Today I made it through the second round (not counting HR), talking with a senior data science manager. They mainly asked a few case-style questions:
If you were sending an email campaign to get people to go get a flu shot, how would you design the campaign? Mainly wanted me to share my thinking and the factors I'd consider.
If you wanted to launch a TV channel campaign, how would you design it?
If you found the results weren't good, how would you go about troubleshooting it?
For the MMX model I'd worked on before, what factors did I think were especially challenging for the model?
What do I think my strength is?
I basically just tried to say everything I could think of — there's no absolute right or wrong answer here, they're mainly looking at your thought process. You might not hit every point they wanted, but they'll give you a hint afterward about whether there's something else to consider, and as long as you can add it in correctly at that point, it's fine. After all, you can't fully know what answer the interviewer is looking for, so giving as complete an answer as possible is the way to get close.
I made it to the next round. I'll post an update with the new interview report later.
Discussion
Loading comments…