The key ideas

Sprint: How to Solve Big Problems and Test New Ideas in Just Five Days

Jake Knapp, John Zeratsky, Braden Kowitz, 2016

A team is arguing over whether to give a robot a face and teach it to dance. Instead of months of development, they get five scheduled days and five hotel guests on Friday. A method invented at Google and refined at the venture fund GV: how to turn a big problem into a question you can test in a week, and get a real answer to it.

Who it is for

For anyone stuck in a long argument, or afraid to pour months into an idea nobody has tested. For teams, founders and anyone who needs a result within a week.

The sprint gives teams a shortcut to learning without building and launching.

What the book is about

At a hotel in California, a guest who forgot a toothbrush gets one delivered. It is brought by a robot a little over three feet tall. Just a week earlier, the team that built it was arguing: should the machine have a face, and should it dance? That argument was not settled in a meeting. It was settled by five hotel guests on a Friday.

This is how a sprint works. Jake Knapp came up with the format while working as a designer at Google. Later, with John Zeratsky and Braden Kowitz, he ran more than a hundred of these weeks with startups from the venture fund GV, by the authors' own count. A sprint is five days on a strict schedule. Each day removes part of the uncertainty, and by Friday, guesses and arguments give way to the real reactions of real people.

One honest note up front: this is a practitioners' method, grown from their experience, not a proven scientific law.

Idea 1. Monday: start at the end

Usually a team rushes to invent solutions on day one. In a sprint, Monday works the other way round: first you figure out where you actually need to end up, and only then look for the road. Hence "start at the end".

The morning begins with a long-term goal: where the team wants to be in a year or two. Then comes an unexpected move. The team takes its fears and rewrites each one as a question. Not "what if we fail", but "will guests feel awkward around a robot?". A fear turned into a question shows you exactly what needs testing.

That is what happened at the startup Savioke, which built Relay, a delivery robot. Its CEO, Steve Cousins, worried that guests raised on movie robots would expect the machine to have feelings and hold conversations, which it could not do. That fear became the main question of the week.

By evening the team draws a simple map of the customer's journey, from the start to the goal, and picks one target on it: one type of customer and one moment to test on Friday.

Livria example: before moving to another city, you can first write down what scares you most and test exactly that, instead of picking curtains.

A question born from a fear can be more useful than a goal: it shows you exactly what needs testing.

Read in Livria

About the author

Jake Knapp, John Zeratsky and Braden Kowitz are designers who worked at Google and at the venture fund GV. Knapp invented the sprint format while at Google, and at GV the authors ran sprints with the fund's startups.

All books by this author: Jake Knapp, John Zeratsky, Braden Kowitz

Terms from the book

  • Facade
  • Sprint: In the book, a five-day process in which a team goes from a big problem to testing a prototype with real customers. A practitioners' method, not a scientific methodology.
  • Decider: A role in the sprint: the person who makes the final decision and owns it. The authors' term.
  • Together alone: The authors' principle: the team sits in one room, but each person thinks and sketches separately, with no group discussion.
  • Supervote: The Decider's final vote, which settles what goes into the prototype. Part of the sprint's decision ritual.

All terms

Similar books