What Must Be True for This to Work?

Strategy is not certainty. It is a reasoned hypothesis about how a company can win.

A leadership team chooses a customer segment, a value proposition, a business model, an operating approach, and a set of capabilities because it believes those choices will create an advantage. But every part of that belief depends on assumptions.

Customers must value something. They must be willing to pay for it. The company must be able to deliver it reliably. Competitors must not be able to match it quickly. The economics must support the investment. The technology must work. The organization must be capable of changing.

When those assumptions remain unspoken, strategy can look sound right up until it fails.

Turn assumptions into tests

The question “What must be true?” is one of the most useful disciplines in strategy.

Suppose a company’s strategy is to win by offering a high-touch, expert-led service at a premium price. Several things must be true: customers must value human expertise enough to pay more; that expertise must remain difficult to replicate; competitors must not offer an acceptable machine-generated alternative at a lower cost; and the company must have the talent, process, and capacity to deliver a consistently superior experience.

That is not a reason to abandon the strategy. It is a reason to test it.

What evidence would confirm or challenge each assumption? What customer behavior should the company monitor? Which competitive moves matter? What changes in AI capability, cost, regulation, data access, or buyer expectations would invalidate the original logic?

A strategy becomes stronger when leaders can identify the signals that would tell them they are wrong.

AI changes the assumptions faster

AI does not affect every industry equally, but it can accelerate change in the assumptions beneath a strategy.

A service that depended on expensive expert analysis may face cost pressure. A product that relied on a slow, manual workflow may be disrupted by a faster alternative that runs on AI. A customer segment that once accepted delayed responses may begin to expect immediate, personalized support. A capability that was scarce last year may become widely accessible through AI tools, agents, or embedded software.

This is why annual strategy reviews are often insufficient. The relevant question is not whether the strategy document is still current. It is whether the assumptions behind the strategy still hold.

Build a learning system

For every major strategic choice, identify a small set of critical assumptions:

• What must be true about customer needs and willingness to pay?
• What must be true about industry cost structures and profit pools?
• What must be true about competitor capabilities and likely responses?
• What must be true about data, technology, regulation, and partner access?
• What must be true about your organization’s ability to build and operate the necessary capabilities?

Then assign each assumption an owner, an observable indicator, and a review cadence.

This is not bureaucracy. It is strategic learning.

The companies that adapt well are not those that predict every change perfectly. They are the ones that state their bets clearly, notice when conditions shift, and adjust before their assumptions become liabilities.