Increase your failure rate

I used to worship success. I used to think failure was something to minimize.

Plan more + Predict better + Get it right the first time - this was my definition of competence.

During my stint working closely with a CPO, I got a mandate that broke this belief completely. "Run as many A/B experiments as we can. The more we fail, the better our odds of success."

I remember thinking, this makes no sense. Failure was supposed to be the thing you avoid, not chase.

But we ran experiment after experiment, every single week. Most failed and some flopped hard and somewhere in that pile of failures, we started noticing patterns. Customer behaviors we'd never have guessed at a desk. Insights that no market research could've given us.

Each failed experiment fed the product team something real and slowly, our odds of building things that worked went up. It didn’t happen because we got smarter overnight, we just finally had proof of what customers actually did, instead of what we assumed they'd do. 

I no longer see failure as the opposite of success.

It's the raw material for it.

The founders and teams unafraid to fail fast are usually the ones who stumble onto the breakthrough.

The rest are still perfecting their first guess.