The water heater
I needed to replace a gas water heater, for a household of three. Nothing dramatic, no flood, no emergency call at midnight. Just a heater that had reached the end of its life, and a decision I had been putting off.
So I did what most people do now. I asked Muse to find a heater company nearby, and what it would cost.
Muse came back fast, with Toronto-area options and broad cost ranges. And then it asked the questions that actually mattered — repair or replacement, gas or electric, how many people in the household — because a price means nothing without those answers. I ended up asking for quotes from three companies, Reliance, Spring Home, and Megacity. And even then, the forms stalled on the small human details: a CAPTCHA to handle, a yes-or-no question about whether I was the homeowner.
That is the whole story, in miniature. The AI can narrow the field in seconds. It can surface the options and the ranges. What it cannot do is decide for the customer. And what it will not do is recommend a company it cannot verify.
I was the customer that day. I was also, without meaning to be, running the experiment every local business is now living through.
The decision happens before the door
A business that depends on people walking through the door does not make the sale at the counter. The sale is made earlier, on a phone, in the minutes between a search and a visit.
Nearly half of all Google searches carry local intent, and most of those searches start on a mobile device. A large majority of people who search for something nearby visit a related business within a day, and a substantial share make a purchase in that same window. Most consumers settle on a local business in under half an hour. A notable minority do so in under five minutes.
That compression is the story. Not how much attention a business can buy, but whether it is legible in the brief moment when a stranger is choosing between two listings. My water heater search took longer than five minutes, because a water heater is a big decision. A tow truck, a dentist, a winery tasting, a furnace repair — those are often decided faster, with less patience, and with no second look.
The same story, in twenty-seven different businesses
The list of businesses that live this way is long, and they look nothing alike. Golf courses. Wineries. Tour operators and attractions. Inns and bed-and-breakfasts. Big spa chains. Home renovation contractors. HVAC companies. Auto shops and car dealerships. Transportation companies, carriers, brokers, and tow trucks. Dental clinics. Moving and storage companies. Apartment rental and leasing. Pest control. Restaurants. Landscaping and lawn care. Roofing, tree service, and snow removal. Plumbing and electrical. Car wash and detailing. Cleaning services. Pet grooming and boarding. Childcare and daycare centres. Gyms and fitness studios. Campgrounds, RV parks, and marinas. Event and wedding venues.
A winery tasting room and an HVAC van share almost no operational detail. A golf course and a dental clinic share even less. What they share is a structure: a physical place, an owner or a manager who spends the day on their feet, and a customer who finds the place on a phone and decides quickly whether to go there.
Being found is being open
The first filter is presence. A listing that is claimed, complete, and current — name, address, hours, services, photographs — reads as a business that is open and operating. A listing that is thin, out of date, or unclaimed reads as the opposite, even when the door is unlocked.
A complete profile is associated with a markedly higher likelihood of being seen as reputable, and a markedly higher likelihood of a visit. Yet while most small businesses have a profile, only a minority actively maintain one. The gap between having a listing and keeping it alive is one of the largest unforced losses in local marketing, and it is invisible to the owner, because the customers who pass over a stale listing never call to say so.
The map compounds it. The small cluster of businesses shown at the top of a local search captures a disproportionate share of attention, clicks, calls, and direction requests. Businesses outside that cluster are not necessarily worse. They are simply not seen in the moment the decision is made. Like the heater companies that never made it onto my shortlist — I never saw them, so I never judged them.
Being chosen is being trusted
Once found, the choice turns on trust, and trust in local search is largely a function of reviews.
Almost every consumer reads reviews for local businesses, and the share who always read them has risen sharply in a single year. The bar has risen with it. A clear majority now require a minimum four-star rating before considering a business, and a growing share require four and a half. Recency matters as well: a review from months ago carries less weight than one from last week, and a business with no recent reviews reads as dormant.
Responses matter too. Consumers are more likely to use a business that replies to its reviews, positive and negative alike. Silence in the review section is read as indifference, and indifference is a reason to choose the listing next door.
None of this is reputation in the abstract. It is the specific, observable signals a stranger uses to decide, in under a minute, whether a place is worth the drive. When I was looking at heater companies, I was doing exactly that — reading what other people had said, and noticing who had bothered to reply.
The return visit
Foot traffic is not only first visits. The businesses that sustain it are the ones where a first visit becomes a second one: a regular at the car wash, a repeat booking at the spa, a seasonal return to the golf course, a winery visit that becomes a club membership, a dental patient who comes back, a tenant who renews.
The same signals that drive the first visit drive the return. A profile that stays current, reviews that stay recent, photographs that reflect the place as it actually is. The second visit is easier to make when the business is easy to find again, easy to remember, and easy to trust a second time.
In the AI era, the answer comes before the click
A new layer now sits between the search and the visit. AI assistants and AI-generated summaries answer local questions directly, often without a click through to a website at all. A person asks for a winery near Niagara, a tow truck nearby, a dentist open on Saturday, and the answer arrives as a short list, already chosen.
That changes the story without replacing it. The signals that feed the answer are the same ones that fed the old listing: a complete profile, consistent business information across the web, recent reviews, photographs, hours that are actually correct. The difference is that the business no longer gets a chance to make its case on its own website. The case is made, or not made, by the data the AI can find.
The shift is fast. In a single year, the share of consumers using AI tools for local business recommendations rose from a small minority to nearly half, and AI summaries now appear on a large share of local searches.
A heater company makes the point concrete. A furnace stops working on a cold evening, and the question put to an AI assistant is not a brand name. It is a problem, a place, and a time: a furnace that has stopped working, nearby, someone who can come today. The answer the AI gives is drawn from what it can verify — a service area that is stated plainly, emergency hours that are actually current, recent reviews that mention furnace repair in that town, and a page that answers the exact problem in plain language. A heater company becomes AI-friendly not by writing about AI, but by being legible to it: the same facts, stated clearly, consistently, and recently, everywhere the AI looks.
That was my experience, from the customer’s side. The AI narrowed the field. It did not decide for me. And the companies it surfaced were the ones it could verify.
