Index

Mental Models

Falsifiability

A statement is falsifiable if there exists an observation or experiment that could demonstrate it to be false.

Falsifiability separates testable ideas from unfalsifiable beliefs, keeping reasoning honest and progress measurable.

Core question

What would we need to observe to know this belief is wrong?

Example

A team claims their new onboarding flow will reduce churn. A falsifiable version states: if 30-day retention does not improve by at least 5% within two months, the hypothesis is wrong.

How to apply

  1. 1.Restate the hypothesis in a form that can be proven wrong.
  2. 2.Define specific observable criteria for failure.
  3. 3.Run the test with a commitment to accept the result.
  4. 4.If unfalsifiable, treat the claim as assumption, not knowledge.

Pitfalls

  • ·Redefining success after the fact to avoid falsification.
  • ·Confusing unfalsifiable with unimportant — some ideas are worth holding loosely even without clean tests.
  • ·Setting falsification criteria so narrow that noise can trigger false negatives.

Questions

Why does falsifiability matter in business?

It prevents teams from running on unexamined assumptions. If a strategy cannot be proven wrong, you will never know to change course.

Are all valuable ideas falsifiable?

No. Values, visions, and some heuristics resist clean falsification. But operational hypotheses and predictions should be testable.

Related