[dr_dshiv]

Have you ever given thought to generalizability in A/B tests? I'm surprised there isn't more of a developing science of constructs that generally work...