HVAC Marketing Benchmarks 2026: CPL, CPC, and Close Rates

"Lead" is the term doing the damage. SearchLight Digital counts a tracked call or form fill inside a managed Google Ads account. First Page Sage defines a lead as "a direct connection via e-mail, phone or in-person introduction to a prospective customer interested in purchasing a product or service." 99 Calls counts "tracked conversions, including phone calls and contact form submissions." Local Services Ads charge per lead by Google's own definition, not per click. Four definitions, four denominators, one shared label.

Read the sample column before the dollar column — not because sampled figures come out cheaper, but because they are the only ones you can interrogate. A published sample predicts nothing about the size of the number: the six HVAC service-category rows printed above all carry an account count and a spend total, and they run from $86 to $354, four of them above the $149 non-branded average and two above $200. What the sample buys is the ability to see that the $354 rests on 25 accounts and $49,000 of January spend while the $198 rests on 432 accounts and $2.74M, and to weight them accordingly. It also lets a publisher flag its own weak row, as SearchLight does with January AC install, which "is either an oversight, or the campaign is not properly named because it is truly catching Furnace installs." The figures published with no account, campaign or lead count — First Page Sage's $92 and $115, 99 Calls' $223 to $353, and WebFX's $153, whose only published sample counts keywords — give you nothing to weight. That is not proof they are wrong; it is the reason they cannot be checked, and the reason they cannot be averaged with the rest.

Averaging this table would produce a number that appears in no source. These are not twenty-one estimates of one quantity. They measure different channels, in different months, under different definitions of a lead, across different client populations. LocaliQ's window closed in March 2025 and publishes medians rather than means; First Page Sage's window closed in June 2025 under a 2026 title.

Channel changes the number more than any other variable, and it is the one variable you control — but the two channel figures cannot simply be divided into each other. SearchLight measured HVAC Local Services Ads at $51 a lead in February 2026 and non-branded Google Ads at $149 in January 2026, then warned on the LSA page itself: "This is why LSA CPL and Google Ads CPL are not directly comparable without context." The denominator is why. On Local Services Ads, "Google sets the lead price" and charges per contact by Google's definition; on Google Ads, SearchLight's own tracking defines and counts the lead. The populations differ too — 888 home services contractors in the LSA report, largest category General / All Trades at 464 accounts, against 816 HVAC and plumbing contractors in the Google Ads report — and the $51 is HVAC-specific where the $149 is not. What survives is a direction, not a ratio: across two consecutive months in one agency's accounts, the LSA lead was cheaper and booked at a higher rate, 44.0% against 37.6%.

Two further provenance facts belong in any HVAC benchmark table.

A widely-cited $45 HVAC cost per lead misreports its own named source. pipelineon.com publishes "Average HVAC cost per lead was roughly $45," along with "Plumbing CPL averaged around $52, electrical around $58, and roofing around $79", and hyperlinks each figure to LocaliQ's home-services benchmark. We fetched both pages again on 18 August 2026. LocaliQ's Air Conditioning Installation & Repair row reads $127.74, plumbing $129.02, electricians $93.69 and roofing $228.15. The lowest cost per lead among LocaliQ's sixteen home-services subcategories is Pools & Spas at $45.15 — which is the only $45 on the page.

The largest public lead marketplace does not publish a dollar figure at all. Angi Inc. discloses Revenue per Lead to the SEC only as a rate of change: "Revenue per Lead increased 11% year-over-year in Q3 2025 (vs. 5% in Q2 2025)." No absolute dollars-per-lead figure appears in the filing. Any article claiming to know the marketplace price of an HVAC lead is not reading the filings.

[VENDOR — DataForSEO] Google Ads auction estimates, United States, language en, pulled 2026-08-17. These are advertiser bid estimates, not amounts any HVAC contractor was observed to pay, and they are not a cost per lead.

Three methodology notes travel with that table, and none of them appears in the benchmark articles that quote a single HVAC CPC.

