The Level Index, by trade
HVAC contractor financial benchmarks
The honest benchmark for HVAC contractors, built from citable public sources. Commercial mechanical margins come from SEC filings, wages from BLS, and market size from industry research. Level layers its blended contractor data (collection, billing, quote conversion) on top, with trade variation unmeasured in these pooled records.
24.1%
Gross margin
a large public commercial HVAC contractor (FIX, FY2025)
$59,810
Median tech wage
HVAC mechanics (BLS OEWS, May 2024)
~90-105 days
DSO
commercial mechanical (SEC filings)
~$159B
US market size
HVAC contractors (IBISWorld)
Productivity answer
What is a scaled public-company revenue-per-employee reference for HVAC?
Approximately $401K per employee
Comfort Systems USA reported $9,101.6 million of FY2025 consolidated revenue and approximately 22,700 employees at year-end. Dividing those filing values produces about $400,951 per employee.
- Definition
- Annual consolidated company revenue divided by approximate year-end headcount. The revenue includes mechanical and electrical operations, and the denominator includes field and office staff. This is not revenue per technician or revenue per truck.
- Evidence and sample
- Tier D Level calculation from Tier A public-company SEC filing inputs. One scaled public commercial mechanical and electrical operator. Use this as a verifiable consolidated-company reference, not as a typical private HVAC contractor target or a pure HVAC measure.
For narrower operating measures, see Level's revenue per technician analysis and revenue per truck benchmarks.
Match the work and accounting definitions
An HVAC margin comparison needs the actual segment definition, not the word service.
Split the word 'service' before you trust any margin. Building-services segments and owner-direct relationship segments use different classifications. EMCOR's mixed building-services segment ran a 6.0% operating margin in FY2025, roughly half its mechanical-construction segment at 12.8%, in the reported segment comparison. That does not isolate the pricing mechanism. Limbach separately reported owner-direct relationships at 31.2% gross margin versus 21.1% for GC relationships in FY2024. This is a customer-relationship split, not a service-versus-construction comparison; it uses gross margin while the EMCOR comparison uses operating margin. The two cannot be read as opposing service results. The retained Comfort Systems activity references do not provide a clean service-versus-construction margin comparison; its FY2025 revenue definition of 'service' spans from 7.3% of revenue (pure maintenance) to 36.8% (all existing-building work). So a quoted 'HVAC margin' tells you almost nothing until you know which work type and which scale it measures.
Read margin by work type and the scale you actually operate at. Owner relationships, facilities contracts and project work are different classification axes; do not treat the reported segments as a controlled pricing experiment.
Source: EMCOR FY2025 10-K, Limbach FY2024 10-K, Comfort Systems FY2025 10-K
How to read this: Level research pools HVAC, plumbing, electrical, mechanical, refrigeration and fire protection. Roofing and general contracting are outside that pool. The sources below are selected for relevance to HVAC contractors; some are proxies or all-contractor references (SEC filings, BLS, CFMA, industry research), each tagged by confidence tier. Level's research below combines those trades. Some older figures have not been checked again against their original sources. Each value explains the time period, businesses included and what was measured.
Different work scopes can produce different margins; these retained ranges are planning references.
Illustrative gross-margin inputs by job type (not a measured cohort)
Source / sample: Retained coaching references; selected scenario inputs, not measured percentiles
Retained planning ranges are service/maintenance 50-58%, equipment install 35-48% and new construction 28-38%. Original source cuts are not reproduced, so these are assumptions for a scope/cost comparison, not verified percentiles or a guaranteed ordering. Compare paid nonbillable time, included parts, equipment, callbacks and subcontract scope once.
Source: CallJolt, Profitability Partners, RunClockwork 2026 (directional)
Customer relationships and business segments answer different questions; compare their definitions first.
EMCOR segment operating margin, FY2025
Source / sample: EMCOR FY2025 10-K segment operating income
EMCOR separately reports its building-services segment (mixed facilities, project and repair work). In FY2025 it ran a 6.0% operating margin, roughly half its mechanical-construction segment at 12.8%, in the reported segment comparison. The aggregate does not isolate pricing or overhead as the cause. Do not read that as 'all service thins at scale.' Limbach's owner-direct relationship segment ran a 31.2% gross margin in FY2024 versus 21.1% for its subcontracted construction (gross, same company). These segment observations do not isolate pricing, bidding or overhead as the cause. Limbach groups customer relationships; EMCOR building services mixes facilities work with retrofits, controls projects and repairs.
