A PE firm just paid a reported $800 million-plus for a garage-door services platform at a reported ~16x EBITDA.
Sixteen times is a software number. It is the multiple you pay for recurring revenue, sticky contracts, and a data moat. It is not the multiple anyone associates with technicians driving vans to fix broken springs and replace panels.
According to PitchBook, the deal closed at that level anyway. And it was not an outlier. There were roughly two dozen garage-door related PE deals in 2025. The category went from invisible to contested in a single year.
So either a serious investor overpaid by a wide margin, or they are betting on something the headline number does not name. I think it is the second one. And I think the thing they are betting on is the data.
The HVAC pattern, repeated
We have seen this movie. A decade ago, residential HVAC was a fragmented field of family-owned shops. Then PE noticed the recurring service revenue, the emergency pricing power, and the sheer number of small operators available to consolidate. The multiples climbed from low single digits to double digits. Buy-and-build did the rest.
Garage doors fit the same template. Highly fragmented. Mostly mom-and-pop. Recurring repair and replacement demand tied to the housing stock. Local brand loyalty that a national platform can absorb. The structural logic of a home services consolidation play is intact.
But the HVAC winners did not win on the structural logic alone. Plenty of firms ran the same playbook and ended up with a holding company that owned thirty businesses and operated none of them as one. The winners won because they turned thirty separate operators into a single legible entity. That is a data outcome, not a deal outcome.
What the 16x is actually pricing
A services company that runs as a federation of acquired shops trades like a services company. Low multiple. Discount for complexity. The buyer assumes you cannot see across the portfolio cleanly and prices that uncertainty in.
A services company that runs as one analyzable entity trades like a platform. The buyer can underwrite cross-sell, route density, technician utilization, and pricing optimization because the numbers exist and reconcile. That is where the multiple expansion lives.
The 16x is not a bet on garage doors. It is a bet that this platform can do what most roll-ups never finish. Take dozens of fragmented local businesses and make them speak one language at the data layer. Same definition of a job, a customer, a service call, revenue, and margin across every operator that gets folded in.
If they pull that off, 16x was cheap. If they do not, 16x was a mistake that reveals itself slowly and then all at once at exit.
What standardizing 30-plus tuck-ins actually requires
Here is what the spreadsheet version of buy-and-build leaves out. Every tuck-in arrives as its own data island.
Different field service software. One operator runs ServiceTitan. The next runs Housecall Pro. The third runs a paper job book and a QuickBooks file the owner’s spouse maintains. The fourth built something custom in 2014 and the developer is long gone.
Different definitions. A “completed job” means one thing where the tech closes the ticket on site and another where the office closes it after payment clears. “Revenue” is recognized on different triggers. A “customer” is a household at one shop and a property at another and a phone number at a third. None of this was documented because none of these owners ever expected to be consolidated.
Different data quality. One operator captures every job with clean line items and a customer record. Another logs revenue in lumps with no service detail. The data you need to underwrite the thesis, recurring revenue, attach rates, repeat customers, route efficiency, only exists if someone captured it consistently, and most small operators did not.
This is the same trap I described in the add-on surge data integration crisis. Each acquisition brings its own system, its own definitions, and its own reporting cadence, and the deal model budgets for none of it. The difference here is scale. HVAC and garage-door roll-ups are not buying three add-ons. They are buying thirty, forty, fifty. The integration problem does not add up. It compounds.
Standardizing across that many operators is not a software migration you run once. It is a set of decisions you have to make and then enforce on every deal that follows. What counts as a job. When revenue is recognized. How a customer is identified across a household, a property, and a phone number. How a service call maps to a category that means the same thing in every market. How margin gets calculated when one operator allocates overhead and the next does not. Each of these is a small decision on its own. Made forty times, inconsistently, they produce a consolidated number that cannot be trusted and cannot be fixed retroactively without going back into every source system.
The work that makes this survivable is unglamorous. A canonical data model the platform defines before it buys anything. A mapping playbook that tells the integration team how to translate any new operator onto that model. A standard for what gets captured at the point of service so the data exists in the first place. None of this is exotic technology. It is operating discipline applied early, and it is exactly the discipline that most roll-ups skip because it does not show up in the deal model.
Why the compounding is the whole game
With two operators on two systems, reconciliation is manual but survivable. One person in finance keeps a master spreadsheet and the board deck looks consolidated.
With thirty operators on a dozen systems, that approach collapses. The reconciliation work scales geometrically, not linearly, and the single person maintaining the master file becomes the most load-bearing and least visible employee in the company. I went deeper on that mechanic in data integration after an add-on acquisition.
This is why the standardization has to be designed at the platform level on day one, not improvised after the tenth deal. The firms that win at this make data integration a condition of the tuck-in, not a project they get to later. Before the ink dries, they know which system the operator runs, how they will map it to the platform standard, and who owns the migration. The standard exists first. The acquisitions conform to it.
The firms that lose treat each acquisition as a revenue line and assume the systems will sort themselves out. They never do. By deal twenty, the platform is a holding company wearing a trench coat, presenting one number to the board that nobody can actually defend.
Smart multiple or expensive mistake
The 16x sorts into one of two outcomes, and the data layer decides which.
In the smart version, the platform standardizes early. By the time it has folded in thirty operators, it can pull route density by region, technician utilization by branch, attach rate by service type, and repeat-customer revenue across the entire footprint, all on one definition, in a day. That capability is the moat. It lets the operator price better, route smarter, cross-sell into the installed base, and underwrite the next tuck-in with confidence. A buyer pays a platform multiple for it because the platform is real. The data proves it.
In the expensive version, the platform grows by acquisition and never standardizes. The consolidated numbers are a monthly act of manual reconciliation that nobody fully trusts. Organic growth cannot be separated from acquired growth. Customer counts are a range, not a number. When the exit process starts, the buyer’s diligence team asks for entity-level revenue on a consistent definition and the company produces a different answer every time it runs the query. If you want to pressure-test how a roll-up holds up under that kind of questioning, the Data Room Survivor tool walks through it.
At that point the multiple does not hold. The buyer reprices the uncertainty, introduces an earnout, or walks. The premium that the original firm paid on the way in evaporates on the way out, because the thing that justified the premium, one legible entity, was never built.
This is the same gap that turns a data carve-out into a cost-of-ownership surprise. The systems and definitions you skip on the way in become the bill you pay on the way out, with interest.
The bet under the bet
Blue-collar trades consolidation is one of the most interesting things happening in mid-market PE right now, and the garage-door 16x is the clearest signal yet. Serious capital is betting that fragmented local services can be repriced as analyzable platforms.
That bet is sound. The HVAC winners proved it. But it is a buy-and-build data bet wearing a services-roll-up costume, and it only pays off for the firms that understand that distinction before they sign the platform deal, not after they have stapled forty operators together and discovered they cannot read their own numbers.
If you are running or backing a trades roll-up, the question is not whether the structural thesis works. It does. The question is whether you have made standardizing the data layer a condition of every tuck-in, or whether you are quietly assuming it will happen on its own.
It will not happen on its own. It never has. And on a platform built to fold in thirty companies, the gap between the smart multiple and the expensive mistake is exactly the gap between a data layer you designed and one you hoped for.