Defining a North-Star and Guardrail Metrics for WhatsApp's Poll Feature

WhatsApp lets people create polls inside their chats: someone asks a question with a few options, and others vote. You are the product manager for polls. Explain how you would set goals for the feature and how you would measure whether it is succeeding, walking the interviewer through your reasoning. ```hint Start from a group that needs to decide Picture the chat moment that leads someone to create a poll, and ask what result would mean the poll actually helped that group, rather than just that people tapped on it. ``` ### Clarifying Questions - Is polls a recently launched feature that still needs adoption, or an established one whose goals we are revisiting? - Do the goals cover polls in group chats only, or also one-to-one chats and channels where an admin polls followers? - What should polls do for WhatsApp as a whole: make groups more useful for coordinating, keep groups active, or something else? - Because chats are end-to-end encrypted, which signals can we measure: counts of events such as a poll created or a vote cast, but not the poll's question or options? ### What a Strong Answer Covers - Why polls matter to WhatsApp and the user problem they solve when a group tries to make a decision. - The creator's and the voter's journeys, and where in them value can be measured. - A primary success metric that shows polls doing their job, supported by adoption, participation and repeat-use metrics. - Guardrails on spam, notification fatigue and group health, and how end-to-end encryption limits measurement. - How the goals shift as the feature matures, and how to tell whether polls cause the outcomes measured. ### Follow-up Questions - The number of polls created is rising, but the share of members who vote in each poll is falling. How do you interpret this? - How would you test whether polls make groups healthier rather than just adding activity? - Would you set different goals for polls in channels than in group chats? Why?