Table of contents
In heating and cooling, the conversion is a phone call and the revenue is a booked job three days later — which is why most HVAC ad accounts are optimising toward the wrong number. In one 130,175-call study, 74.1% of calls went unanswered at some point, and booking rates ranged from 31% to 77%. Here is the 2026 attribution data.
Key Takeaways
- 74.1% of home-service calls went unanswered in a 130,175-call study; booking rates span 31%-77%.
- 38% of answered home-service calls are genuine leads; 45% of those convert on the call.
- Call tracking cuts CPL ~10% and lifts lead-to-close ~7%.
- Enhanced conversions lift reported conversions 5%-15%; GCLID imports attribute 20%-30% more.
- HVAC LSA cost per lead was $51 in February 2026 across 409 accounts, rising toward $63 by May.
- LSA at $53 average is ~49% cheaper than blended Google Ads at $104.
- 67% still use last-touch; multi-touch adoption is 41%-47% with adopters reporting +18% ROI, +22% lead quality, -15% CAC.
- 7-12 touchpoints precede a contractor booking; only 21% of marketers trust their attribution.
1. The Phone Is the Conversion — and the Leak
Analysis of tens of millions of tracked calls and booking-rate research describe a funnel that ends at a human, not a form.
| Call metric | Figure | What it means for spend |
|---|---|---|
| Calls unanswered at some point | 74.1% | Media buys demand you never receive |
| Answered calls that are real leads | 38% | The rest is spam, vendors, existing customers |
| Leads converting on the call itself | 45% | The call is the close, not the handoff |
| Booking rate range across representatives | 31%-77% | 2.5x variance on identical traffic |
| Touchpoints before booking | 7-12 | Last-click understates every upper-funnel channel |
| Marketers confident in their attribution | 21% | Most decisions rest on bad data |
Sit with the booking-rate spread for a moment. A contractor spending $20,000 a month at a 35% booking rate and one spending the same at 70% have identical media performance and a 2x difference in cost per booked job. No bidding change, keyword expansion or landing page test comes close to that magnitude — which is why call recording belongs in the attribution stack before any additional media is bought.
2. What Call Tracking Measurably Returns
Home-services call data quantifies the direct media effect.
| Effect | Measured change | Mechanism |
|---|---|---|
| Cost per lead | ~-10% | Bidding sees call conversions |
| Lead-to-close rate | ~+7% | Better keyword and source mix |
| Wasted spend identification | Immediate | Keywords producing non-lead calls |
| Call source visibility | Keyword-level with dynamic numbers | Ties revenue to query |
| CSR coaching signal | Recording plus transcript | Fixes the 31%-77% spread |
| Spam and vendor call filtering | Substantial share of volume | Cleans conversion counts |
| After-hours leakage | Visible in hour-of-day reports | Justifies answering service |
Dynamic number insertion at the keyword level is the minimum viable setup: a pool of tracking numbers swapped on the site so each call carries the click that generated it. Without that, a contractor can report call volume but cannot attribute it, and Smart Bidding is left optimising toward form fills that represent a minority of real demand. The same infrastructure underpins the conversion-quality work in our data intelligence practice.
3. Local Services Ads Attribution and Cost
LSA is where HVAC lead economics currently look best, and where attribution is most opaque. Cost-per-lead tracking across 409 accounts and 2026 LSA statistics give the current reading.
| LSA metric | 2026 figure | Note |
|---|---|---|
| HVAC cost per lead, February 2026 | $51 | 409-account sample |
| HVAC cost per lead, May 2026 | ~$63 | Rising through the season |
| Cross-vertical LSA average | $53 | All home-service categories |
| Blended Google Ads CPL comparison | $104 | LSA ~49% cheaper |
| Lead dispute / credit process | Manual, per lead | Unreviewed disputes inflate CPL |
| Attribution granularity | Category and geography only | No keyword-level data |
| Booking responsibility | Contractor phone handling | Same 31%-77% exposure |
Two attribution cautions apply. LSA reports leads, not booked jobs, so a headline $51 cost per lead becomes a very different cost per booked job at a 40% booking rate. And because LSA charges for calls, unreviewed and undisputed junk leads silently inflate the metric — weekly dispute hygiene is an attribution activity, not an admin chore. Comparing LSA against search fairly requires both channels measured at the booked-job level, as covered in our HVAC SEO statistics.
