Alberto Savoia had a successful career as Chief Technology Officer in major companies such as Sun Microsystems, SunLabs, and was Google’s first Engineering Manager. As an entrepreneur, however, he realized that building the right thing was more important than building things right. He chats with Barry O’Reilly about the pivotal unlearning moments in his life and his new approach to product development.

The Beast of Failure

You work hard to create a great product, you launch it and the market rejects it. That’s one of the most painful experiences for any software developer. Alberto relates his first experience with ‘the beast of failure’: even though the market told them “if you build it, we will buy”, they did not actually buy. Alberto says that this failure felt as if someone had pulled the rug from under him. [Listen from 3:00]

However, it was also a seminal unlearning moment for him. The first lesson he took away was that if you’re building the right ‘it’, you will find a way to succeed in the market. The second lesson was that you have to own your failures before you can move forward. [Listen from 8:10]

Unlearning Market Research

There is an 80% chance that the original version of any idea will fail. As such, Alberto now goes into a venture expecting failure, and the market has to prove him wrong. Optimizing to be wrong rather than to be right, flips traditional market research on its head. Barry comments that it’s at the heart of the scientific method since you have to conduct experiments to invalidate your hypothesis; if you can’t invalidate it, then it’s probably a good hypothesis. [Listen from 9:25]

Alberto’s most important experiment to test his ideas is his ‘skin in the game meter’. Asking the market if they will buy if you build is due negligence, he argues; that’s just promises and opinions. Instead, he tells them, “If you buy, we will build.” The ultimate demonstration that someone wants a product is when they put down a deposit. Money is the ultimate skin in the game, as Elon Musk’s example proves. [Listen from 12:05]


Engineers usually know whether a product can be built. The uncertainty lies in whether it should be built. Alberto says that when he looked at how creators approached this problem, he saw many examples of pretotyping. A pretotype is something you build before you start to build something that works; for example, how Jeff Hawkins developed the Palm Pilot. The only data that is valuable, Alberto says, is YODa – Your Own Data. Just as Hawkins did, Alberto only counts YODa that is backed up with skin in the game. Barry adds that YODa has the ability to shift mindsets. He has found that the people who own their results, and are continuously learning and unlearning to enhance their product, get exceptional results. [Listen from 18:35]

Alberto Savoia w Barry OReilly

Change Takes Time

Logic does not convince people to change their age-old thinking. It takes time and dedication to get people to buy in to new ideas and methods. Start with one project, Alberto advises, and incorporate some traditional techniques. Let them experience the results firsthand: that will start to open their minds up to a different way of thinking and acting. Barry agrees that logic is not enough to change minds or behavior. “You have to act your way to a new culture,” he says. “You start to see the world differently when you do things differently, and that’s what challenges your mental model and shifts it.” [Listen from 31:00]

Looking Forward

Alberto has written a book to teach entrepreneurs and innovators about pretotyping, so they work on ideas that are likely to succeed. He advises them not to depend on luck and to assume failure. If you iterate enough, however, you will find the idea that succeeds, he says. That is how to play in a systematic way. “Unlearning is learning. It just takes courage to flip it around.”