# Two-cycle campaign review worksheet

## Before the pilot
Workflow:
Owner:
Expected output:
Acceptance rule chosen before testing:
Baseline manual preparation time:
Baseline correction/review time:

## Cycle log
| Cycle | Preparation minutes | Correction minutes | Evidence gaps | Decision usable? | Source records |
|---|---:|---:|---|---|---|
| Manual baseline | | | | | |
| Connected cycle 1 | | | | | |
| Connected cycle 2 | | | | | |

Total working time = preparation + correction time. Compare equivalent work; do not combine a complete review with a partial draft.

## Defect register
| Defect | Category | Impact on decision | Likely cause | Correction | Retest result |
|---|---|---|---|---|---|
| | date / definition / source / arithmetic / interpretation / action | | | | |

## Decision memo
- What improved:
- What became worse or remained unmeasured:
- Did the predefined acceptance rule pass:
- Remaining recurring defects:
- Decision: expand / repeat with fixes / keep manual
- Next owner and review date:

## Fictional arithmetic reference
Week A: $600 spend, 12 inquiries, 6 qualified → $50/inquiry, 50% qualified, $100/qualified.
Week B: $540 spend, 15 inquiries, 3 qualified → $36/inquiry, 20% qualified, $180/qualified.
This illustrates why cheaper inquiries do not establish better business performance. Small samples and qualification timing still need review.
