Customer concentration is the most underpriced risk in lower-middle-market SaaS.

Concentration is the risk that a small number of customers can move the business by leaving, renegotiating, or simply behaving badly. In a traditional private equity hold the question is narrow — can the business survive long enough to be sold again in five to seven years. On a permanent-capital holding period the question becomes much wider. We are underwriting the next decade of compounding, not the next refinancing window. A customer who is 18% of revenue today is a customer whose behavior we will be living with through several contract cycles, several economic regimes, and several rounds of their own leadership changes. The math is different and the discipline has to be different.

The three axes of customer concentration

We look at concentration on three axes, and we never let any one of them carry the analysis. Each axis answers a different question. Two green lights and one red light is not a passing grade; it is a prompt to look harder.

1. Revenue share

The obvious one. What share of trailing-twelve-month revenue is accounted for by the top one, top three, and top ten customers. We also look at the share of net new ARR over the last eight quarters, which is usually more concentrated than the cumulative book. A business with a 22% top-customer share whose net new ARR is 60% coming from that same logo is in a very different position than a business with the same headline percentage whose net new ARR is broadly distributed.

2. Contractual entrenchment

How hard is it for a top customer to actually leave. Multi-year terms, auto-renewal language, switching costs, embedded integrations, regulatory or compliance dependencies, and the presence or absence of a real procurement process at the customer all matter here. We have seen 30%+ top-customer concentration that we underwrote happily because the customer was three years into a five-year term, had built seventeen internal integrations against the product, and operated inside a regulated workflow that made vendor change a board-level decision. We have also seen 9% concentration we walked away from because the contract was month-to-month and the buyer was a procurement-led organization that re-bid annually on principle.

3. Narrative dependence

This is the axis most diligence misses. Narrative dependence is the degree to which the top customers also dictate the product roadmap, the pricing model, and the company's internal sense of who it serves. Concentration in the customer base is one thing. Concentration in the imagination of the company is another. When the founder cannot describe the next twelve months of product work without naming a top account in every sentence, the business is narratively dependent even if the revenue table looks tolerable. That dependence is what makes a customer departure existential, not the percentage.

A customer is a concentration risk when their departure would change what the company is, not just what its revenue line says.

What the three axes produce together

We score each axis on a simple three-band scale — manageable, watch, or disqualifying — and we underwrite the combination, not the average. A manageable-manageable-manageable business is rare and we pay attention when we see one. A manageable-manageable-disqualifying business is almost always a pass, no matter how good the rest of the deal looks, because narrative dependence cannot be fixed by us — it has to be unwound by the operator over years, and we cannot underwrite a multi-year change-of-character program inside the first twelve months of ownership.

Two combinations come up often and are worth naming. The first is a watch-manageable-manageable business — high revenue share but deeply contractually entrenched and roadmap-independent. We have underwritten this profile repeatedly and the businesses have performed. The second is a manageable-watch-watch business — modest revenue share, weak contractual structure, and creeping narrative dependence. These look better than they are on the deal page, and we have learned to be cautious of them.

The diligence workstream we actually run

The framework above is useless without the work. Our concentration workstream is one of the six tracks inside the broader Cobalt Glacier diligence process, and it has the following components.

• Top-account interviews. We talk to at least the top five accounts directly, with the founder in the room when appropriate and without when not. The interviews are not reference calls. We are looking for the customer's own sense of how dependent they are on the product, what the alternative looks like in their head, and what would have to be true for them to consider switching.
• Contract teardown. Every top-ten contract is read by counsel and abstracted into a single spreadsheet — term, renewal mechanics, price escalators, termination rights, assignment clauses, and any side letters. We are looking for the gap between what the company believes the contracts say and what the contracts actually say.
• Usage telemetry. We pull the last twenty-four months of product usage data for the top accounts and compare it to the rest of the book. Stable or growing usage in the top accounts is one of the strongest leading indicators of renewal; quietly declining usage is the strongest leading indicator of a contract event we have not yet been told about.
• Roadmap audit. We sit with the product team and walk the last six and next six quarters of the roadmap. We are asking, without prejudice, how many items on each list exist because a top account asked for them. The answer tells us most of what we need to know about narrative dependence.

What we do once we own the business

Concentration that survived diligence does not stop being concentration after close. It becomes an operating posture. The first hundred days after a Cobalt Glacier acquisition — described in detail in The first 100 days — almost always include a deliberate diversification motion if the top-account share is above 20%. That does not mean firing customers or capping their growth. It means investing the marketing and product effort needed to make the second and third deciles of the customer base grow faster than the first. Concentration falls as a ratio when the denominator grows.

The other operating move is contractual. We renegotiate the top contracts proactively in the first year of ownership, not because we are looking to extract price but because the contracts are usually under-engineered. Multi-year terms, sensible price escalators, real termination notice, and explicit assignment language are good for both sides, and they are easier to put in place when the customer is happy and the relationship is fresh than when something has gone sideways.

Why this matters more on a permanent-capital clock

On a five-year private equity hold, the worst version of a concentration risk is a contract that expires inside the exit window. On a twenty-five-year permanent-capital hold, the worst version of a concentration risk is a customer relationship that slowly turns the company into the customer's internal team. The latter is harder to see, harder to underwrite, and almost impossible to reverse once it has set. It is the reason we put narrative dependence on equal footing with the revenue table and the contract stack. Time exposes risks that short holds never have to see. The discipline is to see them before they exist.

If you are a founder thinking through your own concentration picture in advance of a conversation, we are easy to reach. If you are an LP or investor thinking about how a permanent-capital operator should price concentration, our investor page explains how we think about the broader portfolio shape.