AI Use-Case Scorecard

Score a client's candidate AI use-cases on a weighted impact/effort model a CFO can't argue with. Lock your criteria and weights first, then rank every idea.

Step 1. Lock your criteria and weights

Decide what matters before you score anything. Each axis must sum to 100. This is the step most diagnostics skip, and it is why their rankings fall apart under questioning.

Impact

100/100

Revenue upside

New revenue or revenue protected if this ships.

Cost reduction

Hard dollars or hours removed from the operation.

Risk reduction

Compliance, error, or reputational risk this lowers.

Time to value

How fast the client feels the benefit after go-live.

Effort

100/100

Data readiness

Is the data clean, available, and labeled? Higher = more work.

Integration lift

Systems to connect, APIs, infra. Higher = more work.

Regulatory burden

Approvals, audits, residency. Higher = more work.

Change management

How hard to get people to actually adopt it. Higher = more work.

Both axes sum to 100. You are ready to score.

Why locking weights first matters

When you score first and weight later, you can always nudge the weights to make your favorite idea win. Locking the criteria and their weights to 100 before any use-case is scored removes that bias. The ranking becomes a function of the client's own priorities, which is exactly what survives a CFO's questioning. This is the same impact/effort discipline the Audity diagnostic engine runs at engagement scale.