Residential HVAC is one of the most expensive categories on Google Ads, and one of the easiest to run badly. The problem is rarely the ads themselves. It is structural: repair calls and replacement quotes sharing a single campaign, a service area that includes towns you would never drive to, and conversion tracking that counts every form submit as a lead whether or not anyone answered the phone.
This guide covers how that account should be built and run: keeping service lines separate, planning campaigns, allocating budget across the year, reading published cost data for what it is worth, and tracking the things that make the numbers mean something. It is written for owners and general managers of residential heating and cooling companies in the United States.
What this guide does not claim. It contains no HVAC client results, no campaign screenshots, and no guaranteed outcomes. Every cost figure is labeled as one of four things: a platform estimate, a vendor benchmark, a competitor-reported range, or a worked calculation. The economics section is deliberately short, because the full cost-per-lead model belongs in its own article.
How Google Ads works for a residential HVAC contractor
Google Ads is an auction. You bid for a chance to show an ad against a search, and you pay only when someone clicks. Bid size matters, but it is not the only thing that decides who wins. Relevance between the search, the ad and the landing page counts too, which is why a tightly themed campaign usually beats a broader one at the same bid.
Three ad formats matter, and they are not billed the same way.
Search ads
They appear above and below the organic results. They reach people who have already typed the problem: "AC not cooling," "furnace repair near me," "heat pump installation [city]." Intent is high, and so is the cost per click.
Performance Max with pay-per-lead goals
This is where Google Local Services Ads now live. Google is moving Local Services into the Google Ads interface as a specialized Performance Max campaign type: automated campaigns that, in this format, charge per valid lead rather than per click and serve only on Search and Maps. The rollout is phased: the first phase began in August 2026 for select US home and storefront service advertisers, including HVAC. Broader service-area businesses follow in late 2026, and non-US accounts and remaining categories in 2027 (Google Ads Help).
The billing model does not change. You still pay for valid leads rather than clicks, whether those arrive as calls, messages or bookings. Ads still run only on Google Search and Maps, and targeting stays keywordless, driven by service categories and service areas. What does change matters operationally:
- Local Services performance reports do not carry over. Google recommends downloading or screenshotting your historical reports before the switch.
- Manual bidding and industry-level target CPA are being retired. After the migration, Google Ads no longer supports manual bids, and a target CPA set per trade category is replaced by a single campaign-level figure.
- Weekly budgets become daily averages. Google divides your historical weekly budget by seven, and monthly spend is then capped at that daily figure multiplied by 30.4.
Standard Performance Max campaigns are a different product: billed per click and able to serve across Google's wider inventory, including YouTube, Gmail and the Display Network. The per-lead billing and the Search-and-Maps-only placement described above belong to the specialized Local Services format, not to Performance Max in general (Google Ads Help).
Display and YouTube
They are usually the lowest priority for a residential contractor. They become useful for remarketing later, once search and pay-per-lead are producing work you can actually measure.
Pay-per-click and pay-per-lead are two different bills
A cost per lead from Local Services and a cost per click from a search campaign are not comparable numbers. Blending them produces a figure that describes neither, and it is one of the quickest ways to misjudge an HVAC budget.
Set side by side, the two paid options differ in how they are billed, what they target and what you can report:
| What differs | Search ads | Local Services, as pay-per-lead Performance Max |
|---|---|---|
| Billing | Per click | Per valid lead |
| Targeting | Keywords, match types, negatives and service areas | Service categories and service areas; no keywords |
| Where the ads appear | Google Search results | Google Search and Google Maps only |
| What you can report | Clicks, conversions and cost per conversion | Leads you were charged for and cost per lead; historical Local Services reports do not migrate |
Service-line separation: the decision that shapes everything else
A residential HVAC company sells at least four different things to four different kinds of buyer:
| Service line | What the searcher wants | Typical urgency |
|---|---|---|
| Emergency repair | Someone today; the system has failed | Hours |
| Replacement / install | A quote, financing, a decision timeline | Days to weeks |
| Maintenance / tune-up | A seasonal appointment, a plan | Weeks |
| Indoor air quality, ductwork, other | Information, then a quote | Variable |
These are not one conversation. Someone with no cooling in July is not comparing heat pump efficiency ratings, and someone pricing a replacement is not going to book an emergency dispatch. If both searches trigger the same ad and land on the same page, one of them is being served badly.
When to split into separate campaigns
Google's own guidance is the cleanest rule available: create a separate campaign whenever the budget or the locations differ, and keep them as ad groups inside one campaign when they do not (Google Ads Help). Applied to HVAC, that leaves three decisions:
- Split emergency repair from replacement if you want to cap what you spend on emergency calls, or if you advertise replacement across a wider radius than you will actually dispatch a truck.
