Crossing the $5M EBITDA Line Can Re-Rate Your Whole Multiple

Two companies, same EBITDA, very different price

In this fictional example, both do $5M EBITDA. Applying assumed 7x and 11x multiples gives $35M and $55M enterprise value. Same trade, same region, same number on the bottom line.

The hypothetical $20M difference follows from the selected multiples, not evidence that one recurring-revenue attribute causes it. Renewal quality, size, specialty, management depth and transaction structure all need review.

The three bands, and where the lines actually fall

For HVAC, the published ranges are these. We do not measure transaction multiples and do not publish any of our own: every band below is externally sourced. The bands are illustrative advisory planning ranges assembled on our valuation reference page, not a measured Level transaction cohort or a direct quote of one Capstone table. Operating data can frame diligence; it does not determine a company's multiple.

Note the overlap between the first two. That is not sloppiness in the data, it is the honest shape of the market: a $5M business run one way prices like the top of a tuck-in, and run another way prices like the bottom of a platform. The band you fall into is a judgment a buyer makes, not a bracket your revenue puts you in automatically.

The arithmetic, on round numbers you can substitute your own into

Take an illustrative contractor at $4M of EBITDA and move it to $5M. Nothing about this is a specific company; substitute your own figures.

On those midpoints, 25% more EBITDA is 88% more company. But the midpoint pairing is a choice, not a property of the data, and the honest range is wide because the two bands overlap between 7x and 8x. Run the same crossing at both extremes:

• $4M at 8x ($32M) to $5M at 7x ($35M) is +9%, for 25% more profit
• $4M at 6x ($24M) to $5M at 9x ($45M) is +88% (midpoints)
• $4M at 4x ($16M) to $5M at 11x ($55M) is 3.4x

All three are consistent with the published bands. Anyone quoting you only the third one is selling something.

So the band change is worth a great deal, and where you land inside the band is worth more than crossing it. An owner who grows EBITDA while ignoring what kind of EBITDA it is can cross a boundary and barely move the price.

What a buyer tests when deciding where in the band you land

Recurring maintenance profitability is one attribute to investigate, alongside size, management depth, density and deal structure. Agreement count or a revenue label alone does not settle it.

Measured across 259 contractors with real service-agreement revenue, the median gross margin on that recurring work is 37.9% and the top-quartile boundary is 53.5%. Below about 25% warrants a cost and scope review. Pricing, covered work, cost allocation and customer mix are candidate explanations; the distribution does not identify cause.

Agreement margin is one attribute to test. The 259-company distribution does not explain transaction multiples or establish that a 53.5% margin earns 11x. Scale, density, management depth, customer risk and deal structure also separate outcomes.

Revenue growth on its own does not do this. Doubling installation revenue may change scale and risk. It does not automatically establish a tuck-in classification or a 4x-to 8x multiple.

The three tests a buyer runs before paying the higher band

These checks inform a buyer's judgment; source coverage determines whether each is answerable:

1. Recurring revenue proof. Agreement count, renewal rate, and gross margin per agreement after fully burdened technician hours. Not the revenue category label, the actual burdened margin.
2. Maintenance pull-through. The repair and replacement revenue an agreement customer generates compared with a non-agreement customer. This is what separates an agreement that renews from a discount you give away annually.
3. Install versus service margin split, computed from actual job cost rather than from how the revenue was categorized.

If you cannot produce those three, retain the unresolved questions for diligence. A buyer may apply a risk adjustment, but neither its direction nor its size is established by this dataset.

What to measure monthly if crossing the band is the plan

• Service-agreement gross margin, burdened. Against a median of 37.9% and a top-quartile boundary of 53.5%.
• Share of EBITDA from recurring work. Not revenue share, EBITDA share.
• Renewal rate, and the reason given for every non-renewal.
• Pull-through ratio: additional non-agreement revenue from agreement customers divided by agreement revenue. Compare non-agreement customer cohorts separately.

Four numbers. They are also the four a buyer will ask for, which is the point: the work of becoming worth more and the work of proving it are the same work.

FAQ

What EBITDA multiple can an HVAC company sell for?

Tuck-ins in the $1M to $5M EBITDA range price at 4x to 8x, regional platforms in the $5M to $25M range at 7x to 11x, and national platforms with high service-agreement density at 15x to 18.5x and above (illustrative advisory planning bands, not a company valuation). The bands overlap because a $5M business run one way prices like the top of a tuck-in and run another way like the bottom of a platform.

Does growing EBITDA raise my multiple as well as my profit?

It can, because crossing a band re-rates the multiple applied to the whole business. The honest range is wide, because the two bands overlap between 7x and 8x. Taking $4M of EBITDA to $5M is 25% more profit and, depending where in each band you land, anywhere from +9% more enterprise value to 3.4 times the starting value (+244%), about 88% more at the midpoints. In the fictional comparison, the selected multiple matters more than the size step; actual transactions may differ.

What moves a contractor from the tuck-in band to the platform band?

Buyers may test recurring maintenance profitability alongside scale, density, management depth, customer risk and deal structure. Revenue growth alone does not establish a classification. Median service-agreement gross margin is 37.9% across 259 contractors and the top-quartile boundary is 53.5%. Doubling installation revenue does not automatically establish a classification or multiple; review scale and business quality together.

What will a buyer check before paying the higher multiple?

Three things, each answerable from your own job and agreement data: recurring revenue proof (agreement count, renewal rate, and burdened gross margin per agreement), maintenance pull-through compared with non-agreement customers, and the install versus service margin split computed from actual job cost.

Source and claim note: The band ranges and scopes are EXTERNALLY sourced, not ours. Level does not measure transaction multiples and publishes none of its own; the bands are illustrative planning ranges on Level's valuation reference, not a direct quote of one report. Capstone's July 2026 update reports 9.5x EV/EBITDA for HVAC services during 2024-YTD2026, versus 13.3x during 2021-2023. Its separate broader-HVAC comparison is 11.4x versus 13.4x. Those are different transaction populations and neither is the hypothetical size-band table used here. Private-market multiples vary by deal size, mix and cycle. Businesses under about $1M of owner earnings are usually priced on SDE at lower multiples than the tuck-in band shown here. Service-agreement gross margin is measured from the Level Index across 259 contractors with at least $10K of service-agreement revenue. The worked example uses round illustrative figures so a reader can substitute their own; it is not a specific company's trajectory, and enterprise value in any real transaction depends on far more than one multiple applied to one year of EBITDA.