Turning a vague onboarding goal into research questions

WHAT THIS TESTS This question probes whether you can partner with research rather than treat it as a service desk. The interviewer wants to see you decompose ambiguity into measurable, prioritized questions and choose appropriate methods, not just nod and build.

A GOOD ANSWER COVERS Start by aligning on what 'improve' means in numbers: is success higher activation, faster time-to-value, fewer support tickets, or better day-seven retention? Then map the current onboarding funnel and pull analytics to locate where users actually drop off. Each drop-off becomes a sharp question, for example, why do users abandon at the account-verification step. Pair the quantitative 'what' with qualitative 'why' studies. Prioritize questions by impact and feasibility, and agree with the researcher on which method answers each: usability tests for friction, surveys for attitudes, funnel analysis for scale.

COMMON WRONG ANSWERS Proposing a specific solution like 'add a tooltip tour' before understanding the problem. Picking one method for everything. Treating the PM's wording as the real requirement without negotiating measurable success. Ignoring existing analytics and proposing fresh studies for data you already have.

LIKELY FOLLOW-UPS How do you prioritize when you have ten candidate questions and time for two studies. What if quantitative and qualitative findings conflict. How do you avoid leading questions in the research itself.

ONE CONCRETE EXAMPLE Suppose analytics show forty percent of new users never complete profile setup. You and the researcher reframe the goal into a testable question: what blocks users at profile setup. A quick five-participant usability study reveals the password rules are hidden until submission, causing repeated failures and abandonment. That insight is now actionable: surface validation inline. You measure success against the agreed activation metric afterward, closing the loop from vague goal to verified improvement.