Facebook Mid-Level | May 2021
Location: London, KY
Role: Data Scientist
When did you have this interview?: Last Month
I was contacted by a Recruiter and we scheduled a phone screen call. The screening call was about my experience and my area of focus of my current work. And she spent some time, explained about the position and the work and which product team I might join. After that, she asked some topline Case Analysis about Newsfeed and how would you prioritize between ad frequency & user experience there. The question led to AB testing and she asked what are the actions and steps I would take to evaluate this with the AB testing. We talked a bit about sample size as well.
The next day, I received an invitation to the next interview and a lot of preparing materials for the interview. I scheduled the next interview in 2 weeks.
The second interview has 2 parts: product sense and technical questions which I solved with SQL. The product sense question is about the “People you may know” feature: “Facebook has the people you may know feature to help people connect to more friends, however, it seems like the more friends a user has, people, start to post less. How would you go about it?”
We had a discussion about the product, Ab test, and bias in the Ab testing and then we moved on to SQL. The product sense is longer than the technical part.
What kind of technical questions did you get asked?
The SQL questions: Given a table: User1 | User2 123 | 124 124 | 258 357 | 406
How would you calculate which user has the highest friend count? (Note that the friendship table is one direction: 123 is friend with 124, but 123 wont show in the User2 column, only User1 column)