This is good stuff. I’ve been a pretty vocal critic of A/B testing for a while, and unless you understand the nuances of data and the importance of audience matching, it falls on deaf ears. I see people making incredibly small changes, spurious correlations, and complete ignorance towards the “why.” Basically, when I work with clients, I ask: how can we minimize the amount of stuff you actually have to A/B test? Because it’s impossible to test everything when you could spend that time on leads, development, etc. And just because you experienced a change, lift, or a conversion boost doesn’t mean it came from the change you made.