It seems to me that almost every statistician that isn't in academia is doing some form of A/B testing (I guess more generally, RCTs).
