Design sprint vs Evidence Sprint: when five days isn't enough
Gideon Awolesi ·
The design sprint, which Jake Knapp created at Google in 2010 and wrote up with John Zeratsky in the 2016 book Sprint, packs a product decision into five days: map the problem, sketch competing solutions, decide, build a realistic prototype, and watch real customers react to it. It's a great format. But for some products, five days and a prototype that only looks real aren't enough to answer the question. Here's how to tell, and what we do instead.
01What a design sprint gives you
A design sprint gets the right people in a room for a week and ends with real customer reactions to a prototype. It's very good at getting a team to agree on a direction quickly, and at killing weak ideas before anyone writes code. The prototype is usually a facade, built to look real for the length of a test session.
02Where five days runs out
A facade works when the question is about the concept and the screens. It runs out when the answer depends on things a facade can't fake:
- Real data. People react differently when they see their own numbers, or data that looks like theirs.
- AI features. A demo is a simulated environment. Real users paste in messy data and take answers at face value.
- Regulation. In a medical device or a finance product, you need test results you can trace and keep, not just reactions.
- The build decision. A week of testing tells you if people want it, but not what it'll cost to build properly.
03What an Evidence Sprint does differently
Our Evidence Sprint takes three to four weeks. We interview your team and a few users in the first week and agree on the one flow to build. In the second week, we design and build it as a working app on a live URL, with seeded or real data, and live AI calls where the flow needs them. In the third week, we test it with 5 to 8 people from your user group. Then you get a written test report, and either a fixed price for the full build or a recommendation to stop.
04Which one to choose
Run a design sprint when you need a team to agree on a direction, and the question is about the concept. Run an Evidence Sprint when the question depends on real data or AI, when you need evidence you can keep, or when you need a fixed price for the build at the end. Some teams do both: a design sprint to pick the direction, and an Evidence Sprint to prove the riskiest flow.
Our Evidence Sprint starts from €9,500.