POINT OF VIEW · 8 MIN READ
Value Creation Isn't an Exit Problem
The systems that run the company are the systems that make it sellable.
A high-performing company in your portfolio goes to market. By the time it reaches a process, real value has been built, and your fund built it. Add-on acquisitions folded in, the leadership bench upgraded from founder-era to professional management, new service lines stood up, payer contracts renegotiated, margins expanded, and a referral engine that did not exist at close. That value is real and it was earned over the hold.
Then diligence starts.
The requests come from every direction at once. The same question arrives three ways, and because no one wrote down a canonical answer, it gets three answers. The clinics count a visit when it happens, finance counts it when it bills. The buyer stops trusting both.
Then the value-creation story stops surviving contact with the data. The margin gain you attribute to the operating plan cannot be cleanly tied to it in the numbers. The synergies the add-ons were supposed to produce cannot be separated from everything else that moved. The growth you are selling as durable cannot be cohorted to prove it will hold.
Add-backs the company knows are legitimate get disallowed because they were never documented. Every disallowed dollar is real earnings the transaction multiple no longer applies to. The number that opened the process is not the number that closes it.
By every measure the company runs itself on, it is performing. The deficiency does not exist in the day-to-day. It appears for the first time under a legibility test the company has never had to take. And at the worst possible moment: when the valuation is being set. The value was real and it was earned. It just was not legible, so it gets only partial credit.
Why even good companies fail the test
A well-run company fails this for two reasons, and neither is about how well it is run.
The first is that the diligence bar is a different kind of test. The demands of running a company are vast and relentless, so management runs by exception: if it is not on fire, it is probably fine. That could "work" for years.
But diligence asks something the day-to-day never does, whether a stranger with capital can verify every claim from documentation. Day-to-day operating is graded on outcomes; diligence is graded on evidence.
Think about what clears the first bar and not the second. The metric leadership trusts because it has always been roughly right, though no one can reconstruct exactly how it is built. The forecast that lands every quarter because one person quietly trues it up, and no one else could rebuild how. The customer everyone knows is at risk, a fact that lives in hallway conversation and never in a document. None of it is a problem in the day-to-day, because the people who hold the knowledge are in the room.
Diligence is the first test that stops accepting "trust me," and keeps asking until you actually answer. Every answer now has to reconcile to a document. What lived only in someone's head has nothing to reconcile against, so it gets discounted.
Across roughly twenty transactions at a multi-site physician platform, I sat on the buyer's side of the table. The pattern repeated: the performance was usually real, and the proof of it almost never existed until we went looking under deadline. What the seller carried in their head, we had to reconstruct from scratch, and what we could not reconstruct, we would not pay for.
The second reason is harder to see, because you are the reason. The company is perfectly legible to you. You have sat on its board for five years, absorbed its quirks, and know what every number means and why it moved last quarter. That context does the work documentation would otherwise do, so the gap never shows, to you.
None of it lives in the company's systems. It lives in the heads of the people close to it, and you are one of them. At exit the reader changes. A new buyer arrives with none of your context and no relationship on which to extend trust on credit, and everything you understood by familiarity now has to be proven cold. The hold did not build legibility into the company. It built context into its owners, and context and processes held in heads are not a form a buyer can easily audit. What is a known quantity to you is an unknown to the next buyer, and unknowns get discounted.
What the discount costs
For a long stretch this did not matter much, because the multiple did the work. Cheap debt and rising entry-to-exit multiples covered a lot of operating sins, and a buyer paid for the growth story and sorted out the plumbing later.
The multiple has stopped doing that work, and illegibility now shows up directly in the price. Sponsors themselves put the gap between a company that is continuously ready to sell, audit-ready year-round with nothing to scramble for, and one that scrambles at the end at one to three turns of EBITDA.
On a business doing twenty million of EBITDA at a ten times multiple, a single turn is twenty million dollars of enterprise value. The company performed; the question was whether it could prove it. When the market was lifting every multiple, you could lose that turn and not feel it. You feel it now.
That gap is part of the bid-ask spread. Most of a spread is the market's: financing, competition, the buyer's appetite. The legibility gap is the part you control, the value you left unproven. You close it during the hold, not in the data room.
The Legibility Stack
A buyer pays for performance, and performance is the asset you spent the hold building. But they only pay full price for performance they can validate. Performance is the headline they are shown, the asking price the teaser implies. Underneath it sit three legibility layers, the depth diligence digs through to test the headline, and the deeper a buyer can verify, the more of that price survives. A buyer pays full price only when all three hold. Each comes with a test you can run on any company in the portfolio today.
Is it understood? This is narrative, the story a buyer is told first and the easiest to dress up, which is why the test is built to get underneath the polish. A buyer cannot see inside management's head, so they test understanding the only way they can, through explanation: can the team say why the performance happened, what actually drives the business, and where the risk is, coherently, and the same way whether you ask the CEO or the CFO. A team that can explain it cleanly understands the machine well enough to keep moving it. A team that tells three different stories gets read as having gotten lucky, and the company is priced as if the growth were an accident.
The two-rooms test: ask the CEO and the CFO separately why the business grew and what threatens it. Same answer, same numbers?
Is it real? This is evidence. Can every claim that matters be proven from the data room, tied to the general ledger, reconciled to itself, without someone in the room to explain it. Most value-creation work fails here, not because it did not happen but because it was never documented as it happened, so at exit it has to be reconstructed under time pressure and a skeptical eye.
The tie-out test: pick the three numbers your thesis rests on. Can each be produced from a canonical source, the same way, by more than one person, and tied back to its system of record, today?
Is it durable? This is process, the deepest layer because it is the hardest to verify and the most expensive to get wrong. Real, documented numbers can still be the product of heroics: a sprint into the sale, a few people doing extraordinary things that do not repeat. A buyer is underwriting the years after they own it, and they cannot confirm before close whether the performance is a durable machine or a one-time effort dressed up for the process, so durability is the value they most often hold back. The first thing I check on any company is whether a number survives the person who produces it walking out the door. A data room can be assembled without the processes behind it, and a good buyer feels the difference between evidence generated by how the company runs and evidence manufactured for the sale.
The key-person test: for the performance you are proudest of, is it the output of a documented, repeatable system, or of a few people you cannot afford to lose?
You can build a genuinely better company and still be marked down on every layer, and the discount is just the sum of what a buyer could not verify, could not trust to last, and could not get a straight answer about.
Run the three tests across the portfolio. Where the answer is no, you have found legibility debt, and it comes due in diligence as a discount. Legibility is cheap to build over the hold. It is brutal to retrofit in the data room. Two or three years from a window, there is runway to build the systems instead of scramble; for a company already in a process, it is triage: find the debt before the buyer does, fix the gaps that move the most value first, and frame what you cannot fix so it reads as known rather than hidden. Either way, you stop walking into the haircut blind.
The work is the same work
Value creation isn't an exit problem. The company that is easy to sell is the one whose performance is real, durable, and understood by the people running it. That is built through the hold or not at all. You can do the building and the making-legible as one job, on purpose, across the years you own the company, or you can discover once the data request list arrives and post-LOI diligence starts that you did only half of it, and let the buyer keep the value you could not prove. The company that did both is the only one that gets paid for everything it built.
MARTEL CAMPBELL
Replies welcome: martel@martelhealth.com
© 2026 Martel Health