"Build, Measure and Learn" as Steve Blank says: is much more elaborate than putting software into production to see if it works. The Lean Startup movement is very promising, but for many teams it ends up translating into an important question: "Yeah, but what to build ?""In ThoughtWorks, our response has been a process called an inception. We gather together a good sample of the people who will be affected by the product and have an intensive session to set an initial direction, using a series of exercises focusing on collaboration and the capture of broad goals. We don't attempt a detailed specification, as that is exactly the kind of thing that becomes out of date as soon as code hits production. But we do want to understand what kind of outcomes we are hoping for, the features that we think will drive these outcomes, and how to assess the effectiveness of our product.With The Lean Inception, Paulo has captured his experience in running these inceptions over the last decade. In particular it's focused on his work to boil the inception down to its essence, concentrating the activity on a single, if very intensive, week of work. Paulo shares how he makes this work, through writing a product vision, capturing personas, understanding the user journeys, and developing high-level features. The result isn't a detailed plan of work, which we find quickly rots into irrelevance. It is a guiding set of goals to set us off in the right direction. It doesn't plan out a final product, with all the features that our users will need, instead it focuses on an initial product that we can release and learn from - the Minimum Viable Product. " - Martin Fowler, Chief Cientist at ThoughtWorks