The 25% Cliff

Diane Halloran has been selling for Meridian Climate Systems for 22 years. She closed the Westbrook account in her second year. She remembers the lunch where it happened.

Westbrook Property Services manages 200 commercial buildings across northeast Ohio. They have used Meridian for HVAC service and installation since 2003. They renew the master service agreement every 3 years without going to bid. Diane handles their renewal personally. The current contract is worth $7.3M in annual revenue.

That is 26% of Meridian's $28M revenue base.

Diane has never told Ed it is a problem. Ed has never asked. They both know Westbrook is a big customer. They also both know Westbrook has been with them for 22 years. The relationship is excellent. Nothing about it is going anywhere.

The buyer's commercial diligence team prices it at $275K.

Three legs of a stool

A 4-legged stool tips when one leg breaks. A 3-legged stool tips when one leg wobbles. The fewer the legs, the more weight each leg carries, and the more catastrophic any single failure becomes.

Every founder knows this in their bones for a stool. Most founders do not apply it to their customer base.

A business with 100 customers each contributing 1% of revenue is a 100-legged stool. Lose any one, the stool stands. A business with 1 customer contributing 30% is a stool with 1 leg doing a third of the work. The other legs are decorative.

The founder sees a stable revenue line. The buyer sees the leg.

The 25% Cliff

Buyers tend to price single-customer concentration above 25% as a discrete risk category. Below that line, the customer is a customer. Above it, the customer is a change-of-control event waiting to happen.

Call it the 25% Cliff.

The 25% Cliff exists because of 3 things buyers know that founders rarely think about. The relationship lives in a person, not in a contract. When the founder sells, the relationship is exposed. The new owner has not been to the customer's daughter's wedding.

Change of control gives the customer permission to renegotiate. The procurement team that has wanted to put the contract out to bid for years finally has the political cover to do it.

And concentration risk is asymmetric. The upside is bounded by what the customer is already buying. The downside is unbounded. The customer can leave, demand concessions, or get acquired by someone with a different vendor preference. The buyer prices the downside.

What Ready Looks Like

The 25% Cliff is one of the most preventable items on the deduction list. It is also one of the slowest to fix. The math is straightforward: shrink concentration by adding revenue elsewhere.

A business with 26% concentration that wants to be at 15% has 2 choices. Lose the customer (catastrophic) or grow everything else by a factor that pulls the percentage down (slow). Realistically, the second one takes 18 to 36 months of disciplined business development against a target customer mix.

The other readiness move is documentation. A long-tenured handshake renewal is concentration risk on hard mode. A long-tenured renewal supported by 4 things is concentration risk on easier mode. A written master service agreement. A documented service level history. An executive-sponsor relationship that survives a salesperson change. A clean account-management transition plan. Same revenue dollars. Different deductibility.

Buyers know which version they are looking at within 10 minutes of opening the customer file.

For Meridian, none of these were true.

The Ed Moment

The buyer's commercial associate sends a 1-paragraph note to the deal lead at the end of week 2 of diligence. Customer concentration in the Westbrook account at 26% of revenue. Master agreement is in place but renews on 3-year cycles without competitive review. Relationship held by long-tenured account executive (Halloran). Recommending $275K reduction to cover retention risk over the earnout window.

Ed sees the line item in a deal-points memo from his attorney summarizing the buyer's revised offer. By now Ed has seen this shape twice already. The QoE Haircut before negotiations even started. The Market You Couldn't See in the first finding off the LOI. This is the second finding. Another cut.

He calls Diane at home that night.

"They're taking us down $275K on Westbrook."

Diane has been at this for 22 years. She knows what the buyer is pricing.

"Yeah," she says. "I would too."

There is a long pause on the phone. Ed had been hoping she would tell him the buyer was overreacting. Westbrook loves them. Diane plays golf with their facilities director 4 times a year. The relationship is real.

Diane says, "The relationship is real. The contract isn't great. And I'm 58."

Ed had not thought about Diane being 58.

What this cost Ed: $275K.

Westbrook generates $7.3M in revenue. At Meridian's margins, that is roughly $1.4M in EBITDA contribution from a single account.

The buyer is not predicting that Westbrook leaves. The buyer is pricing the probability that they might. A 22-year handshake held by a salesperson 7 years from retirement, with no documented transition plan, no competitive renewal history, and a change-of-control clause the customer has never had reason to exercise. Over the earnout window, the buyer models a 20% chance that the Westbrook relationship degrades — Diane retires, procurement insists on a competitive bid, or new ownership triggers a renegotiation.

Twenty percent of $1.4M is $275K.

That is how every line on the deduction list works. The buyer identifies an EBITDA contribution at risk, estimates the probability of impairment, and prices the expected loss. No single finding is a prediction. Every finding is a risk discount to EBITDA. Twenty-one of them is $3.5M.

Don't be Ed.