What the Customer List Tells You That the Quality of Earnings Never Will

The cleanest quality of earnings report I have ever been handed came stapled to one of the shakiest businesses I have ever diligenced. Every adjustment tied out. Revenue recognition was conservative. Working capital had been normalized within an inch of its life. The firm that prepared it had done exactly what it was hired to do, which was to confirm that the earnings were real. Nine percent revenue growth, three years running, fully verified.

The customer file took an afternoon to build and most of a weekend to believe. It told a different story about the same company, and it is the reason I still own my capital instead of that business.

The two documents were never in conflict. A QoE answers a narrow question with precision: did the money actually come in, and is the EBITDA you are being shown the EBITDA that exists? The customer list answers the question a buyer should care about more, which is whether the money comes in again after you own it. Most diligence budgets are spent almost entirely on the first question. The second one costs a data request and two days in a spreadsheet, and it gets skipped on more deals than anyone in this industry would like to admit.

Nine percent growth was hiding twenty percent churn, and the QofE had no reason to mention either number.

Paul W. Swaney III

The Anatomy of Nine Percent

Start with one tab: revenue by customer, every customer, each of the last three years. Sellers will offer a summary of the top ten. Take the raw file instead, sort it, and read it the way you would read a story, from the first year forward.

In the deal I am describing, the top of the file looked exactly like the growth chart in the CIM. The bottom explained where the growth came from. Roughly a third of the customers on the books in year one were gone by year three, and gross churn was running near twenty percent annually. The company grew anyway because the founder was an exceptional salesman who refilled the bucket faster than it drained. Blend the two numbers and you get a smooth nine percent line that a QoE will verify without hesitation, because every dollar of it was real.

The trouble is what happens to that arithmetic on the day the bucket-filler retires, which was the entire premise of the transaction. Buy that company and you inherit the drain without the tap. Churn compounds against you the way interest compounds for you, quietly and then all at once. A business that loses a fifth of its book every year is a sales organization with a service attached, and it should be priced like one.

The Name on the Contract

The second thing the customer list surfaces takes more than a spreadsheet, but the spreadsheet tells you where to dig. For each of the top 10-15 accounts I ask two questions. Who signs the contract? And who actually decides to keep buying? They are rarely the same person, and the distance between them is where deals go to die.

In the lower middle market the honest answer, more often than anyone puts in writing, is that the relationship sits with the seller personally. The contract says the customer is a company. The behavior says the customer is one purchasing manager who has had the founder's cell number for 20 years and has never needed a second one. Renewals happen over dinner. Problems get solved before they become tickets. None of this appears in a contract review, and all of it walks out the door at close, in the passenger seat of the seller's truck.

I learned this on an early deal where the two largest accounts had known the founder longer than his own employees had. Reference calls made it obvious within 10 minutes. When I asked one customer what he would do if the founder ever stepped away, he laughed and said he would probably follow him, and he was only half joking. We restructured around that answer: a longer consulting agreement, an earnout tied to retention of named accounts, and a deliberate eighteen-month program to move every key relationship onto people who would still be there in year five. The revenue held. It held because we priced the risk and then managed it, and we only knew to do either because we had asked the customer list who the customers actually belonged to.

One Price Doing All the Work

The third read is pricing dispersion, and it is the one that separates real pricing power from a lucky contract. Take the same product family and chart what every account pays, adjusted for volume. In a healthy book you see spread with a floor: newer accounts near list price, older accounts drifting somewhat below it, nothing wildly out of line. That spread is evidence the company can charge different customers different prices and make them stick, which is the plainest definition of pricing power I know.

What you are hunting for is the outlier. On more than one deal I have found a single account, usually the oldest, paying 30-40 percent above the rest of the book on a contract negotiated a decade ago that nobody has dared to reopen. Strip that one account out and blended gross margin drops from enviable to ordinary. The QoE reports the blended number as a fact, and it is one. It is also a fact that depends entirely on one counterparty continuing not to notice.

So run the repricing test. If every account moved to the median price tomorrow, what is EBITDA? Sometimes the answer kills the deal on the spot. Sometimes it does the opposite, because a book full of accounts priced below the median is deferred upside for a buyer with the discipline to reprice, and I have underwritten deals where that single lever was worth more than the entire cost program. Either way, you want the answer before wire day, and the only document that provides it is the one the data room did not include until you asked.

The Two Days Nobody Bills For

Everything above costs a data request and two days of work, which may be exactly why it gets skipped. The QoE gets commissioned on every deal because lenders require it, committees expect it, and an entire industry exists to produce it. No lender requires the customer tape. There is no line item for it and no advisor whose letterhead attaches to it, so it sits in the category of work that only happens if the buyer personally insists.

As an sponsor, I insist, and the model is the reason. I raise capital deal by deal, which means every investor who backs me is underwriting my judgment on this specific company, with no portfolio effect to hide behind if the book melts after close. That concentration of accountability changes behavior. It also lets me move at the speed of the question: the customer tape goes into my first data request, the read takes a weekend, and I have killed deals in under two weeks that would have consumed a quarter of committee time inside a fund. Full authority and full responsibility, sitting on the same desk, turns out to be a diligence technology of its own.

The QoE tells you what you are buying. The customer list tells you whether you get to keep it. Read both. Only one of them requires you to ask.

Paul W. Swaney III is the founder of Swaney Group Capital, an independent sponsor focused on acquiring and operating lower middle market businesses. LeverUp® publishes at least weekly. More if I have something extra to say.