Latest / The Buyer & Seller Podcast with Fexingo: Business Brokers, Exits, and Private Sales Explained / How a Roofing Company Sold at 130x EBITDA with Recurring Storm Data
Transcript
- Lucas: So we talk a lot on this show about what drives a multiple through the roof. Recurring revenue. Contracts. Customer retention. Route density. All that stuff. But every now and then a deal comes across the desk that makes you rethink the whole framework. Luna: I'm guessing this is the one where the multiple hit triple digits. Lucas: One hundred and thirty times EBITDA. A roofing company in the Midwest. Not a tech platform. Not a SaaS product. A roofing company with trucks, ladders, and crews. Luna: Okay, I have to ask — was there some kind of recurring revenue loop? Like an annual inspection contract? Lucas: No. Zero recurring revenue. They did project-based residential and commercial roof repairs. The kind of business you'd normally see trade at four to six times EBITDA. Luna: So what changed the math? Lucas: Data. Specifically, a decade of proprietary storm-damage roof assessments. They had catalogued every hail event, every windstorm, every roof claim in their service area going back to 2016. Over seven thousand individual roof inspections, with photos, insurance adjuster notes, and repair outcomes. Luna: And the buyer saw that as more valuable than the actual roofing business? Lucas: Exactly. The buyer was a national consolidator — one of the big private equity backed roofing roll-ups. They weren't just buying the cash flow. They were buying a decade of hyper-local weather risk intelligence that would let them price their own jobs more accurately, target marketing to neighborhoods most likely to need a new roof in the next twelve months, and even sell that data to insurance carriers. Luna: So the multiple wasn't really based on EBITDA at all. It was based on the value of the data set. Lucas: Right. The EBITDA was just the ticket to get the buyer in the door. The real negotiation was about the data moat. And the buyer calculated that the data alone was worth more than the entire operating business. Luna: What kind of data are we talking about specifically? Like, what made it so valuable? Lucas: Every inspection included GPS coordinates, roof pitch, material type, age of the roof, the specific date of the storm event, the insurance claim outcome — was it approved or denied? — and the actual repair cost. Over time, that becomes a predictive model. You can look at a new property and say, 'Given the last five hail events in this zip code, the probability of a claim in the next two years is X percent.' Luna: That's essentially building an actuarial table. For roofs. Lucas: Exactly. And the buyer told the seller — this is from the actual deal memo — that they valued the data at $7.2 million. The entire business was generating about $800,000 in EBITDA. So a normal multiple would put the value around $3 to $4 million. The data premium doubled that. Luna: So the seller effectively got paid twice. Once for the business, once for the intelligence they generated incidentally. Lucas: That's a great way to frame it. And it raises a bigger question: how many other small businesses are sitting on similar data assets they haven't realized have value? Luna: I can think of a dozen service industries where that could apply. Pest control. HVAC. Tree service. Any business that does inspections and keeps good records. Lucas: But the key is the data has to be structured, consistent, and longitudinal. This wasn't a stack of paper invoices in a filing cabinet. They had a custom database where every field was standardized. They could query it. That's what made it salable. Luna: And I assume they had to prove the data was proprietary. Not just publicly available weather records. Lucas: Right. Public weather data tells you a storm happened. This data tells you what actually happened to the roofs in that storm. The insurer's decision, the repair cost, the material failure rate. That's proprietary because it comes from on-the-ground inspections. No two companies would have the same data set. Luna: It's kind of wild that the buyer was willing to pay that much for data from a single geographic region. I mean, the Midwest is big, but it's not the whole country. Lucas: The consolidator was already operating in eight states. They planned to use this data as the template for their own data collection in every new market they entered. So the value wasn't just the existing rows of data — it was the methodology. They bought the playbook as much as the numbers. Luna: And you mentioned they also sell the data to insurers. That must be a nice additional revenue stream. Lucas: Post-acquisition, yes. The buyer already had relationships with two regional carriers who licensed the data on an annual basis. That alone is expected to generate about $600,000 a year in pure margin revenue. So the data asset is self-funding the debt they took on to buy the company. Luna: That makes the 130x multiple look almost reasonable. Lucas: It does. And it's a great lesson for owners listening. If you have any kind of proprietary data — customer behavior, job outcomes, pricing elasticity — start thinking about how to package it. Even if you're not planning to sell, having that asset documented increases your leverage. Luna: And if you are planning to sell, this case suggests you should talk to a broker who understands data valuation, not just EBITDA multiples. Lucas: Exactly. A lot of mainstream business brokers would have put this roofing company on the market at 4 or 5 times EBITDA and never even mentioned the database. The seller here worked with a boutique firm that specialized in data-driven businesses, and that made the difference. Luna: That's the kind of thing that makes me think about the broader shift in M&A. It's not just about cash flow anymore. It's about intangibles. Lucas: And that's actually a good place to mention something. We cover these deals because we believe understanding the real mechanics of value helps business owners make better decisions. We don't run ads on this show — never have, never will. If you find these conversations useful and want to support that approach, there's a link. It's buy me a coffee dot com slash fexingo. Totally optional, but it helps keep the podcast independent. Luna: Yeah, and honestly, it's because of listeners who chip in that we can spend time digging into deals like this instead of just reading headlines. Appreciate it. Lucas: Alright, back to the roofing deal. There's another layer here that I found fascinating. The seller — his name is Greg — told me that the buyer also insisted on a three-year earnout tied to the accuracy of the data in predicting future claims. Luna: So the data itself had a performance clause? Lucas: Exactly. The buyer wanted to verify that the data set wasn't cherry-picked. So Greg gets an additional payout if the actual claims in the region match the model's predictions within a certain error band. Luna: That's a smart buyer. And it must have forced Greg to be very honest about the data quality during due diligence. Lucas: It did. He had to let a third-party data auditor review the entire database. They tested for sampling bias, missing fields, inconsistency in inspection methodology. And it passed. That audit report was probably the single most important document in the deal. Luna: So if you're a business owner thinking about a similar exit, what's the first step? Start building the database? Lucas: First, audit what you already have. Most businesses have more data than they think. Second, standardize the fields. If you're not collecting the same information on every job, you can't scale the data asset. Third, timestamp everything. Longitudinal data is exponentially more valuable than one-off snapshots. Luna: And fourth, keep it clean. Bad data is worse than no data. Lucas: Absolutely. Greg's data set had less than one percent missing fields. That's rare. But it's what allowed the buyer to trust it. Luna: I'm curious — did Greg ever think about monetizing the data before the sale? Like selling it to insurers directly? Lucas: He told me he thought about it but never got around to it. He was too busy running the roofing company. And honestly, that may have worked in his favor. If he had started licensing the data, the buyer might have considered the asset already exploited. It was still fresh. Luna: So there's a strategic argument for keeping your data proprietary until you exit. Lucas: If you can afford to, yes. Greg's roofing business was profitable enough that he didn't need the data income. That gave him the luxury of packaging it as a pristine asset. Luna: I think the takeaway for this episode is that the highest-leverage move a small business owner can make is to recognize what they actually own. It's not always the obvious thing. Lucas: And that applies whether you're in roofing, plumbing, or pest control. If your business generates intelligence, that intelligence has a market. The question is whether you've organized it enough for someone to buy it. Luna: Next episode, we're going to look at a deal where the data was actually the liability. A company whose customer list was so outdated it dragged down the multiple. Lucas: That's a good one. But for now, if you want to read Greg's actual deal memo — redacted, of course — we'll put a link in the show notes. And if this kind of analysis is valuable to you, you know where to find us. Luna: Buy me a coffee dot com slash fexingo. Thanks for listening.