HVAC marketing follows the weather, so plan for the quiet months

Every piece of HVAC marketing, from a $28 tune-up mailer to a Local Services Ads budget, exists to make a homeowner call this company when the system quits, and again when the next tune-up is due. HVAC advertising is the paid slice of that: search ads, Local Services Ads, social ads, mailers. The rest runs on lists the company already owns and on whoever picks up the phone.

Take a shop starting from zero. One experienced owner opened a new HVAC company on June 15, 2026, and reported the numbers from the first month on. Most first-month revenue came from Meta ads and direct mail pushing a $28 tune-up special. Replacements mostly went at significant discounts; the one sold at book price made a 56.7% gross margin. The owner's read on that month: "as soon as there is a lapse in any marketing campaign, the phone stops ringing."

After July the owner was ready to cut Meta spend, saying those ads typically bring in "mostly price shoppers, people looking for a bargain." They were also getting a plumbing license to carry the shoulder season and feed HVAC leads.

That August week brought $17,910 in revenue, most of it one install. The weekly report split 14 new customers by source: Facebook ads 4, outbound calling 3, direct mail 2, direct mail prospecting 1, business card 1, not recorded 3. An experienced owner with a weekly report, and 3 of 14 new customers still had no source.

The same owner ran an earlier company that, by their account, scaled past $7M across HVAC, electrical and plumbing. There, Facebook ads converted "10 leads 5 booked 2 sold." Radio and direct mail both ran all year, with mail spend bumped "as seasonal demand starts." Mail was "my personal favorite," easy to track. Marketing as a whole, the same owner wrote, "is more about feelings and can be difficult to track."

So Facebook got two verdicts from one owner: a working funnel at the old company, a price-shopper channel at the new one. That is a reason to distrust any ranking of HVAC channels that leaves out the offer and the month. We made the case for counting calls per source in local business marketing. HVAC adds one more variable to the count: the season.

One owner of an HVAC, plumbing and electrical company in Northeast Ohio ran into that in a mild winter, in February 2024. Their read: "In low demand months, relying on PPC and LSA to drive leads is difficult due to low demand. Online marketing tends to not work as well since it's demand-driven." Search ads and Local Services Ads catch people who are already looking. In a warm February, fewer people look.

So that owner leaned on work the company could start by itself. They listed eight levers for that winter. Here is what six of them brought back, in that owner's words:

• Door hangers, put out by the techs: 2 jobs booked within 24 hours.
• Segmented email, with promos matched to open capacity: "10 to 20 appointments booked per email."
• A text to members due for their annual service: "within 3 hours we had 15 calls and 5 booked jobs."
• Outbound calling: 20 to 30 appointments a day, the lever that owner said was "moving the needle the most."
• Memberships: a "game-changer" for the slow stretch from February to April.
• Postcards: more of them sent, with the payoff from December's round showing.

The other two, promos on inbound calls and on TikTok and Meta, were described as working but came with no number.

Not everyone bought the door hangers. One critic mocked paying "$50/hr techs" to hang them. Fair point. The owner called the hangers low cost but gave no dollar figure, so the wage and the two jobs never meet on one page.

Memberships are the odd one out: the visits are sold before the slow months arrive. An HVAC operator who also sells marketing to the trade put the stakes this way: "October is when you find out whether you built a business or a summer. The companies with maintenance agreements are booked into November." The text to members due for service is the same move as maintenance plan reminders. The list and the due dates already exist. Someone has to send it.

Then the same Ohio owner complicates the picture. Two years later they bought a company in January that was doing about $100K a month, raised its Local Services Ads spend from $1K to $5K a month, and reports: "By February, sales hit $240K." By that owner's account, bigger Local Services Ads spend paid off in midwinter, from a small base of $1K a month on a company that size. Why it worked, the reported numbers don't say.

None of this argues for switching paid ads off when it gets cold. That owner's own January says otherwise. But search demand follows the weather, and no owner sets the thermostat. The members due a visit and the past customers already on file can be reached by text, email, phone or mail on any Tuesday in February. That's the lever an owner actually controls.

Where should the first marketing dollar go? Buy leads you can count this month, or spend on a name people remember when the furnace quits. The HVAC owners in our research argued it both ways. Each row below is one case, and the Middle Tennessee operator gets two.

Read across the rows and the split tracks stage and cash more than taste. The wrap math from Middle Tennessee is an argument: it assumes 10% of bought leads close, a rate the operator called "abysmal". Their own rule was to buy leads direct "until profits/size allow to properly run branded marketing."

The young founder argues that mom-and-pop shops take 40+ years to reach $10M in revenue, and gives the reason: "Because they don't run paid ads." Yet by the founder's own account, their company hit its organic target before it spent "a $ on marketing." And the 50-year company isn't weighing brand against leads at all. Its ad costs keep climbing, private-equity rivals can outspend it, and the owner keeps raising prices to pay.

Reviews sit between the two. The operator quoted earlier on October, who also sells a review tool to the trade, pays techs $20 for a review collected in person on the job and $10 if it comes in after the tech leaves. Their argument: reviews keep the ad budget lower against private-equity bidders on paid search and Local Services Ads.

One owner of a large home-services company saw a department spend over $1,000 for six leads and said the figure "tells me absolutely nothing." To them, cost per lead alone is a vanity number. "A cheap lead that never books, or cancels the next day, is worse than an expensive one that turns into a sold job." So they track cost per booked lead and cost per converted lead. "Your 'expensive' source is sometimes your cheapest."

The same owner wrote, "I thought marketing was our only lever. It wasn't. It was the phone." In their words, "The person who answers our phone has more impact on our success than my entire marketing budget."

By its CEO's count, the contact center of a four-business home-services group took 4,099 inbound calls in the week of December 2 to 8, 2025, its first week with AI on the front line, answering alongside people. Of those, 1,117 were booked live, about 27%, and 390 more booked later took the week to about 37%. Those raw rates count every call, short and spam calls included, so they aren't a close rate. Before the change, human daily booking rates there had run from 19% to 39% depending on the season. With AI answering, the first week's daily rates ran 19% to 28% across its businesses, and the CEO graded the rollout B to B+.

Not everyone wants AI on that line. Someone responding to the multi-trade owner's point about the phone argued that answering calls "may be dead last" on the list, since people come to small businesses for the personal touch. It allowed that AI might be worth it for a company without the scale to have a person available.

The Northeast Ohio owner's report on a month of search ads stops at the lead: $4.59K spent, 698 clicks, 84 phone-call leads. Nothing in it says how many of the 84 booked or sold. The Facebook count from the first section went two steps further, to booked and to sold.

Say an HVAC owner brings us two sources, with $500 spent on each. Source A costs $50 a lead, so $500 buys 10. Half of them book: 5 jobs, or $100 per booked job. Source B costs $20 a lead, so $500 buys 25. One in five books, also 5 jobs at $100 each. On a cost-per-lead report, B looks two and a half times cheaper. At the booked job they tie. Now say two of the five jobs B would have booked came in as evening calls nobody picked up. That leaves 3 for $500, about $167 each, and the cheap source is now the expensive one.

What we'd do is log every call, form and text with its source, whether it booked and whether it cancelled, then read each month as cost per booked job for each channel. Answering and booking those calls quickly is the other half of the job, covered in HVAC lead response.

Whoever is meant to answer, a call that rang out at 7 p.m. still belongs in that log the next morning, next to its source.