Your Customers Are Asking ChatGPT Who to Call. Are You in the Answer?

OHM Editorial Team

A homeowner in your service area had a water heater die last week. Five years ago she would have typed “water heater replacement near me” into Google and scrolled. Last week she opened ChatGPT and typed “my water heater is leaking, who should I call in my area and what should I expect to pay attention to?” She got a calm explanation, a short checklist, and a handful of company names.

Maybe your company was in that answer. Probably you have never checked. That is the uncomfortable part: an entire referral channel has opened up, it is already steering real customers, and most owners have no idea what it says about them.

You do not need to become an AI expert. You need to understand one shift: the recommendation that used to come from a neighbor or a search results page now increasingly comes from a machine that has read everything public about your company and formed an opinion.

Key takeaways

• Homeowners increasingly ask AI assistants like ChatGPT who to call for urgent home repairs, and the answer names a few companies instead of returning ten links.
• AI assistants build recommendations from public evidence: review volume and recency, the plain words on a company’s website, consistent name, address and phone listings, and structured data.
• Google’s own guidance on creating helpful, people-first content reads almost like a description of what AI assistants reward, so the fundamentals of local SEO still apply.
• The shift favors legitimate local businesses with years of real reviews and clear service areas over lead brokers and thin national directories with many pages and no local substance.
• This month, ask ChatGPT and Google’s AI features the questions customers ask, read the website like a machine would, ask for a review after every finished job, and fix stale directory listings.
• Owners do not need to buy an AI visibility tool or write pages for robots; content that reads like a knowledgeable human explaining the work is what the models prize.

The Question Has Moved

People have not stopped searching. They have started asking, and the difference matters. A search hands back ten links and makes the homeowner do the work. An AI assistant hands back an answer: what the problem probably is, whether it is urgent, and who nearby looks trustworthy enough to call.

That answer compresses the entire shopping process into one screen. There is no page two. There is no “maybe I’ll click the fourth result.” Either your company is part of the answer or the job quietly goes to someone whose company is.

For home services this cuts deeper than for most industries, because the questions are urgent and personal. Nobody comparison-shops for fun at 11 PM with water on the floor. They ask, they get a name or three, and they call.

How AI Decides Who to Recommend

AI assistants do not keep a secret list of favorite plumbers. They read what exists in public and reason from it. When a model recommends a company, that recommendation is assembled from a few familiar ingredients.

• Your reviews, in depth. Not just the star average. The volume, the recency, the specifics customers mention, and how you respond to the bad ones.
• Your website’s actual words. Models read pages literally. A site that plainly says what you do, where you work and how to reach you is usable raw material. A wall of vague slogans is not.
• Consistency across the web. Your name, address, phone and services showing up the same way on directories, maps and profiles. Contradictions read as risk.
• Structured data. The machine-readable layer on your site that spells out your business type, service area and offerings in a format models and search engines both trust.

Notice what is on that list: the same fundamentals that have always driven local SEO. The machines did not invent new rules. They raised the price of ignoring the old ones. Google’s own guidance on creating helpful, people-first content reads almost like a description of what AI assistants reward.

Why This Rewards Small, Real Companies

Here is the encouraging part nobody tells owners: this shift favors legitimate local businesses over marketing-heavy national lead machines. An AI model looking for a trustworthy local recommendation wants evidence of a real company doing real work in a real place. Twenty years of local reviews, photos of actual jobs and a clear service area are exactly the evidence it can use.

The lead brokers and thin national directories that gamed old search results have the opposite profile: lots of pages, no local substance. Models are noticeably less impressed by that than the 2015 version of Google was. The playing field is not level, but for once it tilts toward the company with the trucks instead of the company with the ad budget.

What to Do This Month

None of this requires a moonshot. It requires doing a handful of ordinary things deliberately.

• Ask ChatGPT and Google’s AI features the questions your customers ask: “who should I call for a burst pipe in [your city],” “best HVAC company near [your area].” Note who comes up and what gets said about you.
• Read your own website like a machine would. Can a stranger tell in ten seconds what you do, where, and how to reach you? If your pages talk around the work instead of about it, that is the first fix.
• Get your review flow moving. A steady ask after every finished job beats a burst of begging twice a year, and recency matters to machines and humans alike.
• Fix the contradictions. Old phone numbers and stale addresses on forgotten directories quietly undermine you with every system that cross-checks.

If you want the deeper treatment, this is exactly what our AI search optimization work does: making your company legible, verifiable and recommendable to the systems your customers now ask. And keep an eye on this space as it develops; we will keep writing about it, including how Google’s AI Overviews are changing clicks for contractor websites.

What This Does Not Mean

A shift like this attracts snake oil, so two warnings are worth the ink. First, you do not need to buy an “AI visibility tool” from the stranger who just emailed you. Most of what moves the needle is the fundamentals above, done consistently, and anyone selling a shortcut around fundamentals is selling the shortcut, not the destination.

Second, do not start writing your pages for robots. The models are specifically built to prize content that reads like a knowledgeable human explaining things to another human. The company that describes a water heater install the way its best tech would explain it at a kitchen table is doing AI optimization, whether it knows it or not. The company stuffing “best plumber AI recommended” into headers is doing the opposite.

Final Thoughts

Every owner reading this has lived through a version of this shift before. The Yellow Pages gave way to Google, and the companies that moved early spent a decade collecting the customers of the ones that waited. The move to AI answers is the same story with a shorter runway.

You do not have to be first. You have to be present, accurate and worth recommending, in the places the machines read. Check what AI says about your company this week. Whatever you find, at least you will know which side of the answer you are on.

Frequently asked questions