Strategy Needs a Falsifier

Most strategies contain targets.

Few contain stopping conditions.

The plan says what success should look like, which metrics will be monitored, and when the next review will occur. It does not state what result would invalidate the diagnosis, which assumption would collapse the logic, or how long leadership is willing to fund a thesis that keeps asking for more time.

Optimism receives a roadmap.

Disconfirmation receives a meeting invite.

Targets do not test the story

A target tells the organization what it hopes to achieve. A falsifier tests whether the governing interpretation remains credible.

Suppose the strategy says awareness is the barrier. The campaign reaches the intended audience, aided awareness rises, and adoption does not move. The target may have been met. The diagnosis may have failed.

Suppose a premium repositioning lifts average order value while reducing repeat purchase and customer referrals. The revenue target may hold temporarily while the intended brand relationship deteriorates.

Suppose a new product attracts trial only under heavy discount. Unit volume can rise while evidence for the value proposition weakens.

Without a falsifier, every result can be narrated as progress.

People escalate when identity enters the investment

Staw’s classic research on escalation of commitment showed that decision makers can allocate more resources to a failing course when they feel responsible for the initial choice.[1] Arkes and Blumer’s work on the sunk-cost effect demonstrated how prior investment can distort continued commitment.[2]

Organizations add more than money. They add status, public claims, executive sponsorship, team identity, and narrative momentum.

The strategy becomes harder to question precisely when the need for challenge increases.

A falsifier agreed before the outcome protects the organization from the person it becomes after investment.

A falsifier is not one bad week

Poorly designed stopping rules create the opposite problem. Teams abandon sound strategies because early noise is interpreted as proof.

A useful falsifier specifies:

• the claim being tested;
• the metric or observable condition;
• the threshold;
• the time window;
• the minimum execution standard required before interpretation;
• the action triggered if the threshold is crossed.

The point is not to automate retreat. It is to prevent the organization from moving the goalposts without admitting it.

Premortems make failure discussable before it becomes political

Gary Klein’s premortem technique asks a team to imagine that a plan has failed and explain why.[3] The exercise legitimizes dissent by moving the room into a future where failure is already assumed.

A premortem identifies possible failure modes.

A falsifier turns the most consequential ones into governance.

Together, they convert skepticism from attitude into evidence design.

Refusal is part of authorship

Strategy is often described as choice. It is equally a practice of refusal: refusing to fund a story after the evidence changes, refusing to call an unknown a strength, refusing to let the executive sponsor become the metric.

This is where human judgment remains visible.

The model can surface deterioration. A person still has to accept what it means.