AI visibility for home services

Local discovery channels normally shift over decades. Yellow pages gave way to Google over fifteen years, and review platforms took a decade to become table stakes. AI assistants compressed that curve into twelve months. BrightLocal's 2026 Local Consumer Review Survey of 1,002 US consumers found 45 percent had used AI tools to find local business recommendations in the past year, up from 6 percent the year before. In the same survey Google's usage as a discovery channel fell from 83 percent to 71 percent.

The demographics should worry every trade business owner, because adoption is heaviest exactly where the customers are. BrightLocal found the 30 to 44 age group leading at 64 percent, which is the prime span of first-time homeowners discovering what a sump pump is and established owners financing kitchen remodels. These are the people asking ChatGPT who should replace their water heater, and they are asking today.

For a plumber, HVAC contractor, electrician, roofer or cleaning company, the practical meaning is simple: a growing share of your future customers get a shortlist from an assistant before they ever type your trade into Google. That shortlist typically holds two to four names. Being on it is the new version of ranking in the map pack, except almost nobody in the trades is competing for it yet.

Home services demand starts with a problem, and problems produce questions in plain language. My AC stopped blowing cold, who do I call. Best plumber near me who does weekend emergencies. How much should a 200 amp panel upgrade cost. Homeowners used to route those questions through Google and a wall of ads. Increasingly they say them to an assistant, and the assistant answers with a short list of specific companies and a summary of why.

The urgency profile of the trades makes these answers unusually decisive. A burst pipe at 9pm produces exactly one phone call. Nobody with water spreading across the kitchen floor comparison-shops five contractors. If the assistant's answer names three companies and yours is one of them, you get a real share of those calls. If it never names you, the outcome is invisible: your phone simply rings less, and there is no auction dashboard showing you the impressions you lost.

This also changes the economics of visibility. Lead platforms sell the same emergency lead to several contractors and charge per contact. An AI mention costs nothing per call. The work is earning the mention, which is a content and reputation problem rather than a bidding problem, and it favors operators willing to publish real answers over operators with the biggest ad budget.

5W's HVAC and Plumbing AI Visibility Index ran more than 65 consumer-intent prompts through ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews in early 2026 and found that about 87 percent of independent contractors have effectively zero citation share in their own metros. The winners are national franchise networks: Roto-Rooter, ARS Rescue Rooter and Mr. Rooter together pull roughly 19 percent of all consumer-intent citations in the trades.

The franchises are winning on structure rather than quality. They have city pages for every metro, consistent profiles across every review platform, pricing content, and decades of press coverage, which means engines searching for citable sources about plumbers in any given city keep finding the same national names. The independent contractor with better reviews and faster response times often has a single-page website and a Facebook account, which gives an engine almost nothing to retrieve.

The honest read of the 87 percent number is that it describes a vacuum, and vacuums fill fast for whoever moves first. In most metros, a handful of independents with real service pages, current review profiles and published pricing guidance could plausibly become the locally cited names, because the assistant genuinely prefers a specific local answer when one exists. The window is open precisely because almost nobody in the trades knows this game is being played.

When ChatGPT with web search answers a who-should-I-call question, it retrieves and reads sources: review platforms, local directories, community threads, news mentions and the businesses' own sites. It then composes an answer, usually naming a few companies with a sentence of justification each, such as strong reviews, emergency availability or transparent pricing. The companies named are the ones the retrieved sources describe most consistently and specifically.

Reviews carry heavy weight because they are the richest third-party evidence about a local trade. BrightLocal found 97 percent of consumers still lean on reviews, and the engines behave the same way, quoting review counts, ratings and even recurring phrases from review text. A contractor with 400 recent Google reviews mentioning fast emergency response gives an engine a ready-made justification sentence. A contractor with 12 reviews from 2021 gives it nothing safe to say.

Your own site matters in a different way than it does for rankings. The engine is looking for confirmable specifics: which cities you serve, which services you actually perform, whether you handle emergencies, what things roughly cost. Pages that state those facts plainly get retrieved and quoted. Vague brochure copy about quality craftsmanship and family values gets skipped, because there is nothing in it an engine can use to answer a concrete question.

Cost pages are the highest-leverage content in home services, because cost questions dominate what homeowners ask assistants before hiring anyone. What a water heater replacement costs in your city, what a roof replacement runs per square, what an EV charger install costs with panel work. Publish real ranges from your own jobs, dated, with the factors that move the price. Engines cite specific current numbers, and the contractor whose numbers get cited is the contractor whose name rides along in the answer.

Service-plus-city pages come second. One page per meaningful service in each city you serve, stating plainly that you do that work there, your response time, and what the job involves. This is old local SEO advice that now pays twice, because assistants answering who fixes heat pumps in a specific suburb need a source that literally says a named company fixes heat pumps in that suburb.

Diagnostic content rounds out the set. Why is my furnace short cycling, is a leaking water heater an emergency, can I DIY a panel upgrade. Homeowners ask assistants these questions in the hours before deciding to hire, and answering honestly, including when a DIY fix is fine, builds the kind of source engines keep returning to. The contractor cited for the diagnosis is naturally positioned in the follow-up question about who to call.

Most contractors currently assess AI visibility by asking ChatGPT about themselves once, seeing a decent answer, and moving on. That test is doubly wrong. Branded questions flatter you, because the homeowner who already knows your name was never the problem. And single runs mislead, because assistant answers vary between identical runs, so one check tells you almost nothing about what a hundred homeowners actually heard this month.

The measurement that matters is a mention rate on unbranded questions: across repeated runs of best plumber in your city, emergency HVAC repair near me phrased for your metro, and your cost questions, what fraction of answers literally name your company, and who gets named instead. Tracked over weeks, that number tells you whether your visibility work is moving anything, in the same way call volume tells you whether a radio spot worked.

Reachroller runs exactly this loop. It asks your question set through the official ChatGPT API with web search, live today, with Claude, Gemini, Perplexity and Grok built and rolling out, and counts a mention only when your business name literally appears in the stored answer. Every score is auditable back to a real response, and each losing question shows the sources the engine cited, which becomes your to-do list. The 3-day trial with 50 credits and no card is enough to baseline a service business completely.

Every local discovery shift has minted early winners: the contractors who claimed Google Business Profiles in 2010 and built review bases before rivals cared owned the map pack for years. The AI shift is at that same early stage, with 45 percent of consumers already asking and 87 percent of contractors invisible. The math of that gap means the first two or three visible independents per metro capture outsized citation share for as long as the vacuum lasts.

The moat, once built, is real. AI answers lean on accumulated third-party evidence: years of reviews, a body of cited cost pages, recurring mentions in local threads. A competitor who starts eighteen months later has to displace you from sources that update slowly, while you keep compounding. Early visibility in slow-moving citation layers tends to persist the same way old authoritative pages held rankings for years.

The work is also unusually affordable relative to what the trades already spend. Lead platforms routinely charge $50 to $200 per shared lead. Reachroller's Starter plan is $29 per month for 400 credits and 25 tracked questions, with generated fix pages at 10 credits each, and the pages it produces are yours permanently. For the cost of a handful of shared leads per year, an operator can see the whole board and start filling the metro's vacuum deliberately.