I had this interview not long ago. The SQL was the classic Delivery set — the overall question volume was pretty large, but not every question required you to actually write it out. Some of them asked how you'd turn an existing query into an aggregated version, and the interviewer just had me talk through how I'd change the query instead of writing the code, to save time.
The case question was: completed orders in the LA area have dropped, how do you investigate? My takeaway is that you just have to hold onto supply-demand the whole way through, and really think through the entire funnel from order creation to order completion, and where things could go wrong. I threw out a bunch of possibilities, and none of them hit what the interviewer was actually looking for. With only five minutes left, I finally got there after the interviewer gave me a ton of hints. Then we quickly went over how you'd design a test for that cause, and what factors you'd need to consider.
After the interview I was sure I'd failed, but a week later I got notified that I passed. Sharing this in case it helps someone.
Discussion
Loading comments…