I feel it's important to measure the impact of a growth hack right from the beginning instead of going back and doing micro a/b tests. I.e. set things up so tha

I feel it's important to measure the impact of a growth hack right from the beginning instead of going back and doing micro a/b tests. I.e. set things up so that it's trivial to A/B test every feature right from the start.