Amazon's own experiment tool takes two to three months to get results, which is basically useless for anyone moving fast.
>At that pace, you might learn four to six things a year.It's hard to believe they can actually find meaningful insights just by recruiting people from email lists and external sites. I don't see how those
slight edges hold up if the sample size isn't representative of actual Amazon traffic.
article:
https://conversion-rate-experts.com/cro-for-amazon/