The Build-Measure-Learn feedback loop Ries introduced does not work out if only some of the three parts excel. A startup must focus on minimizing total time to complete a turn.
The feedback loop is written in execution order, but the planning actually works in the reverse order. The outcome of each time going through the loop would be used to determine whether to pivot. So we first figure out what we need to learn to decide whether to do the sharp turn or not. And then we figure out what we need to measure to gain that learning. And then we figure out what MVP we need to build to collect those measurements.
We all believe that Facebook has proven it’s two leap-of-faith assumptions, value hypothesis and growth hypothesis, right. People love to see what happens in others life, so they come back to Facebook everyday, and they love it so much that they share it to their friends by word of mouths. But how do I frame my leap-of-faith assumptions?
Randy Komisar has a great way of doing it using the concept of analog-antilog. When Steve Jobs was inventing iPod, Sony Walkman would be the analog, and Napster would be the antilog. People were already used to listen to music in a public place using earphones, but they were not willing to pay for music. In the iPod business, people would pay for music is one of the leaps of faith.
Once the leap-of-faith assumptions are made, we should quickly head to test the assumptions. But instead of going deep and wide to collect information like how Toyota planned for the Sienna 2004, with resources that would not be available to a startup. We could act like how Scott Cook started Intuit in 1982, just grab two phone books and start cold calling people asking if they feel frustrated paying bills by hand, and then growing his vision of having personal computers take care of paying bills.