4. Offline Conversion Imports: Closing the Loop
This is the step that separates contractors optimising for leads from those optimising for revenue. Enhanced conversions documentation and 2026 attribution research describe the recovery.
| Measurement layer | Reported uplift | Requirement |
|---|---|---|
| Baseline form and call tracking | Reference | Tag plus dynamic numbers |
| Enhanced conversions | +5%-15% conversions | Hashed first-party data |
| GCLID offline import | +20%-30% attributed | GCLID stored in CRM |
| Value-based import (booked revenue) | Changes bidding priorities | Job value written back |
| Consent Mode v2 modelling | Recovers consent-limited data | CMP configured correctly |
| Server-side tagging | Improves match durability | Container plus engineering time |
The practical sequence for a contractor is unglamorous and effective: store the GCLID on every web lead, pass it into the field-service system, write back the booked status and the invoiced amount, then import both as offline conversions with values. Once Smart Bidding sees that a $12,000 replacement came from a specific query, it stops chasing cheap maintenance-call clicks. That single change reorders keyword priorities more decisively than any bid adjustment.
5. Where the Revenue Actually Sits
Attribution decisions should follow the money, and the money in HVAC has been shifting. Field-service platform data shows the mix moving.
| Revenue signal | Change | Attribution consequence |
|---|---|---|
| Repair share of revenue | 21.6% → 31.3% | More transactions, smaller tickets |
| Average repair ticket | $818 → $1,205 | Value-based bidding becomes viable |
| Replacement tickets | Five figures, longer cycle | Needs 30-90 day attribution windows |
| Maintenance agreements | Recurring annuity | Lifetime value, not per-lead value |
| Seasonal demand swings | Extreme in both directions | Compare year over year, never month over month |
| Emergency vs planned mix | Emergency skews to phone | Call attribution is mandatory |
Rising repair revenue and a higher average ticket make value-based bidding practical for companies that previously had too little conversion value to work with. It also raises the cost of ignoring maintenance-agreement lifetime value: a $199 plan that renews for six years and generates replacement work is worth far more than its first-year value, and importing only the first invoice teaches the algorithm to undervalue exactly the leads that build the annuity.
6. Multi-Touch: When It Earns Its Complexity
Attribution adoption research and home-service multi-touch analysis set the benchmarks.
| Attribution maturity | Adoption / effect | Fit |
|---|---|---|
| Last-touch only | 67% of businesses | Under $10k/month spend |
| Multi-touch adoption | 41%-47% | Above $10k/month, multi-channel |
| Marketing ROI lift for adopters | +18% | Reallocation gains |
| Lead quality improvement | +22% | Better source weighting |
| Customer acquisition cost change | -15% | Waste elimination |
| Confidence in attribution accuracy | 21% | Industry-wide problem |
| Typical path length | 7-12 touchpoints | Justifies mid-funnel credit |
The honest guidance for contractors is staged. Below roughly $10,000 in monthly media, call tracking plus offline imports capture nearly all the value and multi-touch adds cost without changing decisions. Above that — especially with LSA, search, Meta and organic all running — last touch systematically over-credits branded search and under-credits everything that created the branded query in the first place, which is the exact distortion described in our HVAC branding statistics.
7. Cost per Lead Benchmarks to Measure Against
2026 HVAC cost-per-lead benchmarks give the reference points attribution work should be judged against.
| Channel | 2026 cost per lead | Attribution difficulty |
|---|---|---|
| Branded search | $30-$40 | Low — but credit is often misassigned |
| Local Services Ads | $51-$63 | Medium — no keyword data |
| Performance Max | ~$72 | High — limited placement visibility |
| Blended paid search | $104 | Low with call tracking |
| Non-branded search | ~$149 | Low |
| Paid social | $24-$45 | Medium — view-through inflation |
| Organic and referral | Effectively fixed cost | High — needs assisted-conversion view |
Note how wide the branded-to-non-branded gap is: $30-$40 against roughly $149. Under last-touch measurement, branded search looks like the best channel in the account when it is largely harvesting demand created elsewhere. Any reallocation decision made on that view will cut the channels that generate the branded volume and then wonder why cheap branded leads dried up.
8. The Contractor Attribution Stack
| Layer | Tool category | Priority |
|---|---|---|
| Call tracking with DNI | Call intelligence platform | First — non-negotiable |
| Call recording and scoring | Same platform, coaching workflow | First |
| GCLID capture on web forms | Hidden field plus CRM mapping | Second |
| Field-service system write-back | FSM to ads integration | Second |
| Enhanced conversions | Ads tagging plus consent | Third |
| Value-based offline imports | Scheduled CRM export | Third |
| LSA lead dispute routine | Weekly manual review | Ongoing |
| Multi-touch reporting | BI or attribution platform | Only above $10k/month |
Built in that order, each layer pays for the next. Call tracking alone typically returns its cost within a month through the ~10% CPL reduction and the booking-rate diagnosis; offline imports then unlock value-based bidding; multi-touch reporting is the last mile, and only for accounts large enough for reallocation decisions to matter in dollars. Sequencing this for contractors is part of our growth marketing engagements.
