Diagnosing a Well-Loved Feature That Shows No Measurable Business Impact

You launched a feature that users clearly like and say they find helpful, but it is not producing any business impact you can measure. What would you do? ```hint Doubt the zero first "No measurable value" can mean the feature does nothing for the business, or that the way it was measured could not see what it does. Work out which one you are facing before you decide whether to invest in it, reshape it or retire it. ``` ### Clarifying Questions - What shows that users like it: repeat usage, ratings and survey scores, or mostly qualitative feedback? - Which business outcome was it expected to move, such as daily active users, retention, paid subscriptions or ad revenue, and was that target set before launch? - How has impact been measured so far: against a randomized holdout, or by comparing metrics before and after launch, and over how long? - What does the feature cost to run and maintain? ### What a Strong Answer Covers - An audit of the measurement: whether the metric, time horizon, comparison group and sample could have detected an effect at all. - A structured diagnosis of why value to users is not reaching the business, such as which segment uses it, cannibalization, or no link to the core loop or the paid offer. - Concrete options for connecting the feature to business outcomes, each paired with a test that would confirm or reject it. - A time-boxed decision rule for investing, reshaping, keeping or retiring the feature that weighs maintenance cost against user goodwill. - How the decision is explained to stakeholders and, if the feature changes or goes away, to the users who rely on it. ### Follow-up Questions - Leadership proposes moving the feature into the paid tier. How would you test that without spending the goodwill it has earned? - A holdout test shows a small retention lift that never reaches statistical significance. What do you recommend? - You decide to retire it. How do you sunset a feature people genuinely like?