Selling an HVAC or Plumbing Business: Valuation and Buyers

Most owners land on one of three routes, and the right one depends less on what your business is worth than on what you want next.

Retire and sell outright. You want out. No earnout, no rollover equity, no staying on for two years managing someone else's platform. This usually means selling to an individual buyer, a competitor, or an SBA-backed buyer if the company is small enough to run without you. It's the cleanest exit but often not the highest headline price.

Sell to management or an employee. If you have a service manager or GM who could run the business, an internal sale can work, provided that person can hold or obtain the required license and qualify for SBA 7(a) financing. Standard 7(a) loans run up to $5 million with a 75% SBA guarantee, and the regulatory maximum variable rate is base rate plus 3.0% for loans above $350,000, which puts the ceiling around 9.75% at July 2026 prime of 6.75%. Negotiated rates can land lower.

Sell into a growth story. Roll equity into a larger platform, keep operating for a few years, and take a second bite when the platform itself sells or goes public. This is the PE add-on path, and it only works if your company clears the management-run threshold, usually $500,000 to $1 million in adjusted EBITDA and up.

None of these are mutually exclusive at the LOI stage. A buyer's first offer will tell you which lane you actually fall into.

The one disclosed reference point at real scale: Gamut Capital bought Airtron from NRG Energy in August 2024 for $500 million, reported at 8.6x EBITDA. Airtron was a large residential new-construction and maintenance provider across 13 metro areas, so treat it as a ceiling reference, not a comp for a local shop.

BizBuySell's five-year median for sold HVAC companies was $750,000 sale price on $1.482 million revenue and $304,309 SDE, with 186 median days on market. The 2025 median sale price rose to $800,000. Asking multiples run higher than closed multiples across the board (asking SDE multiple averaged 3.16x versus 2.75x actually paid), which tells you something about the gap between what a broker lists and what a buyer signs for.

The buyer landscape has three distinct lanes, and they don't compete with each other for the same deal.

Residential PE platforms. Apex Service Partners (backed by Alpine Investors), Wrench Group (Leonard Green & Partners, with TSG Consumer and Oak Hill, operating brands including Coolray, Berkeys, Abacus, Parker & Sons, and Morris-Jenkins), Sila Services (Goldman Sachs Alternatives, 28 acquisitions by late 2024), TurnPoint Services (OMERS Private Equity), Heartland Home Services (The Jordan Company, 35 brands across eight Midwest states), Champions Group (Blackstone), and Redwood Services all buy residential HVAC, plumbing, and electrical add-ons, typically in the 4.5x-8x EBITDA range depending on size and quality.

Commercial and industrial consolidators. PremiStar (Partners Group, targets $10M-$100M revenue commercial HVAC and controls), FirstCall Mechanical (SkyKnight Capital, publishes a $5 million-plus revenue threshold, 15 acquisitions since January 2022), Crete United (Ridgemont Equity Partners, 18 add-ons by late 2024), Orion Group (Alpine Investors, 35-plus acquisitions), NexCore (Trinity Hunt Partners), and Service Logic (Bain Capital) buy commercial mechanical contractors, generally 5x-10x EBITDA.

Public strategics. Comfort Systems USA (NYSE: FIX), Limbach Holdings (NASDAQ: LMB), and EMCOR Group (NYSE: EME) are the largest strategic acquirers, but they're realistically fit only for scaled, management-run, profitable operations, not a typical owner-operated shop.

Individuals, competitors, and search funds remain the natural buyer below roughly $500,000 SDE, typically paying 1.8x-3.5x SDE and financing with SBA 7(a) debt. Local strategics can pay a premium over a pure financial buyer where route density, technicians, permits, or territory create real synergy.

Recurring maintenance revenue helps, but only when agreements actually renew, carry positive gross margin, and have documented obligations, not just a plan count on a spreadsheet. Owner dependence is usually the biggest single swing: if you're the qualifier, the salesperson, and the dispatcher, a buyer has to price in your replacement before the multiple even applies. Customer concentration above 15%-20% from a single builder, GC, or property manager can trigger an earnout or holdback. And clean, accrual-based financial reporting with reconciled WIP matters more than most owners expect, because unsupported EBITDA gets repriced during quality-of-earnings review, often at the same multiple that inflated it in the first place.

A prepared Main Street sale typically runs 5-9 months from engagement to closing, though BizBuySell's reported 186 median days on market doesn't include most of the pre-sale preparation work. A PE or strategic lower-middle-market process usually runs 6-12 months. Both are estimates; company quality, financing, and licensing consents swing them considerably.

Preparation matters more than most owners expect. Improving revenue mix, management depth, and clean accrual reporting takes 12-24 months. It cannot be manufactured during a diligence process that's already underway.

1. Treating SDE as EBITDA. The single most value-destroying error, and the easiest for a buyer's QoE team to catch.
2. Counting maintenance agreements instead of underwriting them. Renewal rate, churn, and margin matter more than the raw plan count.
3. Letting technicians hear about the sale from rumors. Key-person flight during diligence can stop financing outright.
4. Blending revenue types. New construction, subcontracted work, and service revenue carry different margins; buyers will unwind the blend and reprice accordingly.
5. Assuming the license transfers automatically. It usually doesn't. Build the continuity plan before you go to market, not after an LOI.
6. Leaving digital assets in your personal name. Google Business Profile, call-tracking numbers, and the main phone line need to be transferable at closing.