Determine if users need a new feature

## Scenario You are a Data Scientist supporting a consumer product team considering launching a new feature (e.g., a new group-calling/chat feature). You have access to product event logs and can also request additional data collection if needed. ## Tasks 1. **Using the data already available**, how would you identify whether users *need* this feature? - Define what “need” means in measurable terms. - If you use an “active user” concept (e.g., **7-day active users**), state your definition clearly. 2. **Assuming you can collect any data you want**, what additional data would you gather to answer the same question better? 3. How would you **measure whether the feature is successful after launch**? - Propose a metric framework with **primary metric(s)**, **diagnostic metrics**, and **guardrail metrics**. - Explain how you would align metrics with the feature’s goal and handle trade-offs (e.g., engagement vs. retention vs. quality). 4. How would you address the **novelty effect** (short-term spike after launch that may not persist) when interpreting results? ## Output expectations Explain your approach end-to-end: problem framing, analysis plan, metrics, experiment or quasi-experiment design, and key pitfalls (bias/confounding, seasonality, logging gaps, metric gaming).