1. Competition is not a price signal. "air conditioning repair" carries a LOW competition label at index 19 and a $42.73 average cost per click. "new ac unit cost" carries HIGH at index 97 and $9.13. The label counts advertisers bidding, not what the auction charges. Treating it as a cost proxy inverts the answer.
2. Google groups close variants, so a naive average double-counts. Our pull requested 40 keywords and returned 37 distinct data rows. "ac installation", "air conditioner installation" and "air conditioning installation" all returned byte-identical figures ($24.74 CPC, $4.72–$46.21 bids, 60,500 volume), and "ac replacement" and "ac unit replacement" shared another. Averaging the 40 rows triple-counts one keyword.
3. August is not a neutral month. The same pull records "furnace repair" at 450,000 monthly searches in January 2026 and 60,500 in June 2026, and "emergency ac repair" at 5,400 in February 2026 and 60,500 in July 2026. Every AC figure above sits near its seasonal maximum and every heating figure near its minimum.

Observed click costs from managed accounts land inside that auction range, which is a useful sanity check and nothing more. 99 Calls publishes per-service cost per click from its own campaigns between January 2024 and June 2026: $8.99 for ductless mini splits, $33.51 for HVAC general, $43.15 for boiler repair, $56.71 for AC repair. LocaliQ's median for Air Conditioning Installation & Repair is $9.68 and for Heating & Furnaces $9.30 — several times below the auction estimates for the same services, because a median campaign's blended click cost and a top-of-page bid estimate are different quantities.

99 Calls also publishes the most instructive click-to-lead pairing in the dataset: AC repair converted clicks to leads at 19.6% and AC installation at 8.7%, while installation clicks cost less than repair clicks ($33.42 against $56.71). The $353 installation cost per lead is a conversion-rate problem, not a click-price problem.

That survey found contractors closing 43% of install jobs — 45% residential, 38% commercial. Separately, it puts the average service ticket at $422, split $390 residential and $516 commercial, as reported by The ACHR NEWS in February 2026. Those two figures describe different jobs and should not be multiplied together. Both reach this page secondhand: the full study is gated to ACCA members, so what is quoted here comes from ACCA's public recap and The ACHR NEWS's report of the survey rather than from the study document itself. Its actionable findings are relative rather than absolute, which is what makes them usable: close rates run 52% when a proposal carries four or more options against 42% at one to three options, and only 10% of contractors offer four or more. Close rates run 38% without financing and 49% with it, while 68% of contractors have financing available and 37% offer it on every job.

TJ O'Connor, president of Farmington Consulting Group, put the options finding plainly when presenting the results: "If you don't offer something to a customer, you have zero chance of making that sale."

Booking rate — the share of inbound calls that become jobs — has exactly one sourced HVAC number, and it is nearly four years old. ServiceTitan published aggregated data from more than 3,000 trade businesses in the US and Canada for June 2022: a typical shop booked 42% of calls across all trades, HVAC specifically booked 38%, businesses with 25 or more techs booked 59% and businesses with fewer than five booked 24%. Two cautions attach. June is summer, and the same article records that "HVAC booking rates varied with the seasons, with higher booking rates in shoulder seasons and lower booking rates in summer." And the 90% figure that circulates as an elite benchmark is not measured: three named ServiceTitan advisors "all feel that a 90% call booking rate is attainable." The highest measured segment in that dataset is 59%.

Angie Snow, a ServiceTitan principal industry advisor quoted in the same report, drew the line between this metric and the marketing budget: "But what kind of training are they putting into their CSRs? The money that they're not investing there, they're wasting on marketing dollars. Because if you get a good CSR, you don't have to spend as much in marketing because your booking rate's up and the calls are coming in."

Average ticket is where benchmark articles do the most damage, because four different quantities share the name.

Housecall Pro's series is the only measured HVAC ticket trend in the set: average repair revenue per job rose "from $818 in 2021 to $1,205 in 2025 (a 47% nominal increase, or 24% after inflation)," and repair revenue climbed from 21.6% of HVAC revenue in 2021 to 31.3% in 2025 on a full-year basis. Housecall Pro also states the limit of its own data, which almost no republisher carries forward: "Findings reflect directional patterns within this subset and should not be interpreted as representative of all HVAC activity on the platform or as benchmarks for individual businesses."

The ServiceTitan row deserves its bold type. That page publishes targets, framed in its own words as "a list of ones many best practice contractors start with," with no sample, no dataset and no date — including the 25–35% close rate on a marketing lead that gets laundered into other articles as the industry close rate, and a $7,700 install target whose own footnote reads "This figure fluctuates highly per market area. Adjust accordingly."