The HVAC tech wage runs from about $19 to $44 an hour, and your local market sets your labor cost.
HVAC mechanic hourly wage percentiles (BLS OEWS, May 2024)
Source / sample: BLS OEWS May 2024, SOC 49-9021 (estimated employment 396,870, not survey sample size)
These are employee wages, not your fully-loaded cost. Payroll taxes, workers' compensation, benefits and paid nonbillable time vary by employer and state. Calculate them from payroll and policies, use the same paid-versus-worked-hour basis, and subtract only cost omitted from starting gross profit.
Source: BLS Occupational Employment & Wage Statistics, May 2024, SOC 49-9021
HVAC benchmarks, by source
7 source-verified rows (Tier A to E)Expand all 7 source-verified HVAC benchmark rowsEvery metric is tagged by confidence level, from Tier A through Tier E, with direct SEC filing, BLS wage, and CFMA source links.
Public commercial-mechanical gross margin
Tier A19% to 24%
Comfort Systems USA (FIX) 24.1%, FY2025; EMCOR (EME) 19.3%, retained FY2024 reference.
Public commercial-mechanical operating margin
Tier A10.1% to 14.4% reported
Comfort Systems 14.4%; EMCOR 10.1% includes an 85-basis-point UK sale gain (about 9.3% excluding it), FY2025. This is not a recurring operating-margin range.
Source: SEC 10-K filings FY2025
HVAC mechanic median wage (SOC 49-9021)
Tier A$59,810/yr
Mean $62,690, ~397,000 employed, median hourly $28.75.
Source: BLS OEWS, May 2024
Commercial mechanical DSO (implied)
Tier D~90 to 105 days
EMCOR ~91, Comfort Systems ~103, retained receivables/revenue calculations, FY2025. These are point-in-time public balance ratios; they do not measure private-shop invoice collection or net financing cost.
All-contractor net profit before tax
Tier B6.3% (Best in Class ~11.9%)
Historical 2023 financials, reported in the 2024 edition (1,290 valid financial respondents). The 11.9% reference is the Best in Class group, top 25% ranked on performance metrics, not the 75th percentile of net margin or a healthy-firm threshold.
Source: CFMA 2024 Construction Financial Benchmarker (n=1,290)
US HVAC contractor market size
Tier B~$159B, ~115,000-120,000 firms
Fragmented; no firm above a few percent share.
Source: IBISWorld, Heating & Air-Conditioning Contractors (2025/2026)
Service vs install gross margin (small/residential)
Tier Cservice/maintenance 50-58%, equipment install 35-48%, new construction 28-38%
Retained planning inputs, not a verified cohort or ordering. Compare complete direct costs and matched contract scope.
Source: Retained coaching references, directional planning assumptions; original cuts not reproduced
Tier A = stated public-company SEC or BLS observations; verify the exact period and scope. Tier B = trade-association survey or research-firm report (CFMA, IBISWorld). Tier C = industry publications or planning references, directional only. Tier D = Level calculations from filing inputs. Tier E = Level pooled research. Public-company figures describe large operators, not typical private firms or attainable upper bounds. Where sources conflict, both are shown, never averaged.
What the public HVAC and mechanical contractors actually earn
These FY2025 filings show consolidated results and selected segments. Work and customer classifications differ; they do not isolate service from construction.
| Company / scope | Gross | Operating | Net | What it tells you |
|---|---|---|---|---|
| Comfort Systems USA (FIX)Consolidated, FY2025 | 24.1% | 14.4% | ~11.2% | Large public commercial mechanical/HVAC contractor. Does not split service vs construction. |
| EMCOR (EME)US Mechanical Construction segment, FY2025 | n/d | 12.8% | n/d | The construction arm. |
| EMCOR (EME)US Building Services segment, FY2025 | n/d | 6.0% | n/d | Building Services includes facilities contracts, retrofits, controls projects and repair work. Operating margin is not a service-only gross margin. |
| Limbach (LMB)Consolidated, FY2025 | 26.2% | 7.7% | 6.0% | Owner-direct relationships, 75% of FY2025 revenue; not exclusively service work. |
EMCOR segment operating margins exclude corporate administration; they are not consolidated operating margins. n/d = not disclosed at that cut. EMCOR FY2025 consolidated operating margin (10.1%) includes an 85 bps one-time gain on a UK-ops sale; the reported US segment margins exclude the separately presented sale gain.
