Why Customer Concentration Isn't Always the Red Flag Diligence Treats It As

A company walks into a diligence meeting with 68 percent of revenue tied to two customers, and the room reacts before anyone asks a follow-up question. That reaction has a well documented basis. Above 30 percent from a single customer, valuation can drop 20 to 35 percent against a diversified peer, and many institutional buyers decline the deal outright rather than price the risk at all. The heuristic exists for a reason. It is also, in a specific and common category of company, measuring the wrong thing entirely.

The rule assumes a market that most applied deep tech companies do not have

Concentration thresholds were built for businesses selling into markets with a large, roughly interchangeable pool of buyers, where a company at 40 percent concentration by definition failed to diversify into an available pool it could have reached. That assumption collapses in categories with a genuinely small buyer universe. A clinical device selling into hospital systems, an industrial sensor platform selling into a handful of manufacturing primes, or a security platform selling into regulated enterprise IT departments is not choosing to under diversify. It is operating in a market where the total addressable list of sophisticated buyers might be a few dozen names, and landing three of them can already represent a meaningful share of the realistic universe, not a failure to sell more broadly.

The number that actually matters is switching cost, not customer count

The standard diligence questions still apply, but they need a different second question layered on top. Buyers will model the financial impact of losing a top client and price that probability into the deal, which is the right instinct. What that model misses if applied uniformly is whether the relationship is a contract that renews out of habit or a workflow the customer has actually built its own operations around. A clinical workflow embedded into a hospital's standard of care, or a sensor platform wired into a factory's control systems, carries a switching cost the buyer has to overcome, not just the seller. That distinction, whether concentration reflects dependency on convenience or dependency on integration, is invisible in a single percentage and decisive in what it actually predicts about revenue durability.

The failure mode is real, and it looks similar from the outside

None of this is an argument that concentration is never a problem. A company at 85 percent revenue from one customer because the founder has a personal relationship that has never been formalized into a real contract, with no evidence anyone else in the category would buy the product, is exactly the fragile version the standard framework is designed to catch. The tell is usually renewal terms and pricing power. A concentrated customer paying full price on multi year terms with real integration depth is a different risk profile than a concentrated customer on a month to month arrangement being carried at a discount to keep the logo. Both show up as the same number. Only one of them should worry a diligence team.

How we actually diligence a concentrated customer base

Applied deep tech in MedTech, IoT, and enterprise B2B runs into this pattern constantly, precisely because these are categories with genuinely narrow buyer pools and long, integration heavy sales cycles. Rather than applying a blanket concentration discount, we look at contract structure, renewal history, and how deeply the product sits inside the customer's own operations, which is where Working Excellence's digital engineering strategy work becomes useful diligence input rather than just portfolio support, since it can surface exactly how embedded a relationship actually is at the systems level, not just the revenue line.

Ask what the concentration is made of before you discount it

A percentage on a diligence memo cannot tell you whether a company failed to diversify or simply landed the hardest accounts in a small category first. Our investment thesis treats concentration as a starting question rather than a conclusion, because the companies worth backing in applied deep tech often look concentrated for the same reason they look defensible: there were never many other buyers who could say yes.