Read the methodology block before the benchmark table, in other words, and read all of it. The same page is the origin of the $153 "industry average" cost per lead, and SearchLight's own comparison table credits it accurately as "WebFX (2025) — $153 avg CPL." A number can be correctly attributed and still be evidence-free, and both of these are.

Two figures on that page cannot describe the same industry. WebFX publishes a lead-to-customer conversion rate of 3.10%. SearchLight, working from its own measured book rate (the share of leads that become a booked appointment) and match rate (the share attributed to a paying customer in the CRM), states on its page: "If your book rate is 38% and your match rate is 42%, roughly 16% of your leads become paying customers." Roughly 16% and 3.10% are not two estimates of one quantity; at least one of them is not measuring lead-to-customer conversion at all.

Lifetime value is also the wrong number to import. It is the most company-specific figure in HVAC — set by your maintenance-agreement attach rate, your repeat rate, your replacement close rate and your ticket — so a national average has no bearing on what you can pay for a lead. Build it from inputs you own, using published figures only as sanity checks:

The arithmetic is simple and the inputs are the hard part, which is why we built a tool for it: run your own lifetime-value numbers rather than adopting anyone's average. What the exercise gives you is the only lead-price ceiling that means anything — the one derived from your own margins. SearchLight demonstrates the method on its own dataset and reaches a specific, bounded conclusion: "you can afford a CPL of up to about $100 before first-job acquisition becomes unprofitable." That number is theirs, computed from a $2,500 ticket at 25% margin. Yours will differ. The method transfers; the figure does not.

Track these four, using SearchLight's definitions so your numbers stay comparable to the largest documented dataset: cost per lead (spend divided by leads), book rate (the percentage of leads that result in a booked appointment), match rate (the percentage of leads attributed to a paying customer in your CRM) and average ticket (revenue per paying customer). Cost per booked job falls out of the first two, and it is the number that decides whether a channel is working.

1. Fix the definition before the arithmetic. Decide what counts as a lead — every call, or every call over 30 seconds from a new number, or every booked appointment — and write it down. Half the disagreement between published benchmarks is definitional, and the same is true inside a single shop across two quarters.
2. Measure by channel, never blended. SearchLight's own data makes the case: across its pooled HVAC-and-plumbing dataset, branded search at $34 a lead and non-branded at $149 sit inside one $104 blended average. A blended number cannot tell you which channel to cut — and, as those three figures show, it cannot tell you which trade it is describing either.
3. Compare month against like month. A January heating benchmark tells a Phoenix contractor little about July. Our own auction pull records "furnace repair" moving between 450,000 and 60,500 monthly searches inside twelve months.
4. Grade every external figure before using it. Ask four questions of it: who published it and what do they sell; what sample is published; what period does it measure; is it a measurement or a recommended target. A figure that fails any of the four is directional at best.
5. Re-benchmark on your own data quarterly, and on published data annually. Most of the "2026" benchmarks fetched for this page measure 2024 and 2025.

Cost per lead is the wrong place to look first when the book rate is the weak link. Work the arithmetic on your own two numbers: a $51 lead booked 20% of the time costs $255 per booked job, while a $149 lead booked 60% of the time costs roughly $248 — the expensive lead wins. The fix in the first case is a phone process, not a media plan. That is the practical content of Angie Snow's point about CSR training above, and it is why an audit should read call handling before it reads bid strategy.

Two questions this page deliberately leaves to other pages. How much of revenue to allocate to marketing is a budgeting decision with its own evidence base: ACCA's survey of 1,000-plus contractors found that the average contractor spends 6% of annual revenue on marketing and advertising, and contractors investing at least 12% reported net profits of 9% against 5% — a correlation observed in a survey, not a proven return on the extra spend. And what HVAC SEO retainers actually cost is a market-price question answered from published rate cards rather than benchmark datasets; how the two channels' economics compare sets the paid and organic figures side by side.

We publish our rates before any call for the same reason this page publishes sample sizes: a number you cannot check is not information. Our own client cost-per-lead and book-rate figures are absent from the table above because HVAC Foundry has no published dataset of its own yet.

For scale, from the one publisher here that sells neither leads nor marketing services: IBISWorld publishes the US Heating & Air-Conditioning Contractors industry at $159.4bn in 2026, across roughly 120k businesses, growing at a 2.6% CAGR between 2021 and 2026. That is a contractor-services industry, not the global equipment-market figures that circulate in HVAC marketing posts.