The Level lens: cost-completeness scenario (fictional)
Fictional HVAC service call: a 55% incomplete contribution estimate becomes 30% after three separately identified costs.
Illustrative cost-completeness bridge, percentage of revenue
Source / sample: Fictional scenario; assumed cost percentages, not a measured cohort
Illustrative assumptions on $100 of revenue: an initial incomplete contribution estimate of $55, then separately omitted costs of $11, $7 and $7, leaving $30. This is not measured gross or net margin. Check your ledger before each adjustment: subtract only cost missing from the starting estimate. Fully burdened direct labor already included in gross profit must not be subtracted again. Allocate shared overhead once, distinguish economic contribution from financial-statement margin, and do not treat a rounded scenario as a trade benchmark.
Public-company anchors used
Comfort Systems USA (FIX), EMCOR (EME), Limbach (LMB). These are large, publicly traded operators. Filing observations and Level calculations labeled Tier D are separately scoped reference points, not a stand-in for a typical private HVAC shop, which runs on different economics.
Level's pooled contractor references
These are pooled references from Level's 2,242-contractor research universe, with a separate eligible sample for each metric. Trade variation is not measured here. The pool covers HVAC, plumbing, electrical, mechanical, refrigeration and fire protection; it has no roofing or GC segment. Match your definitions and work mix before comparing.
Expand all 4 pooled cross-trade contractor reference metrics (n=2,242 universe)Includes blended gross margin, net profit, labor share, AR days, and cash reserve references across mechanical and specialty trades.
n=464 (blended). Cash collected divided by billed revenue across the six-trade pool; top decile 96.0%. The benchmark window is not published. Timing, credits and write-offs need separate reconciliation.
n=733 total; post-completion subset count not published (blended). 1 day with progress billing included; 7 days among post-completion invoicers. Subset and bottom-decile populations are not separately counted.
n=794 (blended). On decided quotes; 38.1% across all quotes.
n=1,747,089 jobs / 1,791 companies (blended). Blended across trades. Compare only jobs where revenue and direct costs are assigned to the same job.
How AI performs on HVAC QuickBooks takeover, close & job-costing checks (LevelBench-100)Why raw AI chatbots score 31% to 71% on messy contractor books, and how Level Playbooks from 2,200+ past experiences prevent bad QuickBooks entries.
Many HVAC owners are testing ChatGPT, Claude, or Gemini on QuickBooks exports to speed up the monthly close. In LevelBench-100 (v3.0) (100 binary checks across 12 real-world messy SMB takeover and CFO workflows), all 27 unassisted foundation models scored between 31.0% and 71.0% accuracy with a 14.3% to 57.1% bad book-entry rate on three common contractor ledger traps:
- Missing Check Stubs & Prior-Bookkeeper Manual JEs (WF01): Unassisted models guess expense categories on paper checks written to supply houses without check images, and double-count `Uncategorized Expense` rows that a prior bookkeeper already adjusted via manual Journal Entry.
- Walk-In COD Swipes vs. Open Vendor Bills (WF02): When a technician makes a walk-in counter debit purchase that matches an open Accounts Payable bill amount, or when an AP bill is paid via ACH/wire, raw AI models frequently misapply the payment or double-expense COGS.
- Unsigned CRM Estimate Options & WIP Cutoffs (WF06): Generic models recognize unsigned customer upgrade options or upfront deposits for future-month installations as current revenue while omitting completed-but-unbilled September field jobs.
Level pairs frontier AI models with Level Playbooks (built from 2,200+ past SMB experiences) and operator review to achieve 100.0% accuracy and a 0.0% bad book-entry rate across all 100 checks, holding unverified paper checks for evidence and reconciling historical QuickBooks subledgers before any entry is posted.
