Google ads for HVAC: what works and what wastes money

Google ads in HVAC are a faucet connected to the weather. Open it during a heat wave and money comes out — homeowners in hot houses click the first credible result and book same-day. Open it wrong, and HVAC’s click prices ($15–50 for emergency terms) will drain a budget faster than almost any trade.

The difference is never enthusiasm; it’s structure, timing, and follow-up. This guide covers the two Google channels that matter — Local Services Ads and Search — plus the weather-trigger routine that makes HVAC ad budgets uniquely time-able. (Where paid fits among all your lead sources is in the ranked list.)

Start with Local Services Ads (Google Guaranteed)

The green-check listings above everything else. For HVAC they should be the first paid dollar, for three reasons:

• Pay per lead, not per click — a browser costs nothing; only contacts bill, and junk leads can be disputed for credit.
• The badge does trust work — Google’s screening (license, insurance, background checks) is visible exactly when a stranger is deciding whom to let into the house.
• Your reviews are the bidding power — LSA ranking leans heavily on review count, rating, and responsiveness. The review engine you’re building for the map pack pays here twice.

Setup: apply at Google’s Local Services portal, pass screening (have license and insurance docs ready — it takes days to weeks, so start now), connect your profile reviews, set service types (AC repair, heating repair, installation, maintenance) and towns, set a weekly budget.

Operating rules: answer instantly — LSA ranks responsive companies higher and bills either way; dispute wrong-service and out-of-area leads every time; and never let the budget cap out mid-heat-wave — the weeks you most want leads are the weeks a low cap quietly turns you off.

Search campaigns: structure before spend

Search ads (the classic pay-per-click results) add precision LSAs can’t offer — specific terms, specific pages, specific seasons. Build the account as four campaigns so each gets its own budget dial:

Campaign 1: Emergency cooling (summer priority)

• “AC repair [city],” “air conditioning repair near me,” “AC not cooling,” “24 hour AC repair”
• Lands on: the AC repair page (emergency strip up top)

Campaign 2: Emergency heating (winter priority)

• “furnace repair [city],” “heater not working,” “emergency furnace repair,” “no heat”
• Lands on: the furnace repair page

Campaign 3: Replacement (steady, research-phase)

• “AC replacement cost,” “new HVAC system [city],” “HVAC financing,” “new furnace cost”
• Lands on: the installation/replacement page with financing framing

Campaign 4: Maintenance & shoulder work (small, seasonal)

• “AC tune up [city],” “furnace tune up,” “duct cleaning [city]”
• Lands on: maintenance-plan page

Why the split matters: emergency campaigns deserve top bids and near-24/7 schedules in their season and near-zero out of it; replacement runs steady year-round; tune-up terms are cheap and feed the membership machine. One blended campaign can’t do seasonal budget shifts — four dials can.

The negative-keyword list: HVAC’s money-saver

HVAC search traffic is polluted with searches you can never book. Load these negatives on day one:

how to, DIY, fix yourself, capacitor, contactor (part-shoppers), filter, filter size, 16x25x1, freon cost, recharge kit, window unit, portable AC, space heater, Home Depot, Lowe’s, Amazon, parts, wholesale, salary, jobs, hiring, school, certification, training, reddit

Then check the search-terms report weekly and keep feeding the list. In most HVAC accounts a disciplined negative list rescues 20–30% of spend — homeowners searching “AC capacitor replacement cost” are fixing it themselves this weekend, and their click costs the same $30 as a real customer’s.

Ad copy: attract the booker, repel the browser

Every click is expensive, so the ad’s job is pre-qualification, not raw clicks:

• Headline formula: service + place + speed. “AC Repair in Plano — Same-Day Service” / “Furnace Out? We’re Open Now” / “New HVAC From $89/Month.”
• Put the service fee in when you can. “$89 Diagnostic, Applied to Repair” filters bargain-hunters before they spend your money.
• Descriptions carry trust: licensed & insured, NATE-certified, review count, “techs in your area today.”
• Use every extension: call button (crucial — many emergency searchers never visit the site at all), location, sitelinks to service pages, review ratings. Extensions are free ad real estate.

Two or three ads per ad group, rotate, replace the loser monthly.

The weather trigger: HVAC’s unfair advantage

No other trade gets to schedule its demand spikes. You can see the heat wave five days out. Build the routine:

1. Watch the 7-day forecast — the first stretch above ~92° or below ~25° is your signal.
2. Two–three days ahead: raise the relevant emergency campaign’s budget 2–4×, extend ad schedules into the evening, confirm LSA caps are lifted.
3. During: watch speed-to-lead like a hawk — the leads you’re buying are perishable within the hour.
4. After: budgets back down, and mine the surge — every emergency customer gets the review ask and the membership pitch while the relief is fresh.

Companies that budget by calendar month miss this entirely; the demand doesn’t arrive monthly, it arrives with fronts. Your ad spend should look like the weather map.

An example month of budgets (so the dials feel real)

Say you’re working with $4,000/month in a mid-size metro. Here’s what the four-dial structure looks like in two different months:

July (cooling peak):

• LSA: $1,800 — uncapped during heat waves, the workhorse
• Emergency cooling: $1,400 — top bids, schedule extended to 10pm
• Replacement: $600 — steady; July emergencies seed August replacements
• Emergency heating + tune-ups: $0–200 — paused or trickle

October (shoulder):

• LSA: $800 — demand is thin; let it idle
• Emergency heating: $600 — building toward the first freeze
• Replacement: $1,200 — “replace before winter” is the season’s message
• Tune-ups: $800 — cheap clicks, feeds the membership machine
• Reserve: $600 held for the first cold snap

The percentages aren’t gospel — your market’s weather and competition set them. The shape is the point: money follows the season, every campaign has its own dial, and a reserve exists so the first forecast extreme gets funded in hours, not after next month’s budget meeting.

