When I was in high school our math and science teachers drilled this into our heads: Always test your hypotheses.
I bet it was the same for you.
And yet — when it comes to experimentation in our startups along the lines of the lean startup methodology, A/B testing or when we place “little bets” we forget all we learned. We head out, do our experiments and go by gut feel if what we did was successful or not. Or we make up the criteria ex post facto.
Don’t do this.
I’m a huge fan of a customer-driven process which places many little bets, takes the results and refines the outcome until we solve our customers problem in the best possible way. Develop deep customer empathy, go broad to go narrow, then run rapid experiments with your customers. And do so with rigor. Formulate and test your hypothesis. And you will soon create true customer delight.