Plumbers, electricians, dentists, gyms, garages, vets. Local service businesses have always won or lost on trust and proximity, and for the last two decades that trust signal lived on a Google results page or a maps listing. It is now, increasingly, living inside a single answer from an AI assistant, with no results page attached at all.
Ask ChatGPT, Gemini or Perplexity "who's a reliable plumber near me" or "recommend a good dentist in Leeds" and you get two or three names, sometimes one, and a short reason why. There is no scroll, no page two, and often no visit to a website before the customer picks up the phone.
Why this is a bigger shift for local service than it looks
Local service businesses already relied heavily on proximity and reviews to win the "who should I call" moment. AI recommendation engines compress that decision even further. Instead of comparing five options on a maps listing, the customer gets one confident answer. If your business is that answer, the effect is strong. If it isn't, you are often not in the conversation at all, not just ranked lower in it.
This matters more for local service than for most other categories precisely because the buying decision has always been high trust and low research. AI assistants are stepping directly into the role a personal recommendation used to play, and most local businesses have no idea what that recommendation currently says about them.
What determines whether you get named
The same underlying signals that decide golf club visibility apply here, with a local twist.
- Review volume and sentiment. AI models weigh recent, specific reviews heavily, particularly ones that mention concrete details like response time, price, or the specific job done.
- Local citation consistency. Your business name, address and phone number need to match across your site, directories and review platforms. Inconsistent listings actively confuse the sources AI models draw from.
- Plain, specific website content. A page that says "serving Leeds and the surrounding area since 2011, specialising in emergency boiler repair" gives a model something concrete to quote. A page that only says "quality service you can trust" gives it nothing.
- Structured data. LocalBusiness schema markup is a direct, machine readable signal of what you do, where, and for whom, and it costs nothing to implement properly.
- FAQ style content answering real buyer questions. "Do you offer emergency callouts," "what areas do you cover," "how much does a boiler service cost" answered plainly on your own site is often lifted almost verbatim into an AI answer.
The gap most local service businesses have right now
In our testing across golf and local service categories, the single most common gap is not reputation, it is readability. Businesses with genuinely strong reviews and a solid local reputation are still frequently absent from AI recommendations, because the information that would let a model confidently name them either doesn't exist on their site in plain text, or is inconsistent across the sources the model is drawing from.
That is a fixable, mechanical problem, not a brand problem. It just requires someone to actually test the real prompts a customer would use, see what comes back, and close the specific gaps that show up.
What to do this week
Open ChatGPT, Gemini and Perplexity and ask the exact question a customer in your area would ask about your category. Write down what comes back, who gets named, and what's said about them. That fifteen minute exercise will tell you more about your real world discoverability than most marketing reports you've paid for.