How do you define success metrics for a change?

Direct answer

What a strong answer covers

How it works

Success is defined before the change so the measurement cannot be bent to fit the result. Start from the behaviour the change is meant to alter, pick an outcome metric that reflects it and that the change can plausibly move, and record the baseline over a period long enough to include normal variation. Set a target and a time window, name the population it applies to, add guardrail metrics that must not get worse, identify the data source and its owner, and agree in advance what result will lead to which decision. Leading indicators give early signal when the outcome metric takes months to move.

Example

For a redesigned onboarding flow the outcome metric was the share of new accounts completing a first project within 7 days, baseline 31 percent over the previous quarter, target 40 percent within eight weeks of launch, measured on self-serve sign-ups only, with support ticket volume per new account as a guardrail. The decision rule stated that below 34 percent the old flow would be restored, which is what happened after the first attempt.

Common mistakes

• Choosing a metric after seeing which one improved.
• Setting a target with no baseline so any number can be called success.
• Tracking an outcome with no guardrails and celebrating a gain that broke something else.

A useful answer structure

Answer from your own work.

Do not memorise a script: use only examples you can explain honestly in follow-up questions.