Use this scorecard to test one proposed AI use case before your leadership team commits more time or money. Score the evidence behind the business decision, workflow, data, ownership, controls, adoption, and measurement. Use the result to choose one decision: stop, fix, test, or fund the next stage.
Define the use case before you score it
Write the use case in one sentence and name the operating decision it should improve. Record the executive sponsor, process owner, and proposed operating owner. Describe where the output will enter the workflow and who can accept, reject, or correct it. List the data sources the team expects to use. Ask the organisation's legal, risk, security, privacy, and compliance owners to define the requirements that apply. Set a decision date and require the team to choose the next action at that meeting.
If the team cannot name the use case or decision, stop the review and define them first. A broad score for the whole organisation will hide the constraints that matter to a specific investment.
Read the leadership guide to AI readiness if you need the questions behind each row.
Use an evidence scale
Score each row against evidence your team can inspect. The scale records evidence completeness for this use case. Use it as an internal decision aid. The scale has no validation as a maturity benchmark and cannot predict project success.
| Score | Evidence state |
|---|---|
| 0 | The team has no answer, owner, or current record for the question. |
| 1 | The team has an answer, but the owner, evidence, or decision consequence remains unclear. |
| 2 | The team has a named owner, current evidence, and an agreed consequence for the next decision. |
Mark each essential requirement before scoring starts and record who made that classification. Treat a missing essential requirement as a decision issue even when the total score looks strong. Your legal, risk, security, privacy, and compliance owners decide which requirements are essential for the selected use case. Revisit the classification when the use case or operating context changes.
AI readiness scorecard
Complete one copy for each use case. Record the evidence reference beside every score so another reviewer can follow the decision.
| Decision area | Leadership question | Evidence to inspect | Owner | Score (0–2) | Essential? | Evidence reference or action |
|---|---|---|---|---|---|---|
| Business decision | Which operating or investment decision should this use case improve? | Named sponsor, current process, baseline, and decision date | ||||
| Workflow | Where will the output enter the work, and who can accept, reject, or correct it? | Workflow map, review points, exception route, and escalation route | ||||
| Data | Can the team access and use the required data for this purpose? | Source systems, access owners, quality issues, retention rules, and permitted-use record | ||||
| Accountability | Who owns the result during delivery and after release? | Business owner, technical owner, risk owner, operating owner, and support responsibility | ||||
| Controls | Which risks require prevention, detection, human review, or a stop rule? | Risk record, test criteria, review rules, incident route, and stop criteria | ||||
| Adoption | What must managers and users do differently for the workflow to operate as intended? | Role changes, manager actions, user preparation, support needs, and feedback route | ||||
| Measurement | How will leaders judge value and risk after release? | Baseline, success measure, guardrail measure, review owner, and review date |
Seven rows at two points each give a maximum of 14. We have not set a pass mark because the total cannot resolve a missing owner, an unacceptable risk, or an essential requirement with no evidence. Use the total to compare revisions of the same use case, then read every zero and one before you make the funding decision.
Turn the score into a decision
Choose the decision that the evidence supports today.
| Decision | Use it when | Record before the meeting ends |
|---|---|---|
| Stop | The use case lacks a credible business decision or an acceptable basis for more work. | Why the team closed it and who approved that decision. |
| Fix | A named constraint blocks a responsible test or production commitment. | The owner, required evidence, and next review date. |
| Test | The use case has enough evidence for a bounded experiment with explicit limits. | The test scope, owner, review criteria, and decision date. |
| Fund | The evidence supports the next delivery stage. | The approved scope, owners, controls, measures, and next gate. |
Do not average away disagreement. When two reviewers give different scores, ask what record would settle the point and assign that evidence request to an owner. Update the scorecard when the evidence changes, then preserve the earlier version with the decision record.
When external assessment can help
Bring in an external assessor when leaders need an independent structure for the review, several owners disagree about the evidence, or the team needs to turn findings into a delivery decision. Require the assessor to list the records they will inspect. Ask who they will interview and how they will handle unresolved requirements. Confirm which decisions the leadership team will be able to make afterwards.
Black & Rhine's AI readiness assessment service is the commercial route for teams that want help to structure that work. The service page explains when an external review may fit. Teams that have already selected a use case can review AI implementation consulting. That page focuses on the decisions needed to prepare the next delivery stage. If this scorecard exposes unresolved partner responsibilities, use the AI implementation partner scorecard.