Choose an app you like and we will focus on this for the rest of the interview. Say that users are suddenly dropping off from using the app (they still have an account but stop using it). How would you research the basis for this churn? Then there were additional follow-on technical questions related to steps in looking at different types of data (log data, surveys, integrating qualitative data, sampling/weighting, reaching out to users for feedback when they are no longer using the app, survey development - questions, etc., trade-offs in response analysis choices (e.g., means vs aggregating percent pos/neg, how to prioritize findings, how to present results to stakeholders, what to do if some factor you cannot measure is the reason for churn, and what to do if you need to get some kind of results fast (e.g., like in 2 days and there is no time to conduct a survey). From my interview, Meta heavily focuses on survey data and user log data and producing simple analyses with great rigor (less focus on 'niche' advanced analytics like network analysis or multilevel modeling) so be prepared to show preciseness, in depth knowledge of survey development and analysis and user log data. Do not waste your time preparing answers using advanced analytical approaches.