Originally Posted by VydorScope
Interesting experiment, but I would say that you used way to short of a window to know.
Yes, I agree that the window was short and the sample size could have been larger. And as I noted in the post, other external factors may have influenced the data.
Nevertheless, I think there are enough sales and a big enough change between the two periods (average increase of 50% per day) to indicate that there was at least some
positive impact from the design change.