Meta·Data Scientist·Onsite - Product Sense / Strategy
Jun 2026
Meta data scientist interview focused entirely on a location-based restaurant recommendation feature, going deep on data requirements, model validation, and privacy tradeoffs. Seven questions, all connected to the same scenario, which made it feel more like a case study than a traditional interview loop.
- How would you use real-time location data from a nearby friends feature to build a new product feature?
- What's the rationale for building restaurant recommendations, and how might user behavior shift as a result?
- What datasets would you need to power this recommendation system?
- How would you validate whether the recommendation model is actually working?
- After launching the feature, which metrics would you track to measure success?
- What negative consequences could this feature introduce?
- How does building a restaurant recommendation system differ from a people recommendation system like friend suggestions?
“I leaned into live coordinates and activity patterns pretty quickly but forgot to tie in the social graph until they nudged me.” 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