Choose one moment worth learning from.
A pilot should have a clear purpose and a bounded population. Keep the first question narrow enough that the evidence can change a decision.
- Moment and audience
- Eligibility and budget
- Reward catalogue
- Success and limitation criteria