- Split maintenance if it runs on a different budget rhythm: a small always-on spend rather than a seasonal surge.
- Keep services together as ad groups only if they genuinely share a budget and a service area. Accept that you will not be able to pace them separately.
The pay-per-lead format has the same constraint in a different form: it applies one target CPA across every trade category, and Google's documented workaround is a separate campaign for each trade (Google Ads Help).
What each service line needs from its campaign
Emergency campaigns need call assets, ad schedules that match the hours someone actually answers, and copy that states your response time honestly. Replacement campaigns need a clear quote path: financing if you genuinely offer it, and a form that asks enough to qualify a lead without filtering out real buyers. Maintenance campaigns need one specific offer with visible terms.
Do not reorganize the account after you start spending
Google's documentation notes that reorganizing an account later "wipes out all the valuable data you've accumulated," and that can affect how the ads perform. Decide the structure before you spend, not after. Whatever you choose, name things so a stranger could read them: Search | Repair | Emergency | [metro] tells the next person exactly what it is. Campaign 1 does not.
Planning a search campaign that reaches the right ZIP codes
The build order is deliberate: geography first, then keyword themes, then negatives, then the landing page, then the assets. Get the service area wrong and every later decision is aimed at the wrong people.
Start with the service area, not the keywords
Target the cities and ZIP codes you actually drive to, and exclude the rest. Then check the distance and geographic reports after two weeks. A surprising share of wasted spend shows up there, in clicks from towns just outside your boundary.
Build keyword themes per service line
Each ad group should hold to one tight theme: emergency AC repair, furnace repair, heat pump installation, maintenance tune-up. Phrase and exact match are the safer starting point for a local service business. Broad match on a head term like "HVAC" pulls in job seekers, parts buyers and do-it-yourselfers faster than you can add negatives.
Negative keywords are a weekly habit, not a one-time setup
The search terms report is where money leaks. In HVAC, the searches that end up being blocked usually fall into a few groups:
- Jobs and salaries.
- Wholesale and parts.
- Do-it-yourself repairs and troubleshooting.
- Rentals and landlord-tenant questions.
- Classes and certification.
- City names outside your service area.
Add negatives based on what you see in the report, not from a list you downloaded. Remember that negatives do not match close variants, so a plural, a singular or a synonym has to be added separately, while casing and misspellings are handled for you (Google Ads Help). Watch too for negatives so broad that they block work you actually sell.
Match the ad to the page
If the ad promises same-day AC repair, the landing page should say so, put the phone number where it can be seen, and name the cities you serve. Mismatch between ad and page is one of the most common reasons a technically sound campaign converts poorly.
Use the assets that cost nothing extra to test
Three asset types are worth setting up before you raise any bid:
- Call assets, Google's current name for call extensions, put a tappable number under the ad on mobile. On desktop, a "Call us" button reveals it. Schedule them for the hours someone answers.
- Location assets reinforce the service area by showing your business location next to the ad.
- Sitelinks let a searcher choose between repair, replacement and maintenance, which is useful precisely because those are three different jobs.
Allocating budget across service lines and seasons
Budget follows capacity and job value, and it moves with the season. An even split across service lines ignores both.
Do not divide the budget evenly
Weight it by two things: the work you can actually deliver, and what the job is worth. If you have two install crews in July, do not buy install leads you cannot service. If the install board is full and the service techs are idle, move money toward repair and maintenance.
Plan the year in three phases, not twelve months
- Heating peak (late fall through winter): furnace repair and replacement demand rises, so heating budgets should be at their highest.
- Cooling peak (late spring through summer): the same picture on AC terms.
- Shoulder months (spring and fall): put budget into maintenance plans, tune-ups and replacement research, the work that fills the calendar and creates next season's repair customers.
Set a floor before you set a target
Bidding needs conversion data to learn from. A budget small enough to produce a handful of conversions a month will not get out of the learning phase, and the account will look like it "does not work" when the real problem is a lack of volume. That is a structural limit, not a promise about results.
Change budgets on a schedule, not in a panic
A weekly review of search terms, budgets and pacing is enough for most single-market contractors. Swinging the budget midweek because of one slow Tuesday creates more noise than signal.
What HVAC Google Ads cost: reading the published evidence
There is no single HVAC cost per lead, and no single source can give you one. What the published figures can do is show the shape of the market and the range of methods behind it. Most guides lose the thread here because four different kinds of number get mixed together, so every figure below is labeled as a platform estimate, a vendor benchmark, a competitor-reported range, or a worked calculation.