The five ways HVAC companies waste money on Google

The same leaks, account after account:

1. One blended campaign — can’t shift budget seasonally, so July dollars subsidize furnace keywords nobody’s searching.
2. No negative keywords — quietly funding part-shoppers, filter-size searches, and DIYers. The list above rescues 20–30% of spend in ten minutes.
3. Ads running when nobody answers — a 9pm click to voicemail is a $35 donation. Match ad schedules to answering capacity, or staff the phones.
4. Everything lands on the homepage — the emergency ad lands on the AC page, the financing ad on the replacement page, always. This one discipline commonly halves cost per lead.
5. Judging on clicks or leads instead of booked jobs — the tune-up campaign’s cheap leads and the emergency campaign’s dear ones can’t be compared any other way.

What about Performance Max and “smart” campaigns?

Google will push automation hard. The honest read for a local HVAC company: automation learns from conversion volume, and a shop with 40 leads a month doesn’t feed it enough — the algorithm’s guesses cost real money, and Performance Max hides the search-terms detail that keeps HVAC accounts clean. Earn automation: run the structured campaigns above until months of clean conversion data exist, then test automated bidding inside that structure against a known baseline. Full autopilot on a fresh account is how shops spend $4,000 learning what a negative-keyword list would have prevented.

LSA vs. Search: which does what

Owners often ask why they need both. They do different jobs:

	Local Services Ads	Search ads
You pay for	a lead (call/message)	a click
Placement	above everything	below LSAs, above organic
Trust signal	Google Guaranteed badge	your ad copy
Targeting control	service types + towns	exact keywords, schedules, pages
Landing page	none — they call you	yours (which must convert)
Ranking lever	reviews + responsiveness	bids + relevance
Best at	emergency capture, steady volume	seasonal precision, replacement terms, symptom searches
Weak at	replacement research, financing messaging	cheap trust with cold callers

The practical split: LSA is your always-on emergency net; Search is your precision seasonal tool. LSA can’t show a financing message to a replacement researcher — Search can. Search can’t borrow Google’s badge for a panicked 9pm caller — LSA can. Most HVAC accounts settle around 40–60% of paid budget in LSA, with Search flexing around the weather — and if budget only allows one to start, it’s LSA every time, because pay-per-lead with a dispute process is the gentler place to learn.

The weather-trigger checklist (tape it to the wall)

When the 7-day forecast shows the first stretch above ~92° or below ~25°:

1. Raise the seasonal emergency campaign budget 2–4×
2. Lift the LSA weekly cap
3. Extend ad schedules to match extended phone coverage
4. Confirm missed-call text-back is on and someone owns speed-to-lead
5. Swap in the seasonal ad variants (“Heat wave this week — same-day AC repair”)
6. After the front passes: budgets back down, review asks out, membership pitches to every emergency customer served

Six lines, maybe fifteen minutes — and it’s the difference between buying the season’s cheapest booked jobs and its most expensive ones.

Tracking: three numbers, monthly

1. Call tracking on ad traffic so calls attribute to campaigns.
2. Form and booking tracking, tagged by source.
3. Cost per booked job, by campaign. Not clicks, not leads — jobs on the schedule. Emergency campaigns will look expensive per click and cheap per job; tune-up campaigns the reverse. The booked-job sheet is the only place the truth lives, and it’s the same funnel math every channel answers to.

Your first month, week by week

Week 1 — plumbing. LSA application in (it takes time). Call + form tracking installed. Four-campaign structure built with negatives loaded, everything paused. Landing pages pass the phone test — the website anatomy is the checklist.

Week 2 — LSA live. Budget on, instant-answer rule assigned by name, dispute habit taught.

Week 3 — seasonal emergency campaign live. Modest cap. Search-terms report checked twice; negatives fed.

Week 4 — read and extend. Compare LSA vs. search on cost per booked job. Turn on the replacement campaign if crews have install capacity. Book the recurring 30-minute monthly review: search terms, negatives, ad rotation, and the booked-job math — same slot every month, non-negotiable, because the accounts that decay are always the ones nobody looked at since launch.

From then on: 30 minutes a month plus weather-trigger days. Constant fiddling resets Google’s learning — build tight, let it run, spend your attention on answering fast.

One last calibration on expectations: give any campaign 60–90 days and a few thousand dollars of data before passing judgment. HVAC ad performance swings hard with the weather, so a “bad” month is often just a mild one — compare against the same season last year, not last month. The account that looks mediocre in a cool June can be printing money by the second week of a hot July, with identical settings. Patience with the structure, ruthlessness with the leaks.

Before any of it: make sure the destination converts. Every leak in the website — a slow page, a buried phone number, a combined services page — is a tax on every click you buy, which is why the foundation always comes before the faucet. The free audit scores your site and profile in 60 seconds; fixing what it finds makes every ad dollar afterward work harder. And if you’d rather run calls than campaigns, we handle the foundation for a flat monthly price — so the only marketing job left on your desk is answering the phone fast.