How HVAC Customers Find Contractors in AI Search (2026 Data)

Scorpion's 2026 State of Home Services Marketing Report surveyed 2,000 US homeowners aged 25+ who had hired a trades professional in the previous 18 months (fielded late 2025 with research firm Dynata). It found 83% of homeowners start their search online, and 22% now use AI tools such as ChatGPT to research or find recommendations. The same report surveyed 944 home-services operators — and 80% of them said they are unsure how to appear in AI-driven search results. That gap between homeowner behavior and operator readiness is the entire story of this page.

BrightLocal's Local Consumer Review Survey 2026 (1,002 US adults, via SurveyMonkey) puts the number far higher: 45% of consumers used AI tools for local business recommendations in the past year, up from 6% the year before. In that survey, AI became the third most-used tool for local business recommendations — behind only Google and Facebook, and ahead of Yelp and TripAdvisor.

Why the surveys disagree (22% vs 45%)

The two flagship numbers measure different things, and it pays to read them precisely. Scorpion asked recent trade-service buyers about their actual hiring research; BrightLocal asked general consumers whether they had ever used AI for a local recommendation in the past year. A homeowner who once asked ChatGPT about a noisy furnace counts in BrightLocal's 45% but may not count in Scorpion's 22%.

The framing also drifts as numbers travel. Contractor Magazine's May 2026 coverage rendered Scorpion's finding as "22% of homeowners now go to ChatGPT first, not Google, when they need a contractor." Scorpion's own release says 22% "use AI tools such as ChatGPT to research or find recommendations" — a broader and softer claim. The honest read is Scorpion's own wording: roughly one in five recent buyers now includes AI in their research, not one in five abandoning Google.

The direction is not in dispute

Whichever survey you trust, every measured trendline points the same way. BrightLocal's year-over-year jump from 6% to 45% is the steepest adoption curve it has recorded for any recommendation channel. Age splits show where it goes next: 64% of adults aged 30–44 asked AI for a business recommendation in the past year, against 24% of the over-60s — meaning the homeowners entering their prime repair-and-replace years are the heaviest AI users. Meanwhile the share of consumers using Google to read reviews slipped from 83% to 71% in the same survey. BrightLocal's CEO Myles Anderson put the operator consequence bluntly:

"If your reputation only exists on Google, you are effectively invisible to the millions of people using ChatGPT to find local services."

None of this says Google stopped mattering — 83% of homeowners still start online, mostly through Google surfaces. It says the path from "AC is blowing warm" to a booked call now routes through more surfaces than the map pack, and one in five to nearly one in two of your future customers is already asking a machine for a shortlist.

Do homeowners trust what the AI tells them?

Trust in AI recommendations splits sharply between people who use the tools and people who don't. Among active AI users in BrightLocal's survey, 63% trust AI tools' business recommendations and only 10% distrust them; among non-users, 53% say they don't trust AI's business recommendations at all. More striking for an industry built on star ratings: 42% of consumers already trust AI platforms as much as online reviews when choosing a local business, and among AI users that figure reaches 64% — parity with a trust signal that took reviews two decades to earn.

AI-generated review summaries pull the same direction. Half of consumers trust AI platforms to accurately summarize reviews from real people (71% among active AI users), and 23% say they are happy to rely on the summary alone when making a decision. For an HVAC company, that means a machine's three-sentence digest of your review profile is now, for a meaningful slice of homeowners, the review profile. What feeds that digest — steady, specific, recent reviews — is suddenly doing double duty.

We verified how far this has already gone on our own primary keyword. On August 13, 2026, we pulled a live US desktop SERP for "ai search for hvac contractors" through a rank-tracking API: Google served an AI Overview above the first organic result, and every People Also Ask question beneath it expanded into its own AI-generated answer rather than a snippet from a single page. For this query — and for a growing share of HVAC queries — the first screen of Google is already AI-written. Which HVAC queries trigger these answers, and who gets cited in them, is mapped in our AI Overviews for HVAC breakdown.

