Skip to main content
Next Step PM
Back to the library

New to product: Practise · Core

Decide how you would measure success and test your idea

A good PM says in advance how they will know an idea worked, and runs the cheapest test that could prove them wrong.

About 1-2 daysEvergreenLast reviewed 9 November 2026

Who this applies to

Skills it builds
Data and measurementDiscovery and research
PM types
All
Product types
All
Starting points
Exploring product management
Ways of working
All
Experience
Beginner (under 1 year)
Size of organisation
All

Steps

  1. Write your hypothesis in this shape: 'We believe that doing X for Y will lead to Z. We will know when we see...'
  2. Choose one main measure and one or two guardrails, which are things that must not get worse. For a club sign-up form, the main measure might be the share of people who finish signing up, and a guardrail might be how long it takes.
  3. Name the riskiest assumption in your idea, then design the cheapest test of it: a paper prototype, a fake front door, a conversation or a mock-up.
  4. Decide before you run it which result would make you carry on, change direction or stop.
  5. Run the test with up to five people and write down what happened.
  6. Record the result against your decision rule and say what you would do next.

Tool examples

Miro
Map your assumptions by importance and how much evidence you have, then pick the riskiest one to test.

Definition of done

  • A hypothesis written down
  • A main measure and guardrails chosen
  • A decision rule written before the test
  • A test run and its result recorded

Common pitfalls

  • Choosing the measure after seeing the results
  • Testing the idea you like instead of the assumption that worries you
  • Treating a small test as proof

Do these first

Sources