9. Reporting Cadence That Survives Seasonality
| Report | Frequency | Primary metric |
|---|---|---|
| Call answer rate and booking rate | Weekly | % booked of qualified calls |
| Cost per booked job by channel | Weekly | Not cost per lead |
| LSA dispute log | Weekly | Credited leads recovered |
| Offline import health check | Weekly | % of leads with GCLID matched |
| Channel mix vs year-ago | Monthly | Never month over month |
| Revenue per lead by source | Monthly | Invoiced value |
| Maintenance-plan attach by source | Quarterly | Lifetime value proxy |
| Attribution model comparison | Quarterly | Last touch vs data-driven |
Seasonality is the reason so many HVAC reports mislead. A July-to-October comparison in a cooling-dominant market will show collapsing performance that has nothing to do with the media. Year-over-year comparison at the same point in the season is the only honest read, and the dashboards that enforce it are covered in our HVAC landing page statistics companion work on conversion measurement.
10. The Bottom Line
- Fix the phone first. 74.1% unanswered and a 31%-77% booking spread dwarf every media lever.
- Call tracking is the floor. ~10% lower CPL, ~7% better lead-to-close.
- Only 38% of answered calls are leads — clean your conversion counts.
- Import offline conversions. +5-15% with enhanced conversions, +20-30% with GCLID.
- Optimise to booked revenue, not form fills — average repair tickets now $1,205.
- LSA at $51-$63 CPL is cheap, but it reports leads, not jobs.
- Last touch flatters branded search at $30-$40 against $149 non-branded.
- Compare year over year. Seasonality invalidates month-over-month reads.
Attribution in heating and cooling is less about models and more about plumbing: connect the phone system, the CRM and the field-service platform back to the click, then optimise toward booked revenue instead of leads. Contractors who do only that outperform competitors running sophisticated dashboards on incomplete data. If you want that loop built and reconciled against your own invoices, talk to our team.
Frequently Asked Questions
Why is attribution harder for HVAC than for e-commerce?
Because the conversion happens on the phone and the revenue happens days later in a field-service system. A homeowner searches, calls, a customer service representative books or fails to book, a technician diagnoses, and a replacement quote may close weeks afterwards. None of that is visible to an ad platform unless call tracking and offline conversion imports connect the phone system and the CRM back to the click. Around 7 to 12 touchpoints typically precede a contractor booking.
What does call tracking actually change?
Measured effects are modest per metric and large in aggregate: roughly a 10% reduction in cost per lead and about a 7% improvement in lead-to-close rate once call data is fed back into bidding. The bigger benefit is diagnostic. Call recordings expose that a large share of lost jobs are missed or mishandled calls rather than bad traffic, and in home services roughly 38% of answered calls are genuine leads with about 45% of those converting on the call itself.
How many HVAC leads are lost on the phone?
A study of 130,175 home-service calls found 74.1% went unanswered at some point in the customer journey, and observed booking rates ranged from 31% to 77% depending on the representative and the script. That spread is the single largest uncontrolled variable in a contractor's funnel — a company at 35% booking is paying twice as much per booked job as an identical company at 70% with the same media spend.
Do offline conversion imports improve Google Ads performance?
Yes, materially. Enhanced conversions typically lift reported conversions by 5% to 15% by recovering matches that cookie-based measurement misses, and importing booked-job or won-revenue values via GCLID commonly attributes 20% to 30% more conversions than form-only tracking. The strategic gain is that Smart Bidding optimises toward booked revenue rather than raw form fills, which changes which keywords and audiences it favours.
Is multi-touch attribution worth it for a contractor?
For companies spending under roughly $10,000 a month, call tracking plus offline conversion imports capture most of the available value. Above that, multi-touch becomes defensible: adoption sits at 41% to 47% while 67% of businesses still rely on last touch, and adopters report about 18% higher marketing ROI, 22% better lead quality and 15% lower customer acquisition cost. Only 21% of B2B marketers say they are confident in their attribution, so the bar is low.
Sources
Invoca — Insights from 60M+ Phone Conversations
CallRail — Home Services Marketing Statistics
NextPhone — Call Booking Conversion Rate Optimization
Searchlight Digital — Local Services Ads Cost per Lead
Elev8 Operations — Local Services Ads Statistics 2026
Digital Applied — Marketing Attribution Statistics 2026
Diggrowth — Marketing Attribution Statistics
PipelineOn — Multi-Touch Attribution for Home Services
Housecall Pro — HVAC Industry Trends
Elevarus — HVAC Cost per Lead Benchmarks 2026
Google Ads Help — Enhanced Conversions
