When a customer types "best dentist near me" into Google, Google checks their GPS coordinates, returns results within a few miles, and pins them on a map. The whole exchange takes half a second and geography does most of the work.
When that same customer asks ChatGPT, Claude, or Perplexity the exact same question, something quietly different happens: the AI has no idea where "near" is. It has no GPS signal, no IP-to-location lookup by default, and no map database it's querying live. The word "near" is essentially invisible to it. This single fact changes how local businesses need to think about AI visibility — and most owners don't know it yet.
What AI Actually Does With a Location Query
When an AI assistant gets a "near me" or "in my area" question, it falls back on one of three things:
Context from the conversation. If a user said "I'm in Tucson" three messages earlier, a good AI will remember that and filter results accordingly. Most conversations don't include that setup.
Account or device location passed by the app. ChatGPT on mobile can optionally share your city-level location with the model. Perplexity has a similar setting. These are opt-in and off by default for many users, and even when on, they pass the city name — not a street address — so the AI still can't distinguish one neighborhood from another.
Training data and general reputation. When neither of the above applies, the AI leans on what it absorbed during training: reviews, articles, directory listings, and forum posts that mentioned a business by name alongside a city. This is where your structured content — or lack of it — determines who gets named.
The result: "best HVAC company near me" often produces a city-agnostic answer, a request for clarification ("what city are you in?"), or a recommendation based on whoever had the strongest written footprint in the training data. That last outcome is the one that costs or rewards local businesses.
Why "Near Me" Is a Dead Phrase for AI Optimization
Traditional local SEO taught businesses to chase "plumber near me" as a keyword. For Google, that phrase is meaningful because Google resolves the location server-side. For AI, the phrase is an empty placeholder.
The phrases that actually work in AI are city plus service: "emergency plumber in Columbus," "best family dentist in Boise," "affordable bakery wedding cakes in Tampa." These queries give the model enough geography to match against its training data without needing a live location signal.
A roofing company in Albuquerque that has built a strong written presence — consistent citations, review responses mentioning the city name, a website that says "serving Albuquerque and Rio Rancho" in paragraph text rather than just in a footer — will be named by AI far more reliably than a competitor with better Google Maps rankings but thin written content.
The Service Area Problem
"Near me" fails AI. But so does vague service area language. Saying you're "proud to serve the greater metro area" teaches the AI nothing. AI models learn from specifics. If your website, Google Business Profile description, and directory listings say "serving Denver, Aurora, Lakewood, Littleton, and Centennial," the model can associate your business name with all five cities.
This is especially important for businesses that cover multiple towns: landscapers, HVAC contractors, home health agencies, mobile dog groomers. A customer in a satellite suburb who asks AI for a recommendation will only get your name if something in the AI's training tied your business to that suburb by name. One well-written service area page — listing each town with a sentence or two of actual context — does more for AI visibility than a map widget ever will. The LocalBusiness schema guide covers how to mark up those area relationships in a format AI crawlers can parse.
What Perplexity Does Differently
Perplexity is worth calling out separately because it actively searches the web at query time rather than relying purely on training data. When someone asks Perplexity for "best pediatric dentist in Springfield," it pulls live results, reads them, and synthesizes a recommendation — then cites its sources.
This means your Yelp page, your Google Business Profile (via scraped web content, not the Maps API), and your own website's text are all in play. Businesses that show up on the first page of organic results for city-plus-service queries tend to get cited by Perplexity. The path to Perplexity citations runs through ordinary on-page SEO and consistent structured data — not any Perplexity-specific trick.
A free scan will flag whether your business is surfacing in AI answers for your key city-service combinations, or whether the AI is drawing a blank when your city enters the conversation.
Three Practical Changes to Make This Week
Replace "near me" with real city names everywhere. Your homepage, About page, and Google Business Profile description should name the specific cities you serve. Not "the surrounding area" — the actual names.
Add a plain-text service area paragraph to your website. One paragraph, readable by a crawler, listing every city or neighborhood you serve. Not a map, not a dropdown, not a JavaScript-rendered component — raw text the AI can read. The plain-text rule applies here: if the AI can't read it as text, it doesn't exist.
Use city names in your review responses. When you reply to a Google review, naturally include the city: "Thanks for trusting us with your heating system in Naperville." These responses are crawled and indexed, and they give AI models another data point tying your business name to a specific geography.
The customers asking AI for local recommendations are already out there. The gap between your business being named and being invisible often comes down to whether the AI had enough city-specific text to make the connection. Fix the text, and the geography takes care of itself.