Skip to main content
EXPERIMENT-DESIGN5 MIN READ

Pick One Primary Metric Before You Look

Choose one primary decision metric and separate it from guardrails and diagnostics before an experiment launches.

Declare the score before the game starts. Primary metric This is the decision metric. It should match the behavior the treatment is supposed to change and the business question the team is trying to answer. Guardrails Guardrails protect against false wins. A treatment that lifts conversion while spiking refunds or support load may still be a bad ship decision. Diagnostics Diagnostics explain movement. They are useful, but they do not get to overrule the primary metric after the fact.

Read the full lesson

Sign up free — one personalized lesson every day, matched to your role and goals.

Already have an account? Sign in

← Back to library
Contact us