Heating and cooling is a repeat-business trade, more than almost anything else in home services. Asked how they picked their contractor, a thousand homeowners who had hired one that year answered mostly the same way: 42.6% called back whoever they used last time, and 26.3% took a friend's word. That is seven in ten who never ran a search at all — and it is also why the September dip blindsides shops. Repeat work and referrals have no dial you can turn up, and no warning light before the phone goes quiet.
The same survey carries a number you should hear from us: 1.8% said they chose their contractor from an advertisement. It is last-touch and self-reported — somebody who clicked an ad, read your reviews, then called will report the reviews — so it undercounts advertising badly. But the shape is right, and the conclusion is not that ads don't work. It is that the fight for new customers happens in the bottom half of that list, and that the largest marketing asset an HVAC company owns is the customer list already sitting in its office. Reactivation and referral belong beside the ad account, not after it.
What follows is the research we use when we build marketing for heating and cooling contractors, in plain English: what your customers type, what your website is missing, where ad money goes to die, and how to hold a map-pack spot. Including two things that moved under the trade's feet this year and are probably still wrong on your site today.
How your customers actually buy
You run two businesses out of the same truck and they behave nothing alike. One is the emergency — not cooling in August, no heat in January — where 73.8% of homeowners expect the problem fixed inside a day and the job goes to whoever answers the phone live. The other is the changeout, on a completely different clock: most homeowners spend about three weeks researching a replacement and talk to at least two contractors before deciding. Two different sales. They should never share a campaign, a landing page, or a definition of success.
The second thing is what your customer walks in expecting. In that same survey, 21% reported unexpected costs or surprise fees on the job they had just hired for, and 5.2% said they were pushed toward upgrades they did not need. One in five of the people across your kitchen table is braced for it. That is also the opening: being plain costs you nothing. Publish the diagnostic fee. Put in writing that you size equipment off a real load calculation, not a rule of thumb — the 2024 residential code requires one where your state has adopted it, and most competitors still eyeball it off square footage. Say you are glad to be the second opinion; a shop with a commission-driven sales floor cannot copy a word of it.
Your next few years of changeouts are already installed, and you know where they are. Every system your shop put in twelve years ago is a job with a date on it — and 42.6% of homeowners simply call back whoever they used last. Nobody else in your market is bidding on that list, because nobody else has it.
Then watch the words. Customers do not say condenser, split system, or changeout — they say “my AC,” “the unit,” “not blowing cold,” and “no heat.” Plenty still say Freon, which stopped being correct decades ago and is still what they type. Repair and replacement searches split cleanly, and the split is worth money: “ac not cooling” is somebody on a hot phone at four in the afternoon, while “ac replacement cost” is somebody at a desk who will not call for a fortnight.
How 1,000 homeowners said they chose their HVAC company
- 42.6% went back to whoever they used last. The largest channel in the trade, and the only one nobody is bidding against you for.
- 26.3% took a friend's or a relative's word. Referral is a system, not luck. Most satisfied customers would recommend you; fewer than one in five is ever asked.
- 17.0% found their contractor on Google. The slice your website, your map listing and your search ads actually compete over.
- 7.7% came through reviews or a marketplace app. Yelp, Angi and the rest — the channel you rent rather than own.
- 1.8% named an advertisement. Last-touch and self-reported, so read it as a floor rather than a verdict.
What your website is actually for
Almost every HVAC site has the same flaw: one page called Services, with repair, replacement, maintenance, heat pumps, mini-splits, ductwork and air quality in a single bullet list. A bullet ranks for nothing and convinces nobody, and it breaks your organic search and your ad landing-page relevance at once. Those are seven different searches needing seven pages, each with its own price signal and photographs of your own work.
Then there is the page missing from most sites in the trade, and the business runs on it. The maintenance agreement's real job is to put you in the house on the day the system finally dies — the day the changeout gets decided. Sell it like a product: tiers, real prices, what each includes, and a signup that works without a phone call. And here is the part nobody says to HVAC owners: when companies in this trade are bought, the maintenance-agreement book is valued at two to three times its annual recurring revenue on top of what the buyer pays for earnings. Every agreement you sell is enterprise value, not just cash flow.
Three things close the big tickets. Financing on the page rather than a footer link, with a monthly number beside the system price. A published diagnostic fee, which in a trade where one in five customers reports a surprise on the bill reads as confidence, not a giveaway. And reviews with names and dates on them, not a floating star widget.
Last, and it takes twenty minutes: search your own site for the copy that stopped being true. The federal 25C heat pump credit terminated on 31 December 2025 and does not cover a 2026 install, and EPA cancelled the January 2026 R-410A installation deadline on 21 May 2026. Both are still sitting on a great many contractor sites.
The pages most HVAC sites are missing
- AC repair — where your mobile ad traffic lands. Phone number first.
- AC replacement and changeouts — financing, brand badges, reviews, quote form. Not the page above.
- Furnace replacement and heating repair — written for October and November, when the heat first fails.
- Heat pumps — shipments passed central air conditioners for the first time in 2025, and held after the credit died.
- Mini-splits and ductless — customers really do type “mini split.” A single zone runs $1,150–$2,800.
- Maintenance plans — tiers, prices, online signup. The highest-leverage page on the site, and usually absent.
- Ductwork and air quality — the attach work that lifts a changeout by thousands. Never has a page.
Google Ads: repair and replacement are not one account
The most expensive mistake in HVAC advertising is running one HVAC campaign. The two halves of the trade need separate budgets and separate definitions of a win — booked calls for one, booked in-home estimates for the other. One number makes the point: across SearchLight's 816 contractors and $14.88 million in Google Ads spend, an AC repair lead cost about $231 while an AC install lead cost about $157, a lead there being a tracked call or form submission rather than a booked job. Emergency keywords are the most contested inventory in the trade, so the small job costs more to buy than the big one — and repair calls still book better and often become changeouts on the spot, which is why the two must be measured apart, not averaged.
That spread runs through the whole trade. Cost per lead varies about four to one by service line in the same benchmark: AC maintenance near $86, heating repair $144, AC install $157, AC repair $231, heating install $354 at the top. So any agency quoting you one “HVAC lead cost” is guessing — the number depends entirely on which lines you sell. And generic campaigns are a tax: “HVAC general” came in at $198, worse than five of the eight named lines, because a broad campaign buys cheap traffic and expensive traffic together and you cannot tell them apart.
Whatever the budget, the first dollar goes to Local Services Ads. In the largest published dataset of those results, HVAC leads ran about $51 with roughly 44% of them booking. The screening is the point: Google wants your license, a million dollars of general liability coverage, and a background check on the owner. That is a real week of paperwork, and a wall the outfit undercutting you with a truck and a cell number cannot climb. Volume is capped, though, so fill Local Services Ads first and fund search with what is left.
Most of what gets wasted after that is avoidable. Negative keywords first: window and portable units, car AC, dryer vents, anyone shopping Home Depot for a capacitor, and the do-it-yourself and job-seeker clusters — how-to, wiring diagram, YouTube, apprentice, EPA 608. But one warning about the lists that circulate online: several tell you to block “compressor replacement,” and you should not, because that is a $1,500 to $3,500 job and one of the strongest changeout triggers there is. Read your own search-terms report weekly for the first two months. Then set the radius by ticket, not habit: a repair campaign inside a 15-to-20-mile core, a changeout campaign 30 to 40 minutes out, because a $12,000 job absorbs the drive and a $150 diagnostic does not.
Local SEO: what actually holds a map-pack spot
For “ac repair near me,” the three-business map pack takes most of the calls, and the gap between first and third is reported to be more than double the call volume. Start with categories, and do not set one and stop. Most shops belong under HVAC Contractor, with Air Conditioning Contractor, Heating Contractor and Air Conditioning Repair Service underneath. Google rewards category-query alignment: for “ac repair,” Air Conditioning Repair Service is a tighter match than HVAC Contractor. But never change your primary category without first writing down where you rank in each town — contractors have swapped a primary and watched a top-three position vanish.
Reviews are the currency and the one lever you can actually pull: proximity you cannot change, velocity you can. As working targets rather than thresholds, a top-three spot in a competitive mid-size market tends to sit between 150 and 400 reviews, and major metros run past 500 — with a rating of 4.5 or better mattering more than the count. A steady fifteen or twenty a month beats a burst of two hundred, which reads as bought. The trade has two natural asking moments and most shops use neither: the maintenance visit, and the day after a changeout runs, when the house is finally comfortable. And it gates more than the map — Trane's Comfort Specialist program requires dealers to hold a 4.6-star Google rating and requalify every year.
Content earns its keep when it answers a question with a dollar sign in it. The highest-traffic informational searches here are cost and comparison questions, both sitting right on the replacement decision: what a new system costs in your metro, heat pump versus furnace, repair or replace. “Why is my AC not cooling” is worth writing purely because it catches organically the customer you would otherwise pay about $231 to reach. Do not chase those queries nationally — Angi, HomeGuide and the manufacturers own every generic cost guide, and several exist to resell the lead. You win locally, which means a real grid of service pages by town, not one Service Areas page listing forty municipalities. Ten pages about towns where you have real jobs and real photographs beat sixty templated ones, which now carry genuine spam-policy risk.
One more opening, bigger than a year ago precisely because the federal credit died: rebate content. Most competitors' pages still describe incentives that no longer exist, so one accurate page on what your state and your utility actually offer can take the query outright. Re-check it quarterly — these programs open and close on a few weeks' notice, and a stale rebate page is the same credibility problem as a stale tax-credit page.