Keyword Planner estimates are not prices
Keyword Planner returns a rounded average monthly search volume and a top-of-page bid estimate for a keyword. For the terms closest to this topic, the numbers are small. "hvac ppc" returned 170 monthly US searches with a $53.80 average CPC estimate and a competition index of 21. "hvac advertising" returned 390 searches at $33.00 with an index of 25 (Keyword Planner via DataForSEO, US, retrieved 7 November 2026). Note what those are: national averages, in volume buckets, for queries that agencies and software vendors bid on. They are not what a contractor in a specific metro pays, and they are not search difficulty. They are also the only measured figures in this guide, and neither is a budget.
Published benchmarks describe someone else's sample
The most detailed public benchmark for this category is SearchLight's HVAC and plumbing advertising benchmark. It covers 816 contractors, 8,077 campaigns and $14.88M in Google Ads spend from 1–31 January 2026, with 143,008 leads and $77.7M in tracked closed revenue. It reports a blended cost per lead of $104, split into branded search at $34, non-branded search at $149 and Performance Max at $72. That Performance Max figure covers standard campaigns, which the publisher describes as serving across Search, Display, YouTube, Gmail and Maps, not the pay-per-lead Local Services format. Heating repair campaigns came in at $144 per lead with a 38.2% book rate and a $3,225 average ticket, and AC repair at $231 with a 37.0% book rate and a $3,174 average ticket (SearchLight Digital, published February 2026).
Those figures are useful, and they are also easy to misuse. Six caveats matter:
| Caveat | Why it matters to you |
|---|---|
| One vendor's client base | It is first-party data from contractors using SearchLight's attribution product, not a market sample. Contractors who buy attribution tooling are not the median contractor. |
| January 2026 only | One month, in heating season. The publisher itself calls heating-repair strength seasonally expected. An AC campaign's economics in July are not in this dataset. |
| HVAC and plumbing are aggregated | The headline $104 blends both trades. It is not an HVAC benchmark. |
| Conversion and CRM limits | Leads depend on each account's tracking, with offline conversions "where available". Two contractors may count a lead differently. |
| No CPC published | The dataset reports cost per lead, not cost per click. You cannot convert one into the other from this page. |
| Not a universal benchmark | It is a comparison point, not an answer. |
Competitor-reported ranges disagree with each other
The gap between those published ranges is the useful signal. EZE Grow's 2026 budget guide reports cost per lead of $25–$80 and cost per booked job of $60–$150, with CPC ranges by keyword type: AC repair $8–18, furnace repair $7–16, HVAC installation $12–25, and emergency keywords $15–35 (EZE Grow). The HVAC SEO Agency publishes a different split by category: emergency repair $25–80, general service $15–50, installation $20–60, maintenance $8–25 and brand $1–5 (hvac-seo.agency). Neither methodology is published, and neither is verified here. That is the point: if two agencies in the same category publish ranges that do not overlap, a benchmark is not a budget input.
How to use a published benchmark without copying it
Treat a published benchmark as a sanity check on your own numbers after 60 to 90 days, in your market, with your tracking. Before that, your cost per lead is unknown, and any number you plan against is a placeholder.
A short economics primer: leads, qualified leads and booked jobs
Four numbers drive everything, and they are routinely confused with each other:
- Leads: every inbound contact attributable to the campaign, whether a call, a form submission or a message.
- Qualified leads: leads that meet criteria you have written down and apply consistently, such as being in your service area, from a decision-maker and for work you actually sell.
- Booked jobs: leads that turn into a scheduled visit.
- Sold jobs: booked jobs that become paying customers.
Two rates follow, and the denominator is the entire argument:
- Booking rate = booked jobs ÷ all inbound leads.
- Qualified-to-booking rate = booked jobs ÷ qualified leads.
Those two rates describe the same funnel from different starting points, so the denominator has to be named every time. Using a cost per qualified lead with a booking rate calculated from all leads overstates the cost per booked job, and the size of that error depends on the qualified-lead rate.
One formula carries the rest of this guide, and it only holds when both terms count the same leads:
Cost per booked job = cost per lead ÷ booking rate, with both the cost per lead and the booking rate counted over the same set of leads.
A $150 lead at a 30% booking rate works out to $500 per booked job. The lead price did not change. The schedule did. That is why conversion work, whether it is answering faster, qualifying on the phone or offering a booking path, moves the number that matters without touching a single bid.
