There's a real cost to starting A/B testing too early, and it's not the one most people expect. It's not that testing is a bad idea, it's that testing without enough data behind it burns time and money on results you can't actually trust. The threshold that actually ...