Interactive example · fictional data
Run a small pilot now, or wait for every feature?
See how inputs, options, objections and outcomes fit into a reviewable decision. These figures illustrate the output; they are not verified results or a live audit trail.
1. Conclusion and next action
Start with a limited pilot. Set the review date and define success before inviting participants.
Reason: a smaller commitment creates an earlier opportunity to learn. The strongest objection: a small sample may miss enterprise requirements.
2. Objective and boundaries
- Objective
- Learn whether people can complete and defend one decision.
- Constraint
- 5 participants, 2 weeks, no automatic public release.
- Outside scope
- Proving enterprise readiness or future revenue.
3. Options and trade-offs
Illustrative ordinal scores from 1 to 5, not probabilities. Higher learning is desirable; higher effort means more work.
| Option | Learning | Effort | Interpretation |
|---|---|---|---|
| Pilot with 5 users for 2 weeks | 4/5 | 2/5 | Preferred in this fictional scenario; success remains untested. |
| Wait until all features are complete | 2/5 | 4/5 | More preparation, with slower feedback. |
| Release to everyone immediately | 5/5 | 5/5 | Wider exposure and greater support demands. |
Pilot with 5 users for 2 weeks4/5
Wait until all features are complete2/5
Release to everyone immediately5/5
Pilot with 5 users for 2 weeks2/5
Wait until all features are complete4/5
Release to everyone immediately5/5
4. Evidence and uncertainty
- User inputDeclared constraint
A pilot of 5 users over 2 weeks is proposed.
- AssumptionUnverified
The target users can spare time for weekly feedback.
- PredictionOutcome not recorded
At least 3 of 5 participants complete one decision.
- Information gapUnknown
Willingness to pay after the trial.
5. Challenge and reversal condition
Five users may give a misleadingly positive impression. Include different roles, record failures as well as successes, and postpone expansion if fewer than 3 complete a decision.
6. Forecast, review and learning
Example forecast: 65% probability that at least 3 of 5 participants complete one decision by day 14. This is a stated forecast, not a measured model accuracy.
- Day 0 · Record the input, forecast and success criterion.
- Day 7 · Record feedback and changes with reasons.
- Day 14 · Record the outcome before assessing forecast accuracy.
Outcome: not recorded. Forecast accuracy: not yet calculable.