Meta Data Scientist, Analytics interview questions

This was my second time applying to this position. Process: (1) Recruiter reached out me via email; didn't make sense to interview when they reached out but I reached out half a year later and arranged a phone screen. (2) Technical phone screen with Facebook data scientist. Three parts: (1) medium-hard SQL questions (no query execution, just write out query on shared text file) (2) metric ideation for measuring success of a new feature on Facebook, and then how to debug if this metric goes down (my answer: make sure this isn't an instrumentation error, then group by geo, user age, seasonality, if a power user, etc.) (3) talk through setting up an experiment for the new feature The recruiter reached out to me and said that I passed the phone screen ... but that since Facebook ran out of headcount for this position, the interview process was coming to a halt. Needless to say, shocked this would happen at a company as large as Facebook! Recruiter recommended checking in 6 months if headcount appeared but I found this suggestion pretty unhelpful since I was actively searching. I tried to see if the recruiter could refer me to an open position at Instagram, but they declined to do so. Maybe they aren't incentivized to pass a lead to another part of the company? Their loss. I ended up recently accepting a position at another FAANG company.

Standard process. I applied through an employee referral. Recruiter reached out and talked to me for about 30 minutes. He was really nice and through the entire process I felt like he really wanted me to succeed. First step was 45 minute Bluejeans interview with a data science manager (2 SQL questions + 2 product sense questions). Second step was onsite at Menlo Park with 4 rounds of interviews (SQL, probability and stats and product sense). Here are some of my observations about this role/ company: 1) Everyone who interviewed me was in their 20's. If you're older, it will work against you (you may not be "fit" for their culture). 2) If you're female and/or Asian, you have a leg up. I noticed more than 50% of people on campus were Asian. FB loves Asians. 3) This role is not really a data science role. It's more like a dumb version of product data analyst. You will work with a product owner and will run queries to answer questions. Don't expect to do sophisticated modeling or machine learning. 4) SQL, stats and probability questions are fair game and if you practice and study you can answer them. Product sense questions however are crap shoot. I personally thought my answers were good but for some reason the interviewer didn't like my answers.