Optionality and the Myth of Strategic Customers

How dependence on a few customers quietly kills resilience.

Every scaleup eventually reaches the same crossroad: double down on a few massive accounts, or cultivate the quiet loyalty of many. The first path feels strategic. The second, humble. Yet history and statistics show that only the humble one endures.

Among the many paradoxes that haunt scaleups, one stands out for how recurrent and destructive it is: the illusion that growth anchored in a handful of major accounts represents stability.
It does not.
It represents dependence. And dependence, in the long run, is indistinguishable from fragility.

Scaleups often inherit the optimism of startups but none of their naivety. They know what it means to survive, to sell, to deliver. Yet as the stakes rise and the pressure from investors intensifies, many organisations fall into a subtle trap: they believe that securing a few “whales” will guarantee predictable growth. In reality, this is how optionality dies.

The asymmetry of risk

As Nassim Nicholas Taleb illustrated in Antifragile, true robustness is built on asymmetry. You can afford to expose 10% of your portfolio to risky bets and keep 90% safe, because the downside is capped, while the upside can be multiplied tenfold. The reverse, however, is fatal: when 90% of your system depends on 10% of your customers, your entire structure becomes one event away from collapse.
It is not risk management. It is risk concentration.

“Wind extinguishes a candle but energises fire.” — Nassim Nicholas Taleb, Antifragile

Optionality is not just a financial principle, but a cultural one. Companies that survive volatility are those that distribute agency, not just revenue. They spread decision-making, learning, and trust horizontally. The more distributed their dependencies, the stronger their resilience when the unexpected happens.

Big accounts distort not only financial resilience but also decision-making. They bend priorities, hijack roadmaps, and skew engineering choices. Entire product lines are re-architected to serve unique needs, while more universal value propositions decay in neglect. The logic is always the same: “if we can keep them, everything else will follow.” But nothing follows. The company turns into a bespoke service provider masquerading as a platform.

The Jaguar lesson

The automotive industry provides a striking analogue.
In the late 1980s, Jaguar conceived the XJ220, a hypercar born from the dream of competing with Ferrari and Porsche on performance and prestige. For a brief year, it even held the title of the fastest production car in the world, and yet, almost no one remembers it today. It was engineered with ambition, designed for the ultra-wealthy, and marketed as a symbol of rebirth for the brand. But the timing was catastrophic. The early-90s recession erased the luxury market the car was meant for. Orders evaporated. Jaguar had designed a masterpiece for a world that no longer existed.

The XJ220 stands as a monument to a structural error: mistaking niche success for scalable value. The same pattern appears in software and platform ecosystems when teams pursue large bespoke deals instead of building generalisable systems. Similar fates befell some of the most iconic names in technology.

Sun Microsystems, once the beating heart of open systems, built its future around a handful of enterprise clients in finance and telecom. When those sectors consolidated and shifted toward cheaper, commodity hardware, Sun’s magnificent engineering culture could not save it. Love for the craft was not enough to offset dependence.

BlackBerry followed a parallel arc. Its devotion to corporate and government clients made it indispensable for a time ... and blind to the consumer revolution that followed. Optimising for elite customers, it lost the mass it needed to remain relevant. By the time it noticed, optionality had already collapsed.

Across decades and domains, the same pattern repeats: excellence turns brittle when it is built to please a few rather than to serve the many.

When optionality collapses

Optionality, in Taleb’s sense, is the freedom to respond to volatility, to adapt, pivot, or seize upside when conditions change. Large accounts, however, tend to freeze optionality. They impose contractual rigidity, technical coupling, and emotional dependence. The organisation starts optimising for their happiness rather than for ecosystem health. Product management becomes reactive. Engineering, defensive. And leadership, addicted to recurring revenue metrics that conceal growing exposure.

Platform teams experience this pressure most acutely. Instead of building leverage, abstractions, automation, self-service capabilities, they end up customising internal tooling for “strategic partners.” Every new integration introduces complexity and erodes coherence. The system loses its purity of intent: to serve all customers equitably and enable everyone to move faster.
What was once a platform becomes a funnel, narrowing with each “special case.”

What begins as structural dependence inevitably expresses itself in culture.

The unseen cost in engineering

The distortion does not stop at strategy. It seeps into engineering culture. Special-case roadmaps produce fragmented priorities and chronic rework. Teams are forced to divert effort from maintainable architectures to quick patches demanded by VIP clients. Technical debt rises not through negligence but through misplaced empathy, the will to please those who shout the loudest. Over time, the organisation stops learning from the median case and focuses on outliers.
That is the opposite of scalability.

Moreover, the social dynamics change. Engineers feel their work is dictated by politics rather than principle. Product managers lose the discipline of generalisation. Even quality assurance starts to drift: instead of testing for universal stability, testing becomes scenario-specific. The company stops building a product and starts managing a portfolio of exceptions.

The human cost of distortion

It takes only one step for these internal frictions to turn toxic. What begins as a natural customer pull soon bends into an unnatural contortion to satisfy a subset of elite clients. Inside the company, this behaviour is perceived as a lack of respect from those customers, a disregard that trickles down and erodes the morale of engineering teams.

Developers find themselves sifting context endlessly, losing their sense of purpose in the process. The organisation then divides: some defend the appeasement strategy, while others suffer cognitive dissonance. Prioritisation drifts away from reasoned market impact towards emotional blackmail and ultimatum.

Fragility disguised as success

Reputation in such an unbalanced environment becomes perilously fragile. It only takes one small but loyal customer, the kind that speaks with authenticity rather than influence, to expose systemic decay and damage trust. In the age of transparency, no major account, however large, can shield a company from the reputational collapse that follows.

When credibility erodes at the base, prestige at the top means nothing.

The cruel irony is that these dynamics often produce flattering short-term results. Revenue increases, customer logos impress investors, and teams appear “busy.” Yet the underlying resilience decays quietly. The moment one of the anchor clients churns, the illusion breaks, revealing that most of the supposed growth was concentrated exposure.
This is not antifragility. It is, as Taleb would say, naïve scaling, mistaking magnitude for diversification.

What starts as hubris ends as exposure.

The alternative: resilience through breadth

Healthy growth does not depend on giant accounts but on a broad and loyal customer base. Mid-tier and small clients may not bring glamour, but they bring stability and truth. They represent the real market, not the one inflated by exceptional circumstances. Serving them well forces the company to standardise, automate, and clarify its value proposition.
It is through them that platforms mature and optionality returns.

In engineering terms, this is the difference between designing for extension and designing for exception. Extension strengthens systems. Exception weakens them. The more a company can scale improvements horizontally, the more antifragile it becomes. Each independent customer becomes a node of resilience, a shock absorber against volatility, a proof of true scalability.

The moral of the theatre

Perhaps IT is a kind of Greek theatre after all, complete with hubris, repetition, and prophecy. The Sisyphus of engineering keeps pushing the boulder of simplicity uphill, while Cassandra, the product strategist, warns that dependence on giants will end in tears. Yet the crowd still believes in the next “strategic deal” or “enterprise transformation.”

The stage changes, the actors age, but the plot remains the same.

True maturity in a scaleup begins when leaders recognise this pattern and choose resilience over vanity, breadth over magnitude, and universality over dependence. In doing so, they exchange the fragile comfort of applause for the enduring discipline of balance, and finally stop mistaking the spotlight for the foundation.

True scale is never about size. It is about how gracefully an organisation can lose one customer without losing itself.