1. Establish the Case
Clarify the objective, decision to be made, boundaries, roles and authority. Record the available material and where it came from.
Method
A Case is prepared for a decision, not just summarised.
Rabbit starts with the purpose and scope of a Case, then qualifies what is known before weighing impact and options. Material uncertainty remains visible to the person who must decide.
The method
2. Qualify information
Separate supported facts from assumptions, opinions and unknowns. Surface gaps, conflicting accounts and dependencies that could change the answer.
3. Assess consequences
Consider the material effect on scope, time, commercial or financial position, and other relevant risks. Avoid false precision where evidence is weak.
4. Prepare a decision
Compare options, expose trade-offs and prepare a recommendation or a decision request. A human with the right authority makes and records the decision.
A Case, step by step · simplified construction example
A delayed delivery with conflicting reports
This illustrative example shows the method. It is not a real Case or a claim about a completed pilot result.
- Evidence
- The site log records a missed delivery. A supplier email gives a different arrival date. Neither source confirms the revised installation sequence.
- Interpretation
- The delay is supported; its duration and cause remain uncertain. The sources conflict on timing.
- Materiality & Case State
- The uncertainty could affect programme and cost. Rabbit keeps it visible as an unresolved information need.
- Impact & options
- Check the supplier position, resequence work where feasible, or wait for confirmation. Each option has dependencies and potential consequences.
- Recommendation
- Seek confirmation before committing to a costly change; prepare a time-sensitive decision request if waiting creates its own risk.
- Human decision & outcome
- The authorised manager chooses an action. The decision basis and later outcome can be recorded and compared.
What the current pilot is testing
Invited reviewers are examining a Web Pilot centred on live decision support. Feedback on its questions, analysis, clarity and friction is part of the validation process. A defined Decision Audit workflow exists in the Foundation, but the current reviewer exercise must not be presented as its validation.