Latest / The Buyer & Seller Podcast with Fexingo: Business Brokers, Exits, and Private Sales Explained / How a Commercial Bakery Sold at 65x EBITDA with Customer Density Data
Transcript
- Lucas: So there's this commercial bakery outside of Portland — let's call it Columbia Crust — that supplied croissants, baguettes, and sandwich rolls to about 140 coffee shops and cafes across Oregon and southern Washington. Luna: I think I've heard of a similar setup. Independent bakeries like that usually sell for 3 to 5 times EBITDA, right? Not exactly a headline multiple. Lucas: That's exactly the baseline. But Columbia Crust sold for 65 times trailing EBITDA. Twelve point five million dollars on roughly one hundred ninety-two thousand in adjusted EBITDA. Luna: Sixty-five times? That's absurd unless there's something else baked into the deal. No pun intended. Lucas: The something else was data. Specifically, customer density data mapped onto delivery routes. The owner, a guy named Doug, had been running the same routes for fifteen years. He knew which stops ordered the same quantity every Tuesday and Thursday, which ones fluctuated seasonally, and which ones had started ordering from a competitor then came back. Luna: But every baker knows that stuff in their head. How did he turn it into an asset that a buyer paid for? Lucas: He spent about six months before listing the business working with a part-time data analyst to build what they called a 'delivery density map.' They took every delivery address, geocoded it, overlaid it on a route optimization software that also tracked order frequency and average basket size per stop. Luna: So instead of saying 'we deliver to 140 cafes,' he could show the buyer that 80 percent of his revenue came from 60 stops clustered in five zip codes, and those stops ordered 98 percent of the time on schedule. Lucas: Exactly. And more importantly, he could identify the gaps. There were two zip codes with high cafe density where he had zero penetration. The data showed the demand existed — same income levels, same foot traffic patterns — but he'd never hired a salesperson to go after them. Luna: That's the lever. A buyer sees that and thinks: I don't need to grow revenue organically from scratch. I just need to replicate the existing route model into adjacent white space. The multiple starts to make sense if the buyer can double revenue without doubling headcount. Lucas: Right. The buyer was a private equity-backed roll-up platform called Artisan Bread Group. They already owned bakeries in Seattle and San Francisco. Their whole thesis was: acquire regional players, centralize back-office, then expand each route network by 30 to 50 percent within two years. Luna: And Columbia Crust's data gave them a ready-made expansion blueprint. The buyer probably valued that at way more than the actual cash flow. Lucas: That's the key insight. The purchase price was twelve and a half million. Of that, about seven and a half million was cash at close. The remaining five million was an earnout tied to hitting specific route expansion targets — adding thirty new accounts within eighteen months in those underpenetrated zip codes. Luna: So Doug didn't just walk away with a check. He had to stay on and help execute the data play. Lucas: He stayed as a part-time route consultant for eighteen months. And he earned out the full five million because the data was accurate. They added thirty-two accounts in the first twelve months. Luna: You know, this reminds me of something I read about a linen supply company that sold for a huge multiple because they had route density data. It's the same principle: if you're a B2B delivery business, the data about where your customers are and how often they order is often more valuable than the actual product you're delivering. Lucas: The product becomes the proof of concept for the data. The buyer is really buying a map with proven demand nodes. And if you're a small business owner listening, this is something you can start building today, even if you're not planning to sell for years. Luna: Honestly, if today was worth a coffee to you, that's the link — buy me a coffee dot com slash fexingo. It's how we keep this show ad-free and focused on real deal mechanics. Lucas: Yeah, listener support makes a difference. We don't run ads, so it's just people who find value in the show chipping in. Appreciate it. Luna: Anyway, back to the bakery. Doug's earnout structure also included a clause that if the buyer added accounts outside the agreed zip codes, those didn't count toward the earnout. That's an important detail. Lucas: Crucial. Because without that, the buyer could just add accounts anywhere and claim they hit the target without actually executing the density play. Doug's lawyer insisted the earnout be tied to specific geographic areas. Luna: So the earnout was essentially a bet on the data's predictive power. And it paid off. Lucas: It did. And the buyer's calculation was: even if Doug's earnout is five million, the incremental EBITDA from those thirty-two new accounts is about one point two million annually. At a conservative 10-times multiple, that's twelve million in added enterprise value. Luna: So the earnout was a bargain for them. They got a five-million-dollar liability that generated twelve million in value. Lucas: Exactly. And that's why deals like this happen. The seller has an asset — the data — that the buyer can monetize better. The earnout bridges the valuation gap. Luna: What about the actual baking operations? Did the buyer keep Doug's head baker? Lucas: They did. The head baker had been there twelve years. The buyer put her on a three-year employment contract with a small equity stake. They knew the product quality was part of the customer retention data. Luna: So the data wasn't just about locations. It also included order consistency. A cafe that orders the same quantity every week is more valuable than one that orders sporadically, even if they spend the same total annually. Lucas: Right. The analyst built a 'recurrence score' for each account. Accounts with a score above 90 percent had a churn rate of less than 2 percent. That's the kind of metric that makes a CFO comfortable paying a premium. Luna: I want to ask about the cost of building that data asset. Doug paid an analyst maybe 15 to 20 thousand dollars for six months of work. That's a tiny investment relative to the outcome. Lucas: It's almost negligent not to do it. Even if you're not selling, knowing your own density and recurrence can help you run the business better. But when you do sell, it can literally add millions to the price. Luna: Are there other examples where this density data play worked? Lucas: Sure. There's a beverage distributor in the Midwest that sold for 40 times EBITDA after mapping their cooler placements — not just delivery stops, but which stores had which products in which cooler. The buyer was a private equity firm that consolidated distributors and used the data to optimize shelf space. Luna: So the principle generalizes beyond baked goods. If you have a B2B delivery business with recurring customers, you likely have valuable data sitting in your delivery receipts. Lucas: Or in your driver's head. But the trick is to extract it and structure it before you go to market. A buyer will discount a seller's claims about 'loyal customers' unless they see hard data. Luna: One risk I see: what if the buyer uses the earnout to squeeze the seller? For example, they could change the route schedule or pricing in a way that makes it harder to hit the targets. Lucas: That's why the earnout contract needs to be specific. Doug's contract specified that the buyer would maintain the same pricing model and delivery frequency for existing accounts during the earnout period. Any changes required mutual consent. Luna: Good. And the target accounts were defined by name and address, not just by zip code. Lucas: Exactly. Zip codes can shift. But a specific list of 35 target cafes was attached to the agreement. If a cafe closed, they'd replace it with a comparable one nearby, also listed. Luna: That level of detail is rare. Most earnouts I've seen are much vaguer. Lucas: And that's why so many earnouts lead to disputes. Doug's broker specialized in data-driven deals and insisted on that specificity. It cost a few extra thousand in legal fees but saved millions in potential litigation. Luna: So the takeaway for a small business owner: start collecting and structuring your customer data now. Lucas: And think about what data would be most valuable to a strategic buyer. Is it density? Recurrence? Lifetime value? If you can show that a buyer could use your data to expand profitably, you're not selling a business — you're selling a map to a goldmine. Luna: And the baker gets 65 times EBITDA. Lucas: Right. Doug is now retired, but he still consults for Artisan Bread Group on route expansion. He says it's more fun than baking. Luna: I bet it is. Alright, thanks for unpacking that one. Very concrete. Lucas: Thanks, Luna. That's all for this episode of The Buyer & Seller Podcast.