HVAC marketing ROI: the 4-step closed-loop attribution chain that makes spend measurable

A thread on the contractor subreddit titled "marketing is just vibes at this point" sat at 345 upvotes and 87 comments when I bookmarked it. The frustration is reasonable. The problem in residential HVAC is not that marketing does not work; the problem is that most shops cannot prove whether it works, and a budget you cannot defend is the first one the owner cuts when the cash month gets ugly.

I run the dispatch-software comparisons at Full Stack HVAC and write the content side of HVAC Know It All. Across operator interviews, podcast guests, and the trade publications I track, the contractors who can speak confidently about their marketing all do the same handful of things: they assign a unique phone number per channel, they ask the CSR to log the source on every call, and they reconcile that source back to revenue at the end of the month. The mechanics are unglamorous. They also turn the vibes question into an arithmetic question.

Why HVAC marketing reads as untrackable

Residential HVAC marketing is genuinely harder to measure than e-commerce, and the reasons are structural:

• Long attribution windows. A homeowner who sees the truck wrap in March may not call until the AC dies in August. By then the touchpoint that started the relationship is invisible.
• Multi-touch journeys. They see the truck, they search online, they read reviews, they ask a neighbor, then they call. Picking the one "responsible" touch is a fiction.
• Offline conversion. The call lives in the phone system, the job lives in the field-service software, the revenue lives in QuickBooks, and the review lives on Google. None of those systems is wired to the others by default.
• Seasonal distortion. Summer AC revenue rises whether the marketing worked or not. Without a baseline, you cannot tell whether the Google Ads campaign moved the needle or the weather did.

The spend benchmark, and the spend-vs-track contradiction

ACHR News reporting in August 2025 puts the HVAC marketing-budget benchmark at 8-12% of revenue.¹ For a $1M residential shop, that is $80K-$120K a year across digital, print, sponsorships, and referral programs. That number is widely repeated and reasonable as a planning baseline.

What gets less press is the contradiction in how the same operators talk about that spend in private. The "spend more" view is that 8-12% is the floor and that anything below it starves the pipeline. The "spend smarter" view is that most shops are wasting half of what they already spend, and that pouring more money on top of an untracked budget just enlarges the leak. Both sides are right at the same time. The reconciliation is that the budget number and the measurement discipline are two separate decisions, and skipping the second one is what makes the first one feel like gambling.

The four numbers that take marketing out of "vibes"

You do not need a marketing degree or an enterprise analytics stack to answer the question every owner is asking. You need four numbers, tracked per channel per month:

1. Lead source. Where did the call originate? Google paid search, Local Service Ads, organic, referral, yard sign, truck wrap, repeat customer. A call-tracking product (CallRail, CallTrackingMetrics) assigns a unique phone number per channel and resolves this automatically.
2. Call outcome. Did the call book an appointment? The CSR or dispatcher logs the result in the CRM or field-service software.
3. Booked-job revenue. What did the ticket close at? The dispatch software already tracks this.
4. Cost per booked job. Total channel spend divided by booked jobs from that channel. Not cost per lead. Cost per lead can flatter a channel that delivers cheap calls that never convert.

If you have those four numbers for each channel each month, the question "is my marketing actually making me money?" has an arithmetic answer.

The channels and what they roughly cost

Pulled from operator interviews, ACHR News reporting, and the practitioner content I track. Treat the cost figures as directional ranges rather than benchmarks; market, season, and account configuration all move them substantially.²

The closed-loop attribution chain

The reason marketing feels like vibes is that almost every contractor stops at lead count. The chain that makes it measurable goes further:

Lead source → phone call → booked job → completed ticket → recognized revenue → post-job review

Each handoff is a real artifact in a real system. The call-tracking product owns the source. The CRM owns the booking. The field-service software owns the ticket and the revenue. The Google Business Profile owns the review. When the four are stitched together for the same customer, the dollar of revenue traces back to a specific channel, and channel-level ROI stops being a guess.

The minimum stack to make that chain operational:

• Call tracking. CallRail or CallTrackingMetrics. Unique numbers per channel, dynamic number insertion on the website, recordings reviewable for source verification.
• CRM or field-service software. ServiceTitan, Housecall Pro, or Jobber. Booked-job and ticket-value attribution lives here.
• A monthly review. Cost per lead, cost per booked job, revenue per channel. Done at calendar-month boundaries so seasonal swings are visible.

Industry consultant Ruth King has written for ACHR News for two decades that knowing true cost per lead by channel is the operating baseline for marketing accountability. The substance of that claim is straightforward: without it, the conversation about marketing is between people who have opinions and people who have receipts, and the receipts win.

What to actually do this month

1. Calculate marketing spend as a percentage of revenue. Compare against the 8-12% benchmark. The point is to know which side of the line you are on, not to hit a number.
2. Stand up call tracking. Unique number per channel. Most call-tracking services land at $50-$150/month for a residential HVAC footprint.
3. Have the CSR log lead source on every call. "How did you hear about us?" answered on the call and recorded in the CRM. Cheap to do, surprisingly rare in practice.
4. Review on a monthly cadence, not a weekly one. Marketing needs a 90-day-plus evaluation window to absorb seasonal noise. Panic-cutting a channel after one slow week is how good campaigns die.
5. Compare channels on cost per booked job, not cost per lead. A $30 lead that never books is more expensive than a $100 lead that converts to a $5,000 install. The channel with the cheapest lead is not the channel with the cheapest customer.

Marketing in residential HVAC is harder to measure than it should be, given how much spend rides on it. It is not, however, unmeasurable. The four numbers above are what separate the operators who can defend their marketing budget from the ones who lose it in the first soft quarter.

For the dispatch-software side of the attribution chain (which CRMs cleanly track booked-job revenue back to lead source), see the dispatch-software comparison post. For the speed-to-lead piece that converts paid calls into booked jobs in the first place, see the 4-hour lead-response window.