Meta·Data Scientist·Onsite - Product Sense / Strategy
May 2026
This was a product/DS case round at Meta for a Data Scientist role, centered entirely on one big open-ended case: should Messenger add group calls, and how would you build, measure, and experiment around that. Dense, multi-part, and the kind of question where you can feel yourself running out of time halfway through.
- How would you determine whether users actually need group calls in Messenger, given that alternatives like group chat, voice notes, 1:1 calls, and third-party apps already exist?
- How would you decide what maximum group size to support for group calls, and what tradeoffs does that decision involve?
- Design a metrics framework for the group calls feature: what would you use as your primary success metric, diagnostic metrics, and guardrail metrics?
- Group members are socially connected, so a standard A/B test will have spillover between treatment and control. How would you design a cluster-based experiment to handle this, and how would you analyze it correctly?
- Lay out a full A/B test plan for launching group calls: unit of randomization, eligibility criteria, rollout approach, and duration. What are the main risks that could invalidate the results?
“I started with behavioral data, things like how often users are doing back-to-back 1:1 calls with members of the same group within a short window, which is a pretty decent proxy for 'I wish I could just call everyone at once.' Also looked at what the survey signals might say about unmet needs.” The rest of the author's notes on Data Scientist interview at Meta, Onsite - Product Sense / Strategy round, covers how they worked through the question, what the panel pushed back on, and what they would do differently.
View Post