Five cost numbers get used in HVAC reporting, and they are not interchangeable. Each one divides the same spend by a different count:
| Metric | Spend divided by | What it tells you |
|---|---|---|
| Cost per click | Clicks | What the auction charged before anyone reached your site |
| Cost per lead | Leads, meaning calls, forms and messages | What a contact costs, whether or not it was worth pursuing |
| Cost per qualified lead | Leads that meet criteria you wrote down | What a contact worth pursuing actually costs |
| Cost per booked job | Jobs on the schedule | What an appointment costs |
| Cost per sold job | Jobs that turned into paying customers | What a customer costs |
What a booked job can afford
It depends on your average ticket and your margin, not on a benchmark. Three sources work through it. SearchLight's benchmark article runs a ticket-and-margin example. Inshalytics publishes a break-even illustration using a $500 repair ticket at 45% gross margin. VelocityPPC describes a budget method built on jobs, close rate and a target cost per booked job (Inshalytics, VelocityPPC, published September 2026).
None of them reconciles an all-lead booking rate against a qualified-lead rate, or works out a maximum allowable cost per booked job for each service line. That belongs in a companion article on HVAC cost per lead, along with the full model, both denominators and the worked examples. If you need the ceiling calculation now, treat it as separate reading rather than something to extrapolate from here. For the account work itself, see HVAC Google Ads management.
Lead quality and conversion tracking
If the tracking is wrong, every decision after it is wrong too. Three things need to be true before you judge a campaign.
Count the actions that represent a lead
At a minimum, count form submissions and phone-number taps, plus inbound calls if you use call tracking. Set them up as conversion actions in Google Ads, which is how the platform learns what a lead looks like, so bidding optimizes toward outcomes instead of page views. Be clear with yourself about what each one proves. A tap on the phone number is a proxy for a call, not proof that one happened, and a form submit is a proxy for a lead, not proof that anyone answered it.
Expect the platform's number to be optimistic
Duplicate form submissions, spam and mis-clicks all inflate reported conversions. Compare what Google reports with what actually reached your dispatch board or inbox. When the two diverge, trust the one closer to revenue.
Import outcomes only if you have the plumbing for it
Google Ads can accept offline conversions imported from a CRM, which is how a booked or sold job gets back into bidding. Whether you can do that depends on a CRM with the data, a reliable match key and someone to maintain it. Many contractors do not have all three, and that is a legitimate place to stop. Without a CRM and an offline conversion process, reporting stops at the lead. You can know cost per call and cost per form. You cannot know cost per booked job, because the booking happens in your dispatch system, not in Google Ads.
Two attribution caveats are worth stating plainly. Last-click attribution credits whichever channel touched the customer last, which under-credits earlier touchpoints and over-credits branded search. Local Services performance reports also do not transfer into Google Ads during the migration, so any comparison across that transition needs deliberate handling rather than assumption.
Operational mistakes that quietly waste budget
These show up in a lot of HVAC accounts, and each one has a cheap check.
| Mistake | What it costs | How to check |
|---|---|---|
| One campaign for emergency, replacement and maintenance | Budget flows to whichever line bids hardest, not the one you need | Open the campaign list. If one campaign contains all three, it is happening. |
| No negatives, or a stale list | Spend on parts, jobs, DIY and rentals | Search terms report, last 30 days, sorted by cost |
| Service area set by a radius you do not serve | Clicks from towns you will not drive to | Geographic and distance reports |
| Broad match on head terms | Volume without qualification | Check match type on the terms consuming most spend |
| Ad promises same-day service; page does not | Paid clicks that never call | Open the landing page on a phone in one hand while reading the ad |
| Budget below the data threshold | Learning never completes; the account "does not work" | Conversions per month against a realistic minimum |
| Performance Max launched before tracking is clean | Budget spent on channels you cannot see, judged on conversions you do not trust | Conversion action audit first, PMax second |
| Slow response to form leads | The lead is bought and then lost in the inbox | Time-stamp five recent form leads from submission to first call |
The last one is not an ads problem, but it behaves like one. A lead that sits for two days makes a functioning campaign look broken.
Working with an agency or freelancer
Most contractors eventually hand this to someone. Three questions separate a useful conversation from a sales call:
- How is the fee structured: fixed, hourly, or a percentage of spend? Get the number and the scope in writing, and check whether reporting and strategy are included or billed separately.
- What exactly will be reported each month, and what cannot be? You want calls and forms counted separately, cost per lead by campaign, and month-over-month movement with seasonality noted. If nobody can tell you what happens after the lead arrives, that limit should be admitted rather than glossed over.
- Who owns the accounts and the tracking? You should hold admin access to Google Ads, Analytics, Tag Manager and your Business Profile, and you should be able to remove a partner without losing your history.
