Once the case is approved, the only thing that really matters is whether the promised outcomes show up. That means clear KPIs, clean baselines, realistic targets, and a simple but disciplined tracking rhythm. This chapter is about turning the value case into a living measurement system that finance, line leaders, and project teams all trust.
16.1 Defining KPIs and Benefit Measures
A KPI is a small, sharp metric that reflects whether the initiative is doing what it said it would do. A benefit measure is the specific way that KPI translates into value for the case (financial or otherwise). Many teams blur those two; you should keep them linked but distinct.
Start from the value story, not from the data you have
Begin with your benefits/value case (Chapter 10) and ask:
- What will actually be different for customers, operations, risk, or financials if this works?
- At what level of granularity do senior leaders care (business unit, segment, geography, product, channel)?
- Which metrics are already tracked vs. which you must create?
You are not building a full KPI catalogue; you are instrumenting this initiative.
Characteristics of good KPIs for business cases
- Aligned: Directly reflect the core objectives in the case (e.g., “reduce handling time,” “increase digital adoption,” “lower incident rate”).
- Attributable: Movement can be credibly linked, at least in part, to the initiative (with clear boundaries and attribution rules).
- Actionable: The initiative team and benefit owners can influence them through their decisions and actions.
- Measurable: Data exists or can realistically be collected at required frequency.
- Simple: Easy to explain in one sentence to an executive or frontline manager.
For each major benefit stream, pick one primary KPI and, optionally, a secondary diagnostic. If you have more than 10 primary KPIs across the entire initiative, you are probably over-measuring.
Typical KPI categories by benefit type
- Revenue-related
- New customers acquired, by segment
- Average revenue per user (ARPU)
- Conversion rate at key funnel steps
- Cross-sell rate, products per customer
- Churn/retention rate and customer lifetime value
- Cost-related
- Cost-to-serve per unit or per customer
- FTEs or hours per unit of work
- Overtime, contractor spend
- Error/rework rate
- Unit production or logistics cost
- Risk and compliance
- Number and severity of incidents (cyber, safety, conduct)
- Near misses or control breaches
- Regulatory findings, fines, enforcement actions
- Loss events and expected loss metrics
- Customer and user experience
- NPS / CSAT for specific journeys
- Average response or resolution time
- Digital adoption vs. assisted channels
- Complaint rates
- Operational performance
- Throughput and capacity utilization
- Cycle time and backlogs
- First-time-right / first-contact-resolution
- ESG and social outcomes
- Emissions, waste, resource usage
- Inclusion/coverage metrics
- Beneficiary outcomes for non-profits and public sector
Tie KPIs to the business case
For each KPI, you must be able to say:
- “If this KPI moves from A to B, that corresponds to $X (or Y risk reduction, or Z outcome improvement) in our case.”
If you cannot make that link, it might still be a useful diagnostic KPI, but it should not be a headline benefit measure.
16.2 Baselines, Targets, and Data Sources
Defining KPIs is only half the job. You must pin down where you start, where you aim to get to, and where the numbers will come from.
Baselines
The baseline is the “before” level of the KPI, typically:
- Measured just before the initiative begins (or at an agreed reference point).
- Calculated over a period that smooths out normal noise (e.g., last quarter, last 12 months).
- Segmented in the same way you plan to measure post-change (e.g., by region, channel, product).
Good practices:
- Use exact dates: “Baseline NPS for SME onboarding, average of Q1–Q2 FY25: 41.”
- Document any adjustments (e.g., removing outlier months due to a one-off event).
- Ensure finance and benefit owners explicitly agree: “Yes, this is the baseline we own.”
If your baseline is weak (short time series, poor data quality), acknowledge it and plan early measurement to refine it. Do not pretend precision you don’t have.
Targets
Targets convert aspiration into clear commitments. For each primary KPI:
- Define the target level (e.g., “from 10% to 7% churn,” “from 25 to 18 minutes average handling time,” “+5 NPS points”).
- Define the timeframe (e.g., “within 18 months of first rollout,” “by end of FY27”).
- Make targets internally consistent with the economics: if your model assumes a 25% improvement but your target is 10%, reconcile the difference—or adjust one of them.
Tiering can help:
- Minimum acceptable: what must be achieved to justify continued investment.
- Planned: what is consistent with the NPV/IRR in the case.
- Stretch: upside scenario for performance incentives or additional investment.
Keep target inflation under control. If everything becomes “stretch,” nothing is.
Data sources
For each KPI, specify a single source of truth:
- System of record (CRM, ERP, HRIS, ticketing, data warehouse).
- Analytics pipeline or BI report (dashboard, cube, report ID).
- Survey instruments (e.g., NPS tools, feedback forms).
- For risk/ESG, specialized tools (GRC, safety systems, ESG platforms).
Also define:
- Data owner: who is responsible for data quality and availability.
- Frