Insights
Discovery Sprint: how to validate a bet before writing code
By Evandro Maciel
Published
Short answer
A Discovery Sprint is a 2 to 3 week cycle to test the riskiest hypotheses behind a bet, such as problem, solution, pricing, demand and messaging, before investing in building it. At the end, the team decides with evidence whether to go, adjust or stop.
Why validate before building?
Because the most expensive mistake is building something the market didn't ask for. In a CB Insights analysis of 101 startup post-mortems, 42% cited no market need as a reason for failure.
Capital doesn't fix that on its own. Quibi raised $1.75 billion and announced it was shutting down six months after launch.
Which hypotheses should you test first?
Every bet rests on a few assumptions. The ones to test first are those that, if wrong, bring down everything else:
- Problem: do people really have this problem, and does it hurt enough?
- Solution: does the offer solve it better than what they do today?
- Pricing: how much would they pay, and who pays?
- Demand: are there enough people who want this now?
- Messaging: which way of explaining it makes people act?
How does a Discovery Sprint work?
It starts by listing the assumptions behind the bet and picking the riskiest. Then come the experiments: interviews with the target audience, prototypes, and demand and pricing tests, designed to give an answer in days, not months.
At the end, the evidence is consolidated into a clear recommendation: go, adjust or stop.
How do you know whether the bet passed?
By setting the criteria before the test. If success is only defined after seeing the numbers, any result looks good. With the criteria agreed upfront, the decision stays honest.
Stopping early is a result too: you save the cost of building something with no demand and free the team for the next bet.
What changes when a team decides with evidence?
Decisions get faster and debates get shorter. At Frete.com, a discovery routine led by Shift23's founder had 73% of its experiments validated and led to 3x freight postings per customer.
Sources
Related services
- Funnel & onboarding experiments
An A/B framework, prioritized hypotheses and an honest read of the results.
- Discovery Sprints
Validate problem, solution, pricing, demand and messaging with fast experiments.
- Zero to MVP
Minimal scope, built with tests and documented decisions, and analytics from day one.