A useful test is whether the person will tell you what they cannot measure. An agency claiming full booked-job attribution without a CRM is describing something it cannot deliver.
Where this approach stops
Structural work improves your odds. It does not control the market. Demand moves with the weather, competitor pricing and local supply, and lead quality depends on how fast the phone is answered and how the person answering the phone handles the call, neither of which happens inside the ad account. For most contractors, attribution stops at the lead. No structure guarantees a cost per lead, a booking or a return, and anyone promising otherwise is selling certainty that does not exist.
FAQs
How much do HVAC Google Ads cost per month?
There is no standard monthly figure, because your spend is the number of leads you want multiplied by what a lead costs in your market. Only the second half of that can be looked up. SearchLight's January 2026 benchmark, built on 816 HVAC and plumbing contractors, put the blended cost per lead at $104, with branded search at $34, non-branded search at $149 and Performance Max at $72. EZE Grow publishes a much lower range of $25 to $80. To see the scale, multiply one of those benchmarks by a lead target: 30 leads a month works out to about $3,100 at the blended $104 and closer to $4,500 at the $149 non-branded figure. That is arithmetic on third-party benchmark data from other companies' accounts, not a forecast and not a recommended budget. Plan against your own numbers once you have 60 to 90 days of them.
What is a reasonable monthly budget for an HVAC company?
There is no single correct figure, and any number quoted without your market and your tracking is a guess. Six things shape the answer:
- What you can afford to spend before booked work starts paying it back.
- Conversion volume. A campaign that gathers very few conversions gives bidding little to learn from, while a larger budget can support several service lines at once.
- Conversion quality. A cheap lead that never books is not cheaper than an expensive one that does.
- Campaign structure. One combined campaign is cheaper to fund; separate service lines buy control and cost more to run.
- Seasonality. A heating budget in January and a maintenance budget in May are not the same number.
- Operational capacity. If the crews are booked three weeks out, more leads will not help.
Review the figure each season rather than each month, and treat any published minimum as someone else's starting point rather than a threshold you have to clear.
Why do HVAC clicks cost so much?
Because you are bidding against other contractors for a search that often leads to a job worth several thousand dollars. Demand concentrates in the weeks when systems fail, so the same click costs more in a July heat wave than it does in April. National top-of-page estimates for related marketing terms sat between roughly $33 and $54 in November 2026, and peak-season auctions in competitive metros run higher. The click price itself is a poor way to judge any of this. Cost per lead is the next step up, and cost per booked job is the one that decides whether the campaign is working. A $40 click that turns into a booked replacement is cheap. A $12 click that never reaches a dispatcher is expensive no matter how good the auction looked.
Should I pay an agency a percentage of ad spend or a flat fee?
Either can work, but the fee model matters less than what gets written down. A percentage of spend gives the agency a reason to spend more; a flat fee gives it a reason to do less work in a slow month. Neither is dishonest by itself. What protects you either way is a stated fee, a defined scope, and ownership of the Google Ads account, the tracking numbers, the CRM and the analytics property. Ask what happens to your history if you leave, and settle the reporting shape before the first invoice rather than after it.
Are Google Ads worth it for an HVAC company?
They can be, once three things are true: the account is split by service line, the service area is enforced, and the tracking can tell one stage from the next. Those stages are not interchangeable. A click is a visit you paid for. A lead is someone who contacted you. A qualified lead is one that meets criteria you wrote down. A booked job is an appointment on the schedule, and a sold job is revenue. If your reporting stops at clicks or forms, you are guessing about everything downstream. With reliable booking data, the test is straightforward: compare cost per booked job with what a booked job is worth to your business. If you cannot measure bookings yet, that is the work to do first, not a reason to raise the budget. No structure or budget guarantees a result.
Do I need Local Services Ads as well as search ads?
Not necessarily, and the two do different jobs. Local Services, now moving into Google Ads as Performance Max campaigns with pay-per-lead goals, targets people searching by service category in your service area and charges per valid lead. Search campaigns let you choose the services, terms and landing pages you appear for, and charge per click. Plenty of contractors eventually run both, but the sensible order is to get the phone answered reliably and the tracking clean first. Budget them separately, because a cost per lead and a cost per click cannot be compared, and blending the two hides which one is working.
About this guide
Written by Khandokar Riajul Islam, a performance marketing specialist with more than five years of experience in paid search, paid social and analytics. No HVAC client case study is published, and none is implied here. All benchmark figures are third-party, dated and attributed, and the operational guidance is general paid-media practice applied to residential HVAC rather than a report of campaign results.