Validation Sprint.
Ten working days to answer the question your roadmap has been avoiding. Cheaper than a quarter of building the wrong thing.
Two weeks. Three milestones. One answer the team can act on.
Useful for, and not.
A framework earns its keep by being clear about what it doesn't do. The Validation Sprint is for one specific kind of question.
Use it when
- A six-figure decision is sitting in the roadmap, defended only by belief.
- The team is mid-debate; another month of arguing will cost more than the test.
- The risk is technical or behavioural, and a small experiment can make it cheap to know.
- You want a written answer your board, not just your engineers, can read.
Don't use it when
- The decision is already made and you want validation as theatre.
- The question is a strategy question dressed as a tech question.
- The team needs the headcount, not the answer.
- Ten days will not change anyone's mind regardless of result.
Two weeks.
One contract.
A fixed shape your team can plan around. The work happens with your engineers, not in parallel to them.
Frame the question
Co-written brief · one page
We sit with you and the team and write the question down. Stop conditions. Evidence required. What "yes" and "no" each look like, in the architecture, in the org, on the P&L.
Build the test
Smallest thing that gives evidence
A throwaway prototype. A load test. An interview round. A scoped data study. Whichever is cheapest to make the answer credible. The test ships; we keep the learning.
Land the answer
Written memo · all-hands optional
A short, blunt memo. What the evidence said. What to do next. What we still don't know. Your team owns it from there.
A small number replaces a large argument.
Validation Sprints have shifted six- and seven-figure roadmap decisions in two weeks. The work that follows isn't faster. It's just aimed at the right thing.
"Ten days is almost always enough to know what a year of building wouldn't tell you."
Jędrzej Tabaczyński
Is this the question you're sitting with?
If a decision is stuck and the cost of being wrong is large, ten days is almost always the right next step.