of things you can actually control. Reviews are central to almost all of them.
How each AI platform actually works
The first thing to understand is that ChatGPT, Perplexity, Gemini, and Google AI Overviews all use different underlying data sources — which means visibility in one doesn't guarantee visibility in another.
The practical implication: Gemini is your highest-leverage target because it pulls directly from Google Business Profile and Google Maps — data you already control. A fully optimized GBP with strong reviews is the single highest-ROI action for AI visibility overall. ChatGPT is the hardest to crack and requires broader web presence beyond your GBP.
The five things AI looks at when recommending a local business
AI platforms treat reviews as a confidence filter before they'll recommend you at all. ChatGPT won't recommend businesses averaging below 4.3 stars. Perplexity's floor is 4.1. Gemini will go as low as 3.9 — but below that, none of them will surface you regardless of how optimized everything else is.
Beyond rating, volume and recency matter. A business with 200 reviews averaging 4.7 stars will be recommended over a competitor with 12 reviews at 4.9 — the AI reads volume as a signal of how established and trusted a business is. Reviews from years ago also count less than recent ones: AI tools are trying to give current recommendations, and stale review profiles look like stale businesses.
Gemini is grounded directly in Google Maps data, which is why businesses appearing in Gemini recommendations average 100% profile accuracy — versus 68% on ChatGPT and Perplexity. Every incomplete or inconsistent field on your GBP is a reason for Gemini (and Google AI Overviews) to pass on you.
This means: primary and secondary categories completed, services listed with descriptions, business description written as a clear factual statement (what you do, where, who you serve, what makes you different), hours current, photos added regularly. AI pulls from every structured field when deciding whether your business is a fit for a given query.
ChatGPT pulls from Bing's index and its training data — which means it's looking for your business across the broader web, not just your GBP. If your business is mentioned in local news, industry directories, review platforms like Yelp, or any third-party editorial content, those mentions teach AI models that you're real, established, and worth recommending.
Critically: if your business isn't indexed by Bing, ChatGPT often can't recommend you at all. This is a surprisingly common gap — many businesses optimize for Google and ignore Bing entirely.
AI doesn't just count your reviews — it reads them. When someone asks ChatGPT "who's a good plumber in Austin who handles emergency calls," the AI synthesizes review content to find businesses whose customers have described exactly that. A review that says "called at midnight, arrived within an hour, fixed the pipe, fair price" is more useful to an AI recommendation engine than "great service, highly recommend."
This means the specificity of your reviews matters as much as their rating. Reviews that mention your service type, your location or neighborhood, and specific details about the job are the ones AI can actually use to match you to a query.
This one sounds obvious, but many businesses (and website themes) have robots.txt configurations that block AI crawlers by default — including GPTBot (OpenAI), ClaudeBot (Anthropic), and PerplexityBot. If your site blocks these, those platforms cannot crawl your content and are far less likely to recommend you.
Where to start if you're doing nothing yet
The hierarchy is clear: fix your Google Business Profile first. Gemini is grounded in GBP data, Google AI Overviews trigger on 68% of local queries, and a complete GBP is the foundation everything else builds on. A business with a fully optimized GBP and strong, recent reviews is already in position to appear in the majority of AI-driven local recommendations.
After that, Bing Places for ChatGPT coverage, NAP consistency across directories for credibility signals, and review content quality for the increasingly specific queries AI is being asked.
Find out if AI can
even find you right now
Get a free audit of your AI search visibility — we'll show you exactly where you stand across Google AI, Gemini, and beyond.
Get Your Free Audit →