Frequently asked questions (5)Answers on HVAC gross vs. net margins, data sources, and safe QuickBooks automation.
What is a good profit margin for an HVAC contractor?
It depends on whether you mean gross or net, and on commercial vs residential work. The best hard reference points are public-company actuals: Public commercial-mechanical gross margin: 19% to 24%. Comfort Systems USA (FIX) 24.1%, FY2025; EMCOR (EME) 19.3%, retained FY2024 reference. This is a scoped reference, not a measured distribution of private HVAC firms.. The retained CFMA 2024 Benchmarker (2023 financials, 1,290 valid financial respondents) reports 6.3% net income before tax for all respondents and 11.9% for its Best in Class group, the top 25% ranked by performance metrics rather than a net-margin percentile. These are historical all-contractor references, not a measured HVAC-only cohort or a current target. Small private shops vary widely, so use the tiered sources on this page rather than a single number.
Where does this HVAC benchmark data come from?
Named sources selected for relevance to HVAC contractors, including separately labeled proxies and all-contractor references: public-company SEC 10-K filings, BLS wage data, CFMA's Construction Financial Benchmarker, and industry research, each tagged by confidence tier with a source link. Level's pooled research covers HVAC, plumbing, electrical, mechanical, refrigeration and fire protection; it excludes roofing and general contractors. It includes HVAC operators among these segments, but Level does not publish a separately measured HVAC-only cohort. Collection, billing and quote metrics are pooled cross-trade references; the broader 2,200+ research universe is not the contributing sample for each metric.
How can HVAC contractors safely use AI for running the books and the monthly close without corrupting QuickBooks?
Unassisted AI models pose serious risks when connected directly to a contractor general ledger. Across all 27 unassisted foundation models evaluated in LevelBench-100 v3.0 (100 binary checks across 12 messy SMB QuickBooks takeover and CFO workflows), raw takeover accuracy ranged from 31.0% to 71.0% and every model posted a 14.3% to 57.1% bad book-entry rate: double-counting expenses already adjusted by a prior bookkeeper's manual Journal Entry or guessing expense categories on paper checks without check images (WF01), confusing walk-in counter debit purchases with open Accounts Payable vendor bills (WF02), and recognizing unsigned CRM estimate options or upfront customer deposits as current-month revenue (WF06). Level avoids these risks by pairing AI with Level Playbooks (built from 2,200+ past SMB experiences) and operator review (100.0% accuracy, 0.0% bad book-entry rate) that require check-image evidence, AP subledger matching, and WIP cutoff verification before any entry reaches QuickBooks.
Is HVAC service work higher margin than installation?
The retained directional planning references are service/maintenance 50-58%, equipment install 35-48% and new construction 28-38%, not a measured cohort or a universal ordering. Materials, paid nonbillable time, warranty scope and cost allocation can reverse the comparison. At scale, distinguish contract types: EMCOR's mixed building-services segment ran a 6.0% operating margin in FY2025 versus 12.8% for its mechanical-construction segment, in the reported segment comparison. The owner-direct customer-relationship comparison shows why segment definitions need matching; it does not isolate service pricing. Match contract type and gross-versus-operating definitions before comparing.
Why is the public-company HVAC margin so much lower than the 50%+ I hear about?
Two different things are being measured. The 50%+ figures are job-level or service-line gross margins on residential work before overhead. Public-company numbers like Comfort Systems' 24.1% are company-wide GAAP gross margins across a commercial-heavy, equipment-heavy, new-construction mix, after all job costs. Both can be true at once. Use the one that matches your work type, and never compare a job-level number to a company-wide one.
Tier A and B rows name their source and link the record or source index where available. The full tiered dataset for all six trades (public-company actuals, BLS wages, CFMA and market figures, each with its source URL) is downloadable and free to cite with attribution. Download the per-trade benchmark facts (JSON).
See where your HVAC numbers land
We pull your margin, labor economics, and collections and compare your matched-scope figures with public references and the pooled Level Index, with each sample identified, then show you the first leak to fix.
No commitment. Real numbers, not generic advice.