meta | Data Scientist | Interview Experience
Interview Date: Not specified
Result: Not specified
Difficulty: Not specified
Interview Process
The interview process consisted of three rounds with DoorDash, Private, and Netflix.
DoorDash:
The focus was on product sense, particularly related to restaurants. Questions included how to increase user loyalty (not engagement or volume), metrics used for launched restaurants, and analysis of a drop in restaurant revenue. The model involved a classic orders model with tables for orders and order items, including follow-up questions about market share metrics and handling multiple menus for a restaurant.
Private Account:
Product questions were similar to those covered in other interviews, including inquiries about engagement drops and user privacy factors. SQL questions involved filtering active user actions from a user action table, specifically how to handle multiple records for the same action.
Netflix:
All questions were based on previous problems, focusing on SQL involving two tables related to user video views and video types. A more complex question required updating a third table daily, necessitating the use of full outer join and coalesce functions.
Technical Questions
-
DoorDash:
- How to increase user loyalty?
- SQL: Analyze pickup and delivery ratios, top 3 restaurant types.
- Python: Model for order steps with specific input/output requirements.
-
Private Account:
- SQL: Filter out multiple records for the same user action.
- Python: Friend recommendation problem using a dictionary.
-
Netflix:
- SQL: Queries on user video views and types.
- SQL: Update a third table daily with user video view statistics.
Tips & Insights
Time management is crucial during interviews. Be mindful of the time spent on discussions, especially if interviewers tend to elaborate on their introductions. If the interview is dragging on, politely remind the interviewer to keep on track. Good luck to everyone!