### Prompt A travel-oriented social app is considering a new **Groups** feature that lets people who do not already know one another form communities around destinations and interests. The existing app already has individual discovery, messaging, and trip-planning surfaces. Product leadership asks you to define how the team should decide whether Groups is successful and whether it creates incremental value rather than moving activity away from existing surfaces. ### Constraints & Assumptions - The feature has not launched, so historical group behavior does not exist. - Success must include user value and sustainable product health, not only feature clicks. - Existing discovery, messaging, and planning surfaces may be substitutes or complements. - Groups can create network effects: one person's treatment can affect other people's experience. - You may request additional instrumentation, but you must identify it rather than assume it exists. - Do not assume a particular revenue model. ### Clarifying Questions to Ask - What user problem is Groups intended to solve, and which user segment is the first target? - What actions are possible inside a group, and what makes a group experience meaningfully successful? - Is the immediate decision concept validation, limited launch, or broad rollout? - Which existing surfaces are most likely to lose traffic, and are they strategically important? - Can users participate across several groups, and how are invitations or recommendations delivered? - Are there trust, safety, privacy, or moderation constraints that could limit the launch? ### Part 1: Build the Metric Hierarchy Define one north-star outcome, diagnostic metrics along the activation-to-retention journey, and guardrails. Explain the unit, denominator, measurement window, and segmentation for each important metric. #### Hints Begin with the user problem and the behavior that would demonstrate it was solved. Distinguish feature adoption from durable value. #### What This Part Should Cover ```premium-lock What This Part Should Cover ``` ### Part 2: Measure Incrementality and Cannibalization Describe an experiment that estimates the incremental effect of Groups. Show how you would identify substitution from existing product surfaces, complementary effects, and net product impact. Address spillovers between treated and untreated users. #### Hints Track where displaced activity came from and whether the total user outcome changed. Choose a randomization unit that matches how group membership spreads. #### What This Part Should Cover ```premium-lock What This Part Should Cover ``` ### Part 3: Set a Decision Rule Explain what result would justify launch, iteration, or stopping. Include how you would handle a positive feature metric paired with a negative product-level or safety outcome. #### Hints State the trade-offs before seeing results. A launch decision should be able to reject an apparently successful feature. #### What This Part Should Cover ```premium-lock What This Part Should Cover ``` ### What a Strong Answer Covers ```premium-lock What a Strong Answer Covers ``` ### Follow-up Questions 1. When can cannibalization be acceptable or even desirable? 2. How would you measure success if most groups are small and outcomes are highly skewed? 3. What would you do if user-level randomization is operationally easy but interference is substantial? 4. How would your framework change if Groups were a standalone app instead of a feature? 5. What early signals would you use before long-term retention matures?