A company can be very good at what it does and still be heading in the wrong direction. Strong results, efficient operations and reliable delivery usually signal a well-run business, and they also build confidence in the decisions that produced them. Over time, that confidence makes the business model itself harder to question.
VARTA illustrates the problem well. The German battery manufacturer has more than 130 years of engineering history and is one of relatively few companies able to produce rechargeable button cells. Its CoinPower cells went into Apple’s AirPods, and its Nördlingen plant was built largely around that demand.
When Apple ended the supply agreements, VARTA was exposed. Reports cited cost competition from Chinese manufacturers that German production found increasingly hard to match. Weak demand, currency effects and a complex corporate structure also played a part, so the loss of Apple alone does not explain what followed. It does show how the business had been shaped. A plant designed around one customer can run efficiently for years, and its dedicated capacity, tuned processes and scale are exactly what make it hard to redeploy once that customer leaves.
A strategic bet that looked like a capacity decision
Building a plant around a single customer’s product is a bet on that customer’s future sourcing decisions. Inside most companies, it does not look like one. It arrives as a capacity plan, a capex request or a response to an order forecast, and it is reviewed on the terms those formats invite: payback period, utilization, unit cost. Each of those tests can be passed with distinction while the underlying bet goes unexamined, because none of them asks what happens if the customer stops buying.
Many of a company’s most consequential strategic decisions are taken this way. They enter the organization through an operational channel, they are judged by operational criteria, and their strategic weight only becomes visible when the assumption behind them fails.
Same dependency, different outcomes
GT Advanced Technologies invested heavily in sapphire production after its 2013 agreement with Apple. The expected business did not materialize, and the company filed for bankruptcy protection in October 2014. The investment rested on a commercial opportunity that depended largely on one customer’s product decisions.
Dialog Semiconductor faced a variation of the same problem. Apple accounted for a large share of its revenue, and signs that Apple was developing its own power management chips put Dialog’s future in question. Dialog negotiated a technology and engineering transfer agreement with Apple and worked to broaden its customer base. Renesas acquired the company in 2021 for approximately €4.9 billion.
The three cases differ in circumstance and outcome, and they are not fully comparable. What they share is a dependency that became a strategic risk well before it showed up as an operational problem. Dialog acted while it still had room to move. By the time such a risk reaches the financial results, most of that room is usually gone.
Why strong performance hides the risk
Operational performance is measured in detail. Costs, delivery, quality and utilization are tracked, and when one falls behind target someone is expected to explain why. Strategic assumptions get no comparable scrutiny. A company can depend on one customer for most of its revenue, compete in a market where cost advantage is shifting, or invest in a capability its customer may later bring in-house, and still post excellent results while those exposures grow.
Operational issues demand attention now, while strategic risks are uncertain and tend to mature beyond the current planning cycle. When the numbers are good, challenging the direction feels unnecessary. Operational excellence shows how well the business runs under its current assumptions and says nothing about whether those assumptions still hold. A business can improve its efficiency year after year without becoming any less vulnerable to the decision of a single customer.
Measuring the dependency where it sits
Most companies already track customer concentration in revenue terms. Under IFRS 8, listed companies must disclose when a single customer accounts for 10 percent or more of revenue. That figure is useful, but it describes the income statement. VARTA’s exposure sat largely in its factory.
A more revealing measure is asset concentration: the share of fixed assets, production capacity and specialized capability whose output has only one realistic buyer. Revenue concentration can fall within a year if a company wins new customers. Asset concentration moves far more slowly, because a line built for one product rarely converts cheaply to another. It is the better indicator of how much room the company would have if its largest customer left.
For each significant production asset, management can ask who else could buy what it makes, how long redeployment would take and what it would cost. Where the honest answer is “no one, years, and most of the original investment”, the asset is a strategic position and deserves to be reviewed as one, alongside its utilization and unit cost.
Concentration remains a legitimate choice. A dedicated line can make good commercial sense, and a large customer can provide the scale needed to compete. The point is to make that choice visible and to revisit it while results are strong, when the company has more financial room and more time to act than it will have in a downturn.
Running the business and questioning its direction
Operational excellence matters, but it cannot compensate indefinitely for a business model that is losing relevance or becoming more exposed to forces the company does not control. The difficult part is recognizing when a strength has started to create a weakness. VARTA’s dedicated capacity was a source of efficiency and, at the same time, the reason it had so little flexibility when Apple left.
The companies best placed to handle change understand which dependencies they have chosen, what those choices cost and how they would respond if circumstances shifted. That understanding rarely comes from operational reviews. It has to be built into the way strategic decisions are recognized in the first place, including the ones that arrive looking like something else.