artifactStrategy & discovery

Product Hypothesis

A testable statement about how a product change is expected to create a measurable user or business result.

Download Markdown (.md)

01What it means

A product hypothesis connects a proposed change or intervention to a target audience, expected behavior, and measurable result under stated conditions. It is a structured belief to investigate, not a prediction guaranteed by confidence or authority.[1][2]

A testable belief connecting intervention to outcome A proposed product change is expected to alter behavior for a defined audience
Proposed product change under stated conditions
ExposureDefined audience encounters and can use the change
OutcomeExpected behavior and measurable user or business result

Test the mechanism, magnitude, uncertainty, and unintended effects

The hypothesis is not validated by confidence or a moved metric alone; design a credible comparison and state conditions, mechanism, threshold, and guardrails.

02Why it matters

Hypotheses turn vague ideas into learning questions and make it possible to stop, adapt, or scale work based on evidence.

03When to use it

Use one when a team is choosing among solutions or needs to learn before scaling. Keep exploratory hypotheses provisional and avoid pretending every long-term strategy can be reduced to one experiment.

04How it works

  1. Name the audience and context.
  2. State the change being considered.
  3. Predict the behavior or outcome.
  4. Choose a measurable signal and threshold where possible.
  5. Test, interpret, and revise the belief.

05A practical example

Example

We believe showing delivery dates before checkout for first-time buyers will increase completed purchases without increasing cancellations; the team prototypes it and compares both measures.

06Where it breaks

A result may be confounded, delayed, or invisible in a small test; a hypothesis can be technically supported while the overall product remains undesirable or unviable.

Common mistakes

  • Writing an unfalsifiable aspiration.
  • Measuring clicks instead of the intended outcome.
  • Treating one experiment as universal proof.

07Sources

  1. 1. Understand users and their needsGOV.UK Service Manual · Government service standard1. Understand users and their needs
  2. The Lean StartupHarvard Business Review · Authoritative business publicationThe Lean Startup
View saved topics

Start typing to search titles, definitions, categories, and frameworks.