Hacker News Reader

Few questions:

1. How do you determine who fits into the A cohort and who fits into the B cohort? If I refresh could I fit into a different cohort and thus see a different price? If I go Incognito?

2. Instead of a 50/50 A/B test, can you multi-arm bandit a handful of different price points and shift a higher proportion to the "best" one? And thus be able to change over time as well?

View on HN

1. Using random id + localstorage, so yes, incognito could result in a different price.

2. I'd love to implement MAB, but focused on nailing core product for now.

How do you avoid customers changing their behaviours by observing pricing changes? Or when they share it with a friend and go "cute shirt for 10 dollars" and their friend is like "wat it's 15 dollars"

Could cohort based on geo-location instead of completely at random. Makes analysis a bit more challenging though and doesn't completely solve the problem of course.

Reply on news.ycombinator.com