Your KPI Changed: What Should Happen Next?
Trace the signal, inspect the evidence and build a sample decision record.
Customer 1: presenter fallback
Trace the signal, inspect the evidence and build a sample decision record.
The signal and its boundary
Northline Retail is a fictional two-store retailer. All figures and source descriptions are synthetic.
| Scope | Prior | Current | Change |
|---|---|---|---|
| Store A | $60,000 | $54,000 | −10% |
| Store B | $40,000 | $39,000 | −2.5% |
| Combined | $100,000 | $93,000 | −7% |
Store A accounts for $6,000 of the $7,000 decline; Store B accounts for $1,000. This locates the change. It does not prove its cause.
The evidence
- E1 · Sales export: authoritative for the displayed totals only.
- E2 · Inventory log: incomplete supporting evidence. Gaps do not prove stock was unavailable.
- E3 · Promotion calendar: context only. Timing does not establish an effect.
- E4 · Price-change note: unverified explanation. No cause has been established.
Challenge, review and record
Reject “Store A explains the entire decline”: Store B also fell $1,000. Keep “A price change caused the decline” unsupported. Missing evidence is not the same as a contradiction.
Proposed next step: a bounded inventory check and data reconciliation. Operations is the fictional decision owner; Finance is the fictional reviewer. The human reviewer may revise or reject the proposal. No operational action is executed and no recovery is promised.
Specific action owner to be named before the check; Operations owns the decision. Next review point: after the inventory check and data reconciliation; exact review time to be agreed.
Outcome: Not yet measured. Roles and permissions are conceptual, not real access-control enforcement.
This session’s lens
Follow a changed KPI from a first signal to a bounded, human-reviewed next step.
Product readiness is not established. Baseline measurements and customer validation still need to be collected; no demo selection counts as validation.