1. What the published HVAC cost numbers tell you
LocaliQ's 2025 report analyzes 3,211 U.S. Home-service Search campaigns from April 1, 2024 to March 21, 2025. AC CPL is $127.74, heating/furnaces CPL $129.02; CPCs are $9.68 and $9.30. The page was updated July 2026, but an updated page date does not turn the underlying data into a 2026 sample. The report covers Google and Microsoft Search. A Google-only price has to come from your account.
Your auction price also reflects Quality Score, landing-page relevance, and whether Smart Bidding sees qualified installs or raw call clicks. Two HVAC companies in the same metro can report wildly different CPLs with identical "budgets."
Treat benchmarks as a conversation starter with ownership: "If we need 12 changeouts a month, here is the lead volume the funnel implies."
- → Use for: First-draft media sizing and board-level expectations.
- → Do not use for: Firing a vendor because one week exceeded a blog's average.
- → Replace with: 90-day cost per booked install from your CRM.
2. Seasonal budget logic unique to HVAC
HVAC is one of the few trades where a 10-degree temperature swing can double emergency search volume. Raising bids without on-call techs simply buys voicemail.
Build three postures: heat-wave emergency, cold-snap emergency, and shoulder-season replacement selling. Shoulder season is when changeout CPLs often look "worse" but sales cycles are calmer and attach rates for IAQ or warranties can improve.
Recommendation, and this is editorial judgment: pre-approve heat-wave coverage and technician overtime before raising media spend. A forecasted surge is an operations decision as much as a bidding decision.
| Posture | Bid/budget bias | Ops check |
|---|---|---|
| Heat-wave repair | Higher emergency spend | On-call techs + capacitors stocked |
| Cold-snap heat | Higher heat repair spend | After-hours answering live |
| Shoulder replace | Steady install campaigns | Sales capacity for estimates |
3. Campaign architecture that protects changeout margin
Split search into at least: emergency repair, planned maintenance/tune-up, and replace/install. Add brand defense separately so competitors cannot siphon your name cheaply.
Emergency ads should talk response windows and truck fees. Replace ads should talk load calculation, financing, and system options. Shared copy trains Google, and your team, to send the wrong intent to the wrong page.
Negatives that repeatedly matter for HVAC: DIY parts ("capacitor for sale"), employment, "window unit" if you do not sell portables, and pure informational "SEER rating explained" queries if they never convert in your account.
- ☐ Emergency vs install campaigns with separate budgets
- ☐ Landing pages with different forms (repair vs replace)
- ☐ Call recording graded by job family
- ☐ Weather-aware weekly budget reviews
- ☐ Offline conversion import for sold changeouts
4. Worked example: from 10 changeouts back to budget
Illustrative goal: 10 residential changeouts a month, with a 35% estimate-to-sale close rate and a 50% lead-to-estimate rate. You need about 10 / 0.35 / 0.50 ≈ 57 replace-intent leads, and at a $150 net CPL the media is roughly $8,550 before management, if those assumptions hold.
If half your "replace" budget buys emergency repair clicks because campaigns are blended, you will hit the lead count and miss the install count.
5. HVAC PPC mistakes
The usual misses are one campaign for every HVAC keyword, bidding to any phone call, ignoring the overnight miss rate in a heat wave, and sending "AC repair near me" and "new HVAC system financing" to the same page. Fix the tracking first, then the campaign structure, then the budget.
Pair Search with instant follow-up
AI or live answering turns expensive HVAC clicks into scheduled troubleshoots before a competitor answers.
6. Landing pages and offer logic for HVAC Search
Emergency repair pages should lead with tap-to-call, a short "what is happening" form (no cool / no heat / strange noise), and a visible truck-fee policy if you charge one. Long brand stories belong lower on the page.
Replacement pages should ask for home size, system age, fuel type, and photos of the outdoor unit and the furnace data plate. Those questions take a minute, and they filter tire-kickers before an estimator drives across town.
A maintenance plan page needs the price, how often you visit, and what the plan leaves out. Send maintenance keywords to that page. When those keywords land on a replacement page, people leave, and you still pay the CPC.
Page speed on mobile matters during heat waves when LTE is congested. Compress hero images; keep call buttons sticky.
Test one primary CTA per page. Competing "chat," "form," and "call" equal-weight buttons often reduce emergency call volume.
- ☐ Sticky tap-to-call on emergency LPs
- ☐ Photo fields on replace LPs
- ☐ Separate maintenance LP
- ☐ Mobile LCP checked on 4G throttle
- ☐ CTA hierarchy decided before creative tests
7. Bidding signals: what HVAC should optimize toward
If every call longer than 30 seconds counts as a conversion, Smart Bidding will happily buy tire-kickers who chat about filters. Create conversion actions for qualified estimates and closed installs with different values.
Value rules can weight changeout-intent calls higher than tune-up calls once your CRM can identify them. Until then, keep campaigns separate so budgets cannot cross-contaminate.
Enhanced CPC without clean offline data is a common HVAC trap. It scales spend into whatever the account historically called "success," including junk.
When you test AI Max inside Search, start on the install campaign with clean conversion values. Leave the emergency campaign alone while overnight staffing is thin.
- → Primary conversion: Write it down before you optimize: a booked estimate or a completed install.
- → Secondary: Qualified phone connect with minimum talk time and job-family tag.
- → Exclude: Nav clicks, 5-second calls, known spam patterns.
8. Brand defense and competitor terms
Bid on your own shop name. During a heat wave, competitors will bid on that name when a homeowner searches for the company a neighbor recommended.
Competitor-name bidding is a legal and policy-sensitive area. Follow Google's trademark policies and your attorney's guidance. Many HVAC advertisers prefer brand defense plus strong non-brand emergency coverage instead of aggressive competitor conquest.
Measure brand CPL on its own. That number should look "too good" next to non-brand. When brand CPL looks ordinary beside non-brand, check the search terms that landed on the brand campaign, and check whether the sitelinks still match the page.
9. A monthly scorecard for HVAC ownership
Use separate rows for emergency repair, tune-up and replacement. Record spend, qualified conversations, appointments, completed visits, sold jobs, gross profit and cancellations. A campaign with more calls and fewer completed visits needs a look at the jobs. The extra calls are the wrong thing to celebrate.
Attach the date the lead arrived and the date the sale closed. Otherwise June replacements may be attributed to July spending, making both months look misleading. Report cohorts as well as calendar totals.
Before an experiment, write the success criterion: lower media cost per qualified replacement appointment without reducing sold gross profit. Keep technician capacity and service area consistent when comparing periods. If a heat wave changes demand, write that down next to the result so you can see that the weather moved the numbers.
When volume is small, do not stop a campaign after two expensive days or declare victory after one sale. Set a test spending limit you can afford, review actual call quality, and stop immediately if tracking breaks or dispatch cannot fulfill promises. Set a spending limit you can afford, review call quality, and stop if tracking breaks or dispatch cannot keep the promise.
10. Membership campaigns need their own payback window
A maintenance membership can make sense even when its first visit contributes little margin, Pull the future value from your own records. Use actual renewals, repeat work, and cancellation history from your CRM when you assign a conversion value.
Illustrative decision: a campaign books ten discounted tune-ups, but seven cancel after the first visit. Counting ten annual members overstates acquisition value. Reconcile active contracts and completed visits before increasing spend.
Do not train the same bidding objective to value a cheap tune-up and a sold system replacement equally. Keep membership experiments separate until your data supports a defensible lifetime-value model. This also makes an honest conversation with ownership easier: the test buys retention potential. Immediate replacement revenue is a different result.