this comparison is of course a bit nonsense because of selection into Prime by heavy users the estimation of the actual revenue lift from prime for a representative user is quite tricky. i was once asked how i would identify this figure in an interview, long agohttps://twitter.com/TrungTPhan/status/1357703917648322560 …
-
-
You'd also have to account for the lost revenue from shipping fees, which would be variable based on size of post-prime orders
-
But in order to do that you'd also have to credit the before side with the additional sales used to bump an order past the $50 free shipping limit
- Show replies
New conversation -
-
-
Imo getting a clean counterfactual is a really high bar to reach. Something like A/B testing could give you a clean counterfactual (random free trial for prime) but then you incur real expense
-
I would assume that A/B testing would not be the correct answer because they want you to be clever with cuts of data and save them that expense. This is a very good interview question-- the first instinct was probably fine for first pass. But they would want to dig deeper
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.