# Value Case Canvas

Use this before a Copilot rollout, agent, or AI workflow gets funded or scaled.

## 1. Business Mechanism

How should AI create value?

- Faster cycle time
- Higher throughput
- Better decision quality
- Fewer errors
- Lower service load
- More revenue capacity
- Better employee experience

Write the value hypothesis in one sentence:

If we use [AI solution] for [work unit], we expect [metric] to improve by [target] without making [guardrail] worse.

## 2. Work Boundary

Define the exact work.

- Task or process:
- Role or team:
- Trigger:
- Systems involved:
- Monthly volume:
- Current owner:
- AI owner:
- Risk owner:

## 3. Counterfactual

Choose the comparison.

- Previous period
- Holdout team
- Similar team
- Randomized task split
- Same task with and without AI
- Manual review sample

Explain why this comparison is credible:

## 4. Primary Metric

Choose one main metric that would justify scaling.

Metric:

Baseline:

Target:

Measurement source:

## 5. Guardrail Metrics

Pick guardrails that prevent weak value claims.

- Quality:
- Risk:
- Customer impact:
- Employee effort:
- Compliance:
- Review burden:

## 6. Total Cost

Count the full system.

- License cost:
- Build cost:
- Runtime/model cost:
- Prompt/process design:
- Review and correction time:
- Training:
- Support:
- Maintenance:

## 7. Decision Rule

Write the threshold before the pilot starts.

Example: scale if median cycle time improves by at least 20 percent, p90 does not get worse, quality stays equal or better, and cost per accepted output improves.

Decision rule:

## 8. Final Decision

- Scale it
- Change it
- Retest it
- Stop it

Reason: