7-Day Sprint
A one-week operating rhythm for agentic MVP experiments.
Key takeaways
- A disciplined seven-day sprint is enough to move from hypothesis to evidence.
- The day-by-day plan runs from hypothesis and success criteria on day 1 through launch and decision review on day 7.
- Daily review questions guard against scope creep, unclear hypotheses, and uninstrumented critical paths.
- Done means the test surface is live, primary and diagnostic events are captured, critical paths are verified, and the decision rule is written before results arrive.
A seven-day sprint is enough to move from hypothesis to evidence if the scope is disciplined.
Sprint Plan
| Day | Focus | Output |
|---|---|---|
| 1 | Hypothesis and success criteria | Experiment brief |
| 2 | Context pack and repo inspection | Implementation plan |
| 3 | First build slice | Working path |
| 4 | Instrumentation and copy | Measurable test surface |
| 5 | Quality and safety | Fixed critical issues |
| 6 | Launch preparation | Traffic plan and support notes |
| 7 | Launch and decision review | Evidence report |
Daily Review Questions
- Is the primary hypothesis still clear?
- Did implementation create unnecessary scope?
- Is the critical path instrumented?
- What would make the evidence ambiguous?
- What must be fixed before real users see it?
Done Criteria
- The test surface is live or ready for controlled release.
- The primary event and diagnostic events are captured.
- Critical UI, auth, payment, or data paths are verified.
- The decision rule is written before results arrive.