This should be the first onsite writeup for their company anywhere online.
Overall it left a really good impression. The recruiter was very professional and moved things along and coordinated quickly, and the interviewers were great too. Everyone I met was very friendly and good at guiding me.
The onsite was four rounds in total: HM + stats + case + coding.
The HM round asked in a lot of detail, with lots of follow-ups about impact. I think I failed this round.
The stats round was a bit confusing. They gave a scenario and asked what to watch out for in modeling, for example what to do about multicollinearity and whether it affects the results, what to do if you need to determine whether a certain feature can improve retention, and how to look at feature importance.
Case: the company is launching a personalized offer recommendation system, and I had to design an experiment. It was basically product sense plus decision-making in different situations. The interviewer was really nice in this round, and the guidance and feedback were very positive.
Coding was SQL + Python. You can use AI to fix syntax; they mostly look at your thinking, and whether the syntax is correct matters less. The Python was not pandas, more like simple algorithm problems. The interviewer in this round was especially nice and warmly answered a lot of my questions about the company, and I'm grateful to him.
Fetch felt a bit like a startup to me, with a flat internal structure where everyone has to take on a lot of ownership, so they probably want people with richer hands-on experimentation experience. Of all the rounds, the HM round asked the most detailed questions and had the most follow-ups, so I'd recommend preparing well for that one.
Discussion
Loading comments…