Microsoft Copilot and Claude trail those five in BrightLocal's data but are being tried at far higher rates than a year earlier. The practical consequence, as Footbridge Media's analysis of contractor AI search notes, is that a contractor cited inside Perplexity can be invisible inside ChatGPT and vice versa — different tools, different source piles, different answers to the same question.

How AI questions differ from a Google search

Homeowners ask AI engines longer, context-loaded questions than they ever typed into Google. Footbridge Media's write-up contrasts the classic search — "plumber Pensacola" — with what the same homeowner now types into ChatGPT: the symptom, what they already tried, the age of the house, the neighborhood, the review requirement, and their availability, all in one prompt. The HVAC equivalent is no longer "AC repair Dallas" but "my upstairs AC is blowing warm air, the unit is a 2009 Trane, I'm in Lake Highlands, and I need someone this week who's good with older systems."

That question format changes what wins. A single prompt carries the trade, the equipment, the location, and the constraint — and the engine matches all of it at once against what it can read about every company in the market. A generic "cooling services" page gives it almost nothing to match. Which is exactly why the next question — what the engines actually cite — decides who gets named.

One term needs pinning before the numbers: a "citation" here means a source link the AI engine displays under its answer — not the name-address-phone directory listings that local SEO also calls citations. BrightLocal ran 800 manual local business searches in ChatGPT in November 2024, recording the first ten sources behind each answer, using the same search terms as its earlier Google study so the two are directly comparable. The data is a snapshot — engines change fast, and figures this granular should be dated, not treated as permanent — but it remains the most transparent public look at where local AI answers come from.

Four findings in that study should reset an HVAC owner's assumptions:

• Your own website is the single biggest input. For transactional local searches, business websites made up 72% of ChatGPT's sources. The engine is reading contractor sites directly — service pages, FAQs, location details — and building recommendations from them.
• The directories that show up are not the ones you'd guess. Yelp, Facebook, and Google Maps did not appear as directory sources anywhere in the study. Three Best Rated (24% of directory sources) and Expertise (18%) led instead — sites most contractors have never optimized for.
• Mentions carry real weight. Third-party mentions — trade press, local news, Wikipedia (39% of all mention sources) — took a 27% share, far above their 16% share in Google results. Companies that exist only on their own domain give the engines less to corroborate.
• Reviews still feed the machine. BrightLocal notes ChatGPT sources local business review information, including Google reviews, even while Bing's index does most of the retrieval. Reviews are read by engines assembling answers and by the humans double-checking them.

The mix also shifts with the kind of question asked. For transactional searches ("hvac repair chicago"), business websites dominated at 72% of ChatGPT's sources — the engine goes straight to contractor sites. For informational searches ("best hvac company chicago"), the split levels out: 43% business websites, 35% business mentions, 22% directories. In plain terms, when a homeowner asks who is best rather than who is nearby, the engines lean harder on what third parties say about you — press coverage, roundups, and directory profiles you may never have looked at.

The pattern behind all four findings: AI engines recommend companies they can verify from multiple independent directions. A crawlable website that answers real questions, consistent business details across profiles, reviews on surfaces beyond Google, and third-party mentions each corroborate the others. The mechanics of that pathway — the Bing index gate, entity consistency, which profiles ChatGPT actually reads — and how to influence each one are the subject of our ChatGPT recommendation guide for HVAC companies; this page's job is the evidence for what gets counted.

The click-loss evidence is real and worth stating plainly. Pew Research Center's July 2025 study tracked 900 US adults across 68,879 real Google searches in March 2025: 18% of searches produced an AI summary, and users who saw one clicked a traditional result on just 8% of visits, versus 15% without a summary. They clicked a link inside the AI summary on only 1% of visits, and ended their browsing session entirely on 26% of AI-summary pages versus 16% otherwise. Ahrefs' 300,000-keyword analysis found the presence of an AI Overview correlated with a 34.5% lower average click-through rate for the top-ranking page, with position-1 CTR on AI Overview keywords falling from 0.073 to 0.026 between March 2024 and March 2025. Footbridge Media reports a roofing client losing 31% of organic traffic on a single query while holding the #1 ranking — the result page simply reshaped around the answer.

One detail inside the Ahrefs data deserves more attention than it usually gets. Position-1 CTR was already declining before AI Overviews arrived: Ahrefs forecast it would have fallen to 0.040 by March 2025 without AI Overviews, versus the 0.026 actually measured on AI Overview keywords. So the honest attribution is that AI Overviews explain a large share of the click loss — but not all of it. SERPs were getting more crowded with ads, packs, and widgets either way, and blaming AI for the entire decline overstates its bite.

Two qualifiers bound that evidence before you rebuild your budget around it. First, both large studies measured broad and mostly informational search behavior — not "AC not cooling emergency" queries, where the searcher needs a truck, not an answer, and a phone call is the end goal. Second, fewer clicks is not the same as fewer customers: Contractor Magazine reports Semrush data showing traffic arriving from AI recommendations converts at 4.4 times the rate of standard search. BrightLocal's survey adds the behavioral reason: 97% of AI users at least sometimes double-check AI recommendations against real reviews, and 42% always verify on the native review platforms. AI is the first step of research, not the final authority — which means the recommendation gets you onto the shortlist, and your review profiles still close the job.

Weigh both sides honestly and the 2026 picture looks like this: AI surfaces are intercepting research-stage queries and shrinking clicks on informational content, while high-intent local queries still resolve to a call — increasingly influenced by what the AI said first. Whether the click losses have actually reached HVAC lead counts, including the counter-evidence that they mostly haven't yet, gets a full data treatment in our honest analysis of AI Overviews and HVAC lead loss.

This is the demand-side page, so what follows is the owner's checklist, not the full playbook:

1. Run the homeowner's test. Open ChatGPT, Perplexity, and Google's AI Mode and ask what your customers would ask: "who should I call for a heat pump repair in [your city]?" Run each prompt more than once — answers vary between runs, which is itself worth knowing before you pay anyone to change them.
2. Read what the engines read. The evidence above says the inputs are your website's service pages, your reviews across platforms, your directory profiles (including unglamorous ones like Three Best Rated), and third-party mentions. Check each for accuracy and consistency before spending on anything new.
3. Keep reviews compounding — beyond Google. Scorpion's data shows 87% of homeowners will not hire a business rated below 4 stars — while 67% of business leaders in the same survey admit they struggle to collect reviews consistently. BrightLocal's data adds that fact-checkers head straight to review profiles after an AI recommendation. Reviews now work twice: as engine input and as human verification.
4. Measure AI referrals now, even while small. Referrer traffic from chatgpt.com, perplexity.ai, and copilot.microsoft.com shows up in analytics today. Establish the baseline before you invest, so any program you buy can be judged against it. Scorpion's operator survey found 67% of home-services businesses cannot connect marketing spend directly to revenue — the businesses that can are the ones that will know whether AI visibility work actually paid.
5. Refuse guarantees. No vendor controls what ChatGPT or Google's AI answers say. An agency promising "we'll get you into ChatGPT" is selling certainty that does not exist; the credible version is measured — baseline prompts, documented changes, re-measured prompts. That standard is exactly how we've structured our own AI search visibility service, and you should hold us to it too.

The same homeowner expectations that drive AI adoption keep running after the recommendation lands. Scorpion's survey found 56% of homeowners want 24/7 scheduling or a way to communicate after hours, while 66% of home-services businesses name after-hours customer service as their top challenge. A ChatGPT recommendation delivered at 9:40 pm is worth nothing if the recommended company's phone rings out — the visibility problem and the answering problem are two halves of the same lost job.

One more honest note: parts of this work overlap heavily with traditional SEO, and some agencies argue AI visibility is nothing more than good SEO done well. There is real truth in the overlap and real differences at the edges — we've laid out where each argument holds in GEO vs SEO for HVAC companies, so you can decide what to buy (or not buy) with the trade-offs in view.