B2B decision contract¶
This contract operationalizes PDAMR without redefining it. Pending fields must be completed by people with operational knowledge and authority, not by the CDI-BoK.
Instantiated PDAMR¶
| Element | Case contract |
|---|---|
| Priority | Recover commercial compliance after April–June deterioration |
| Decision | Choose the first funnel intervention and whether to act now |
| Action | Translate the option into scope, owner, date and stop conditions—pending |
| Metric | Proposal conversion and compliance; expected range; 1–2 month horizon |
| Risk | Regression to mean, confounding, small channel n, operational overload and false learning |
Decision rights¶
The proposed Commercial Director decision owner is unconfirmed. Finance approval is conditional for reallocation; sales execution, committee escalation and Analytics/Decision Steward audit remain unassigned or unconfirmed. Without a confirmed owner there is a pilot design, not an organizational pilot.
Alternatives¶
| Option | Action | Best argument | Main risk | Reversibility | Status |
|---|---|---|---|---|---|
| A | Improve proposal quality, speed, coaching and follow-up | Addresses largest observed decline and can be bounded | Capacity; real issue may be pricing or mix | High | Provisional preference |
| B | Shift budget from mass digital to event/referral | Observed rates are higher | Small samples; difference imprecise | Medium | Insufficient evidence |
| C | A + B | Addresses conversion and source quality | Cannot identify which component worked | Medium | Not recommended first cycle |
| D | Do not act yet | Avoids reacting to anomaly | Another month may deepen the gap | High | Active comparator |
Option A dominates as a first test, not as truth: it is more reversible and proximate to the signal. The conclusion changes with credible seasonality, a pricing explanation or no capacity to intervene.
Required preregistration¶
Before action, freeze chosen and rejected options; seller/region/opportunity scope; action, accountable actor, date and stops; approved baseline and sensitivity; lower/central/upper expectation; horizon; intervention and comparison; known confounders; and process-quality score.
Process-quality rubric¶
Before outcome, score 0–2 for real alternatives, classified evidence, uncertainty, trade-offs and prior expectation. A high score does not guarantee a good outcome; it records that basic omissions were avoided.