The Blended Number

Ed Kowalski's revenue looked like one number. It was two.

Meridian Climate Systems reported $28M in annual revenue. The P&L showed a single line. The blended gross margin was healthy. Ed talked about the business as one business because that is how he had always run it.

The buyer's commercial team did not see one business. They saw two — and the two were not equally valuable.

The blender problem

A blender takes different ingredients and produces one smoothie. You cannot taste the kale separately from the mango. That is the point.

Ed's P&L was a blender. It took $18M in maintenance-contract revenue and $10M in project revenue — new installations, retrofits, tenant buildouts — and produced one blended gross margin. The margin looked healthy. Ed pointed to it in every meeting.

The buyer's commercial associate asked for the revenue broken out by type. Ed said he did not track it that way. Janet pulled it together in a week.

When the numbers came back separated, they told a different story.

Two businesses, two margins

Maintenance contracts are the engine of an HVAC services business. The technician shows up on a schedule. The scope is defined. The pricing renews annually with a built-in escalator. Labor is planned. Parts are stocked. The customer calls when something breaks, and the contract covers the visit.

Gross margins on a mature service book run 45% to 55%. The revenue is recurring. The cash flow is predictable. The buyer underwrites this stream at full value because it behaves like a subscription.

Project revenue is a different animal. New installations, retrofits, and tenant improvements are won through competitive bidding. The bid compresses the margin. Scope creep eats the estimate. Labor is variable — project crews are hired up and laid off with the backlog. Materials pricing fluctuates with the supply chain. A bad job can go negative.

Gross margins on project work run 18% to 28%. The revenue is lumpy. The backlog is the only forward visibility, and backlog in commercial HVAC is typically four to six months.

Ed's blended gross margin was 38%. That number was real. It was also meaningless to the buyer.

What the buyer actually models

The buyer's commercial team splits the revenue into its component streams and runs EBITDA contribution by type. The question is not what is the total margin but where does the margin actually come from.

When they ran Meridian, the maintenance book — 64% of revenue — was generating roughly 78% of the EBITDA contribution. The project book — 36% of revenue — was generating roughly 22%.

The project revenue looked like $10M on the top line. Its EBITDA contribution looked like a business half that size.

The buyer then applies a different multiple logic to each stream. Recurring, high-margin, contractually protected revenue gets underwritten at full value. Lumpy, low-margin, competitively bid revenue gets haircut — lower growth assumption, higher attrition rate, compressed margin in the forward model.

Ed's growth projections assumed the blended margin held. The buyer's model assumed the mix would shift toward project work as the service book matured. The difference between those two assumptions was the finding.

What Ready Looks Like

The founder who understands margin by revenue stream before diligence controls the narrative. The founder who presents a blended number lets the buyer write the story.

Separate the P&L by revenue type. Maintenance contracts, project work, emergency service calls, parts sales — whatever the streams are. Each one gets its own gross margin, its own growth rate, its own forward assumption.

Know the contribution split. If 60% of revenue generates 80% of EBITDA, that is the story the founder should tell first. The buyer will find it anyway. The founder who presents it proactively demonstrates financial fluency. The founder who does not demonstrates the opposite.

Price the project book honestly. Project revenue is not bad. It is volatile, lower-margin, and harder to underwrite. A founder who can show three years of project-level profitability data — win rates, margin by job, backlog trend — gives the buyer something to underwrite. A founder who cannot forces the buyer to assume the worst.

Show the service book's durability. Renewal rates, average contract tenure, annual escalator history, customer retention by cohort. The service book is the crown jewel. Prove it.

For Meridian, none of these were true.

Ed sat down with his investment banker to review the buyer's initial findings.

The banker pointed to a line in the commercial summary. "They're splitting your revenue two ways. Maintenance and project."

"We've always run it as one P&L," Ed said.

"I know. They're not."

The banker showed him the buyer's margin analysis. Maintenance contracts: 52% gross margin. Project work: 23%. Ed's blended 38% had masked a 29-point spread between his two businesses.

"They're haircutting the project revenue in the forward model," the banker said. "Lower growth, higher attrition, compressed margin."

Ed stared at the page. "How much?"

What this cost Ed: $375K.

The buyer's commercial team is not penalizing Ed for having project revenue. Every HVAC services business has a project book. The buyer is pricing the risk that Ed's blended margin overstates the quality of his earnings.

Maintenance revenue at 52% gross margin is worth more per dollar than project revenue at 23%. When the buyer re-ran Meridian's forward EBITDA model with the streams separated, the projection came in lower than Ed's blended forecast. The delta — the gap between Ed's blended assumption and the buyer's stream-weighted model — priced out at $375K. The buyer framed it as a forward earnings adjustment: the EBITDA Ed projected assumed margins that only held if the revenue mix stayed constant. The buyer's model assumed it would not.

One more cut. The deduction list is at $3.5M not because any single finding is large enough to walk away from, but because twenty-one of them add up.

Don't be Ed.