What Must Be True for This to Work?

Strategy offers no certainty. It is a reasoned hypothesis about how a company can win.

A leadership team picks 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. Every part of that belief rests on assumptions.

Customers have to value something and be willing to pay for it. The company has to deliver it reliably, competitors have to be unable to match it quickly, and the economics have to support the investment. The technology has to work, and the organization has to be capable of changing.

Left unspoken, those assumptions let a strategy look sound right up until it fails.

Turn assumptions into tests

Asking "what must be true?" is one of the most useful habits in strategy.

Say a company intends to win with a high-touch, expert-led service at a premium price. Several things have to hold. Customers must value human expertise enough to pay more for it. That expertise must stay hard to replicate. Competitors must not be able to offer an acceptable machine-generated alternative for less. And the company must have the people, process and capacity to deliver a consistently better experience.

None of that argues for abandoning the strategy. It argues for testing it.

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

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

AI changes the assumptions faster

AI does not hit every industry equally, but it speeds up change in the assumptions underneath a strategy.

A service built on expensive expert analysis comes under cost pressure. A product that relied on a slow, manual workflow meets a faster alternative running on AI. Customers who once accepted a two-day turnaround start expecting an immediate, personalized answer. A capability that was scarce last year becomes available to anyone through AI tools, agents or embedded software.

That is why an annual strategy review is often too slow. Whether the strategy document is current matters less than whether the assumptions behind it 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 our ability to build and run the capabilities this requires?

Then give each assumption an owner, an indicator you can observe, and a date to review it.

That is strategic learning, and it takes less paperwork than it sounds.

The companies that adapt well rarely predict every change. They state their bets clearly, notice when conditions shift, and adjust before their assumptions become liabilities.