strategyplaybook.aiSqwyz
Sign in
← All 124 tools
Risk · Choose

Falsification test

Asks what evidence would prove the strategy wrong, and then whether anyone is looking for it. A strategy nothing could falsify is a belief.

The printable canvas and the handoff into Claude Code are part of a paid plan. See what a plan includes. It needs the Claude desktop app on this machine, and your team and the prompt library already installed in that project — we cannot see your disk, so open it there.

Stage
03 Choose
Works at
All levels
Maturity
Any
Time
An hour

What it is

An application of Popper's falsifiability criterion to strategic claims: for each significant assertion, what observation would show it to be false, and would you see it? Claims that cannot be falsified cannot be tested, and should not carry weight in a decision.

The claim
A specific strategic assertion, stated precisely enough to be checked. Precision is most of the work.
The falsifying condition
What would have to be observed for the claim to be wrong. Not what would make it harder — what would establish it as false.
Observability
Whether that evidence would actually reach you, through which channel and who would notice.
Timeframe
By when you would expect to know. Without a date, claims are never tested, merely carried.
Unfalsifiable claims
Assertions nothing could disprove. Legitimate as values, dangerous as strategy, and worth labelling as one or the other.
The mistake people makeThe test only works if someone is genuinely prepared to change their mind. Run as a formality it produces a set of falsifying conditions nobody monitors, which is worse than not running it because it creates the impression of rigour.
What it’s forYou want to know what evidence would prove the strategy wrong, before you commit.
What it’s not forYou are not willing to change your mind. Then it is a formality and everyone can tell.

How you run it

  1. State each major claim preciselyVague claims cannot be falsified, which is part of why they survive. "We will win on service" needs to become something checkable.
  2. Ask what evidence would disprove itNot what would make it harder — what would establish it was wrong. If nothing could, the claim is not doing analytical work.
  3. Check the evidence would be visibleA falsifying condition you would never observe is no better than none. Name where the signal would show up and who would see it.
  4. Set a dateBy when would you expect to know? Open-ended claims never get tested because there is never a moment when they are due.
  5. Flag the unfalsifiable onesClaims nothing could disprove are values or hopes. They may be fine as either — they should not be load-bearing in a strategy.

The prompt

Two ways to run it

Run this tool in your own Claude

The short prompt starts your partner against the library on your disk. The long one carries everything with it and needs nothing installed.

🔒 Prompt locked

Everything above is free — what this tool is, what it is for, what it is not for, and how to run it as a workshop. Three tool prompts a month are free with an account; beyond three, and for the printable canvas, it is the paid part.

See plans — from £19 a month

Your playbook

It lands in the earliest stage this tool suits. Move it on the Playbook page.

You’ll need

  • A stated strategy or a specific claim
  • Genuine openness to being wrong
  • A forum where the answer will be acted on

You’ll end up with

  • A falsifying condition for each major claim
  • How you would detect it, and by when
  • The claims that turn out to be unfalsifiable

Runs well with

Acquisition screen and synergy thesisAI opportunity mapBoard paperBusiness case and driver model