How do you diagnose why a new feature's adoption is flat?

WHAT THIS TESTS: Whether you can build a structured diagnostic plan that separates awareness problems from activation and retention problems. Interviewers want to see that you do not treat adoption as a single metric but as a funnel, and that you know how to pair quantitative signals with qualitative context to generate testable hypotheses.

A GOOD ANSWER COVERS: Four layers in order. First, define the adoption funnel stages for this specific feature, such as discovery, first use, activation, and habituation, and pick one leading indicator per stage. Discovery could be page views or tooltip impressions; first use could be clicking the entry point; activation could be completing the core action; habituation could be seven-day retention. Second, segment users by cohort, role, or entry surface to see if the flat KPI hides a bimodal distribution, for example power users adopting while casual users ignore the feature. Third, combine behavioral data with qualitative signals. Quantitative events show where users drop off, while session replays, surveys, or user interviews explain why they drop off. Fourth, rank hypotheses by impact and effort, then validate with a targeted experiment such as an A/B test on onboarding placement or an in-app product tour.

COMMON WRONG ANSWERS: Jumping straight to usability fixes like redesigning the button or adding more tooltips without first checking awareness. Another red flag is proposing a generic analytics dashboard instead of a time-bounded diagnostic plan. Saying you will just interview five users without any behavioral scoping is also weak because it lacks the funnel structure that tells you which users to interview and when.

LIKELY FOLLOW-UPS: The interviewer might ask how you would measure awareness for a feature that has no dedicated UI surface, or how you would distinguish between a feature that users do not need and a feature they need but cannot find. They might also ask how long you would run the diagnostic before deciding to sunset the feature, or how you would design an experiment to prove that better onboarding actually changes activation rates.

ONE CONCRETE EXAMPLE: Imagine a new reporting dashboard that is flat at five percent weekly usage. Your funnel shows that eighty percent of eligible users never open the reports tab, so the barrier is discovery, not usability. You segment and find that users who enter from the home banner have a thirty percent activation rate, while users who do not see the banner have a one percent rate. Session replays of non-adopters show they search for export functionality in the old table view instead. You form the hypothesis that users do not know the dashboard replaces the old export flow. You validate by running an A/B test where the old export button shows a contextual tooltip pointing to the new dashboard; the test group shows a twelve percent lift in first use, confirming the awareness barrier.