Google Ads for HVAC companies: seasonality changes demand and capacity quickly, so spend needs a verdict date before the busy period has passed.

Contents (10 sections)

• A no-heat call and a replacement quote are not one job
• What an HVAC enquiry is worth
• Seasonality is a hypothesis, not a calendar
• Response handling is part of the campaign
• Where HVAC ad spend leaks
• What I manage
• The first test has a boundary
• Search Ads and Local Services Ads are separate decisions
• Maintenance agreements change the seasonal math
• Start with three numbers

A no-cooling call in July and a furnace replacement in October do not share urgency, price, or sales time. The emergency repair is won by whoever answers today; the replacement is a proposal that can take weeks. Both have to earn their place in one account without the season drowning the verdict.

I run Google Ads for HVAC companies around the work your crews can actually serve: emergency repairs, air-conditioning and heating service, system replacement and installation, maintenance plans, indoor air quality, commercial work, and the locations that make the trip worthwhile. Every engagement gets a proving period, a written spend cap, and a verdict date before the next dollar moves: what I call The Payback Window.

The account should tell you which service the customer needed, whether the enquiry became a booked or dispatched job, what the job was worth, and whether the advertising earned the right to continue.

Start a conversation or ask for the free audit.

A no-heat call and a replacement quote are not one job

“HVAC” is a seasonally split business: the emergency call and the replacement quote do not share urgency, price, or sales time.

An emergency repair, a planned system replacement, and a maintenance agreement arrive with different urgency, ticket values, sales steps, and capacity requirements. A campaign that blends them can make a cheap repair enquiry look as valuable as a replacement opportunity, or let replacement clicks consume the budget meant for the calls your team can answer today.

I separate the intent that needs separating. The structure follows your service mix, service area, seasonal demand, margins, response process, and available crews. It does not follow a generic HVAC account template.

What an HVAC enquiry is worth

The value of an enquiry starts with the service and the outcome, not the keyword.

A no-cooling call may lead to a dispatch and a same-day repair. A replacement may need a diagnosis, an equipment decision, financing questions, an estimate, and a scheduled installation. A maintenance enquiry may become a plan the customer keeps for years, or a single tune-up. Those are different purchases wearing similar search language.

So I do not assign value from the keyword. The monthly review records the requested service, whether it became a booked, dispatched, or completed job, and what you confirm it was worth.

Stage 1 Tracked enquiry

A call or form arrives, tied to a search and an ad.

Stage 2 Answered and qualified

The service fits, the location fits, the timing fits.

Stage 3 Dispatched or scheduled

A technician is assigned, or an appointment is set.

Stage 4 Completed job

The repair, installation, or plan service was delivered.

Stage 5 Confirmed value

You confirm the ticket. This is the first honest revenue figure.

Cost per lead is a platform number. Cost per booked job is the one the decision rests on, and it is a division you can do yourself with the cost per booked job calculator before you commit a budget.

Seasonality is a hypothesis, not a calendar

Cooling demand, heating demand, and shoulder-season demand do not arrive on the same schedule. Testing cooling terms into summer and heating terms into winter, and resting the budget between the peaks, is a reasonable starting plan.

It is a plan, not a law. Climate, housing stock, equipment age, service mix, commercial contracts, and your own capacity all change the shape of the year. The search-term report, the enquiry record, and the booked-job record decide whether the hypothesis held. A flat year-round budget is not automatically wrong, and a seasonal schedule is not automatically efficient.

What matters for the money question is where the proving period sits. A test that opens as a season closes will produce a verdict about the calendar rather than about the account.

Response handling is part of the campaign

When a customer has no cooling in a heatwave, no heat in a cold snap, or a system that has failed outright, the response process decides whether the enquiry becomes a job.

The ad should not promise same-day or 24-hour service unless the business can provide it during the stated hours. The landing page should make the phone number, the service area, the response window, and the next step easy to find. The monthly review should show whether calls were answered, returned, scheduled, and booked.

Where HVAC ad spend leaks

Season mismatch. A flat budget can keep spending after the useful demand has moved on, or arrive too late for the weeks when your crews could have converted it. Scheduling is a decision to make with the account’s own seasonal record, not with a generic industry calendar.

Irrelevant intent. DIY and how-to searches, filters, parts and equipment shopping, manuals and model lookups, jobs, salary, and training research sit close to service language while representing people who are not hiring an HVAC company. These are categories to test in the search-terms report, not a fixed percentage of waste to repeat without evidence.

A blended service line. Repair, replacement, maintenance, indoor air quality, and commercial work can need different ads, landing pages, hours, follow-up, and job economics. Where the services differ materially, separate them enough to judge them honestly.

Geography. A broad metro target generates enquiries from locations that take too long to reach or fall outside the service area. Google Ads supports area and radius targeting, but location matching uses multiple signals and is not perfectly accurate. The account gets checked against actual enquiry locations, travel, and booked-job economics.

The unqualified call. A lead count can rise while booked jobs stay flat: the caller needs equipment you do not service, the property is too far, the crew is unavailable, or the follow-up breaks. Tracking the call is useful. Confirming what happened after it is the business answer.

What I manage

Repair and emergency campaigns

Built around the problems people call about now: no cooling, no heat, a failed compressor, a frozen coil, a system that will not start. The work is to control the search terms, match the ad to the service and the response window, keep the service area honest, and review whether calls were answered and whether they became booked work.

Replacement and installation campaigns

A replacement carries different economics from a diagnostic visit. The landing page explains what the business assesses, what it installs, the service area, the estimate path, and the next step. The acceptable acquisition cost comes from your confirmed ticket, labour, equipment cost, and stated margin, not from a generic industry benchmark.

Maintenance and recurring demand

A maintenance plan may continue past the first visit, which changes what a signup is worth. If you can confirm plan signups and the revenue that follows them, the review reports that outcome separately instead of hiding it inside a blended lead count. If the record stops at the first service, the page says so rather than assigning value the evidence does not support.

Commercial and light-commercial work

Commercial enquiries may involve a facility manager, a site visit, a proposal, access requirements, compliance questions, and a longer approval. If the business serves them, the campaign and landing page speak to that buying process rather than treating it as a larger residential repair.

Search terms, calls, forms, and booked jobs

I review what people searched, which ad they saw, where they landed, whether the enquiry was answered, what service was requested, and whether it became a booked or dispatched job. Call conversion tracking covers calls from ads and calls to a website number, and imported call conversions let a business count calls against its own outcome criteria.

The platform record tells me where the enquiry came from. Your confirmation tells me what happened after it arrived.

The first test has a boundary

There is no honest universal budget for an HVAC company. The right first test depends on service mix, average ticket, equipment and labour cost, service area, seasonal demand, response speed, margin, and how many jobs your crews can take right now.

Before launch, the fee, the working budget, the proving period, and the verdict date go in writing. The verdict is continue, adjust, or stop. If the numbers do not justify more spend, I will say stop.

I do not call revenue profit. Profit is estimated only when your margin is on file, and it is labelled estimated profit at your margin. Without that margin, the Ledger stops at confirmed revenue and says so.

More calls do not automatically justify more spend, if the calls do not become jobs the business can serve and defend.

Search Ads and Local Services Ads are separate decisions

For HVAC the two channels behave differently in July than they do in October, and that is the part a blended number hides.

Local Services Ads charge per lead, so in a heat wave you are buying leads at peak price against every other company in the metro, and the constraint stops being budget and becomes whether a crew can get there this week. Search Ads charge per click and let you separate a repair search from a replacement search, which matters when one is a $300.00 visit and the other is a $9,000.00 install.

So the two get separate budgets and separate outcome definitions. The install pipeline is judged on signed work over a longer window; the repair channel is judged on dispatched jobs inside the week.

Maintenance agreements change the seasonal math

A book of maintenance agreements is the closest thing HVAC has to stable revenue. Agreement customers produce tune-up visits, tune-up visits surface failing equipment, and failing equipment becomes the replacement work that does not need to be bought at peak auction prices.

So in the shoulder seasons the campaign buys the agreement, not just the emergency. Peak-week capacity still caps everything, but the plan base decides how much of the busy season is booked before the weather arrives, and that changes what a new customer is worth.

The monthly record shows the enquiry source, requested service, caller location, answer status, dispatch or appointment status, completion status, plan status where it applies, and confirmed value.

I count conservatively. Every untracked job is your upside, not my credit.

The Profit Ledger

A one-page answer, not an activity report

Money out Advertising spend Recorded to the dollar, for the period the verdict covers.

Money in Booked or completed HVAC jobs and plan signups Jobs, plans, and revenue you confirm. With your margin on file, profit is estimated at your margin and labelled that way.

Verdict Continue, adjust, or stop Decided on the verdict date, before the next dollar is committed.

The Ledger does not upgrade a lead into a booking, or a booking into profit. It records what the evidence supports and says where the record ends. That is the standard.

Start with three numbers

Tell me what the last marketing attempt cost, what a common booked HVAC job is worth after the costs you track, and how many jobs your crews can take now. If one number is missing, working it out is part of the audit.

Start a conversation or ask for the free audit. The first step is a read-only review. If the numbers do not justify more advertising, I will say so.

Questions HVAC businesses ask

Can HVAC Google Ads work outside the busy season?

They can be tested outside the peak, but the test has to match the demand you actually have. Heating, cooling, repair, replacement, and maintenance should not be judged as if they share one calendar. The verdict date is set before launch so a quiet month is not mistaken for a weak account, and a seasonal spike is not mistaken for a permanent result.

When should the proving period start for a seasonal HVAC account?

Start it where the demand you want to judge actually exists, and set the verdict date so it lands while that demand is still live. A proving period that opens as a season closes produces a verdict about the calendar rather than the account. The timing is written down before launch, alongside the spend cap, so nobody re-argues it later.

Should repair, replacement, and maintenance use separate campaigns?

Usually, when their economics, sales steps, or capacity requirements differ materially. Separation makes it possible to see which service line bought the booked work and which one consumed spend without earning its place. A smaller account can start simpler and add separation as the search-term and booking data justify it.

How does a maintenance plan change what a signup is worth?

A plan may continue past the first visit, so the first ticket is not the whole value. If you can confirm signups and the revenue that follows them, the review can report cost per plan signup separately from cost per repair call. If your record stops at the first visit, that is where the Ledger stops too, and it says so.

What HVAC searches should be excluded?

Begin by reviewing DIY and how-to searches, filters, parts and equipment shopping, manuals and model lookups, jobs, salary, and training research. Exclude the terms that do not represent a service you sell. Google Ads supports broad, phrase, and exact negative match types in Search campaigns.

What should an HVAC company track besides leads?

Calls, forms, answer status, requested service, caller location, appointment or dispatch status, completed jobs, ticket value, and the source you can defend. Cost per lead is a platform number. Cost per booked job is closer to the decision. Nobody deposits a lead.

What if my crews are already full?

Then more enquiries are not automatically the answer. A full calendar changes the question from how to get more enquiries to which enquiries are worth buying. The budget concentrates on the service lines your team can serve at the ticket and margin you need, and the verdict judges each line on that basis.

How will I know whether the advertising paid for itself?

With your margin on file, estimated payback is the date estimated profit at your margin crosses cumulative advertising spend. Without your margin, the honest milestone is revenue crossing spend, and revenue crossing spend is not profit. The Ledger names the difference rather than blurring it.

Last verified August 23, 2026