Excel template · Prove
AI Value KPI Set by Layer
A starter scorecard that takes KPIs from the library and maps them to the five layers of AI value management, with baseline, target and owner for each.
Excel (.xlsx) · version 0.1 (2026-10-27) · a proposal from this site, not a standard · CC BY 4.0
- What it is
- A scorecard of 23 KPIs from the KPI library, grouped by the five layers and led by attribution coverage, with columns for owner, source, frequency, baseline, target and current value.
- Who it is for
- A value management lead, FinOps lead or finance partner choosing what to report and to whom.
- When to use it
- When setting up AI value reporting for the first time, or when a dashboard reports activity and the board asks what it proves. Pick a handful per layer; nobody needs all of them.
- Not for
- A benchmark. Targets are yours to set, and the library's benchmark notes are directional.
AI Value KPI Set by Layer
Excel (.xlsx) · version 0.1 · CC BY 4.0
The full structure
Every sheet in the file, in order. The download adds space to write, the formulas, input shading and drop-down lists.
Instructions
How to use the workbook: cream cells are inputs, grey cells are formulas, blue rows are illustrative.
- Choose two to four KPIs per layer. Delete or hide the rest rather than reporting all of them.
- Name an owner role for each KPI you keep, and the system the figure comes from.
- Set the baseline before the target. A target without a baseline cannot be checked.
- The Status column compares current with target in the direction you choose (higher or lower is better).
Scorecard
One row per KPI, grouped by layer, with attribution coverage first. The KPI name, formula and library ID come from the KPI library.
| Layer | Library ID | KPI | Formula |
|---|---|---|---|
| All layers (headline) | Headline | Attribution coverage | Outcome-linked AI spend / total identified AI spend × 100 |
| 1. Usage Transparency | 1.1 | AI Spend Allocation Rate | AI Spend Allocation Rate = (AI spend allocated to owners / Total AI spend) × 100 |
| 1. Usage Transparency | 1.2 | Shadow AI Spend Ratio | Shadow AI Spend Ratio = (AI spend outside formal IT budgets / Total AI spend including shadow) × 100 |
| 1. Usage Transparency | 1.4 | AI TCO Completeness Score | TCO Completeness Score = (Number of TCO layers with active cost tracking / 7) × 100 |
| 1. Usage Transparency | 1.6 | Unallocated Shared AI Capacity Cost | Unallocated Capacity Cost (%) = (Monthly capacity cost pool − cost allocated to teams, batch work and policy headroom) / Monthly capacity cost pool × 100 |
| 1. Usage Transparency | 2.1 | Cost per Inference | Cost per Inference = Total inference cost for a service / Total number of inference calls |
| 1. Usage Transparency | 7.1 | AI Budget Forecast Accuracy | Forecast Accuracy = (1 - |Actual AI spend - Forecasted AI spend| / Forecasted AI spend) × 100 |
| 2. Output Quality | 2.4 | Token Efficiency Ratio | Token Efficiency Ratio = Tokens in accepted/useful outputs / Total tokens consumed (input + output + retries + overhead) |
| 2. Output Quality | 2.6 | Coding-Agent Cost per Merged Change | Cost per Merged Change = (Coding-agent spend + extra review time cost + rework cost) / Merged changes that clear the quality bar. Also report the cost of the additional changes: the same cost divided by merged changes above the pre-rollout baseline. |
| 2. Output Quality | 3.2 | Model Routing Efficiency | Model Routing Efficiency = (Queries routed to optimal-cost model / Total queries) × 100 |
| 3. Productivity Value | 5.1 | AI Baseline Definition Rate | Baseline Rate = (AI initiatives with documented pre-deployment baselines / AI initiatives that have reached production) × 100 |
| 3. Productivity Value | 5.2 | AI Adoption Rate | Adoption Rate = (Active users of AI capability in period / Target user population) × 100 |
| 3. Productivity Value | 5.3 | Capacity Redeployment Rate | Redeployment Rate = (Hours redeployed to defined higher-value activities / Total hours claimed as saved by AI) × 100 |
| 3. Productivity Value | 2.3 | Cost per Business Outcome | Cost per Business Outcome = Total AI cost for the capability (all seven TCO layers) / Number of completed business outcomes |
| 3. Productivity Value | 5.4 | AI Benefit Realisation Accuracy | Benefit Realisation Accuracy = (Actual measured benefit / Forecasted benefit in business case) × 100 |
| 4. Delivery Alignment | 4.1 | AI Initiative Ownership Rate | AI Initiative Ownership Rate = (AI initiatives with named economic owner / Total AI initiatives) × 100 |
| 4. Delivery Alignment | 4.2 | AI Business Case Coverage | Business Case Coverage = (AI spend governed by approved business cases / Total AI spend) × 100 |
| 4. Delivery Alignment | 4.4 | AI Governance Span | AI Governance Span = Count of distinct governance functions involved in pilot-to-production approval |
| 4. Delivery Alignment | 7.3 | AI Run/Grow/Transform Ratio | Report as three percentages: % Run / % Grow / % Transform |
| 5. Portfolio Strategy | 6.1 | AI Portfolio Balance Ratio | Report as three percentages: % Predictable ROI / % Strategic Innovation / % Exploratory |
| 5. Portfolio Strategy | 6.2 | AI Portfolio Concentration Risk | Concentration Risk = (Largest single dependency as % of total AI spend or total AI value) |
| 5. Portfolio Strategy | 4.3 | AI Initiative Kill Rate | AI Kill Rate = (AI initiatives intentionally stopped in period / Total AI initiatives reviewed in period) × 100 |
| 5. Portfolio Strategy | 6.3 | AI Portfolio Review Cadence | AI Portfolio Review Cadence = Number of structured AI portfolio reviews per year |
| 5. Portfolio Strategy | 1.5 | AI Vendor Consolidation Ratio | AI Vendor Consolidation Ratio = Total AI vendors in use / Number of production AI use cases |
The workbook adds columns for owner (role), data source, frequency, better when, baseline, target, current, status. Full definitions and watch-outs are in the KPI library.
Layer summary
For each layer, how many KPIs have a baseline and how many are on target. A layer with no baselines is where to start.
| Layer | KPIs kept | With a baseline | On target | Behind target |
|---|---|---|---|---|
| [ ] | [ ] | [ ] | [ ] |
Layers
The question each layer answers, as set out in the five layers of AI value management.
| Layer | Question |
|---|---|
| [ ] |
Illustrative
Illustrative rows
Hypothetical. Every figure is invented and describes no real organisation. They are the blue rows in the workbook.
| KPI | Baseline | Target | Current | Status |
|---|---|---|---|---|
| Attribution coverage | 18% | 40% | 26% | Behind target |
| AI Spend Allocation Rate | 61% | 80% | 72% | Behind target |
| AI Baseline Definition Rate | 35% | 75% | 80% | On target |
| AI Initiative Ownership Rate | 70% | 100% | 100% | On target |
Terms used in this template
Read alongside