Defining and Measuring Project Success: Goals, Primary Metrics and Guardrails

How do you judge whether a project you worked on was a success? Explain how you define success and how you measure it. ```hint Decide it up front Success is easiest to judge when it was written down before the work began. Think about what you would agree at kickoff, with whom, and what would count as harm elsewhere. ``` ### Clarifying Questions - Should I describe my general approach, or anchor it in one specific project I have worked on? - Is the project a customer-facing product launch, or could it also be an internal or platform project with no external users? - Should success include how the project was run (date, scope, cost, team health), or only the outcome it produced? ### What a Strong Answer Covers - Tying success to the user and business goal the project serves, agreed with stakeholders before work starts - A primary metric with a baseline, a target and a time window, plus guardrail metrics that catch harm elsewhere - How the project's impact is separated from noise and other changes, for example through an experiment or a credible comparison - The difference between outcome, delivery and learning as separate measures of a project - How mixed or disappointing results are read and turned into a decision to scale, iterate or stop ### Follow-up Questions - How would you measure success for a project whose impact takes months to appear, such as a retention initiative? - Your primary metric hit its target but a guardrail regressed. Was the project a success? - How would you measure an internal tooling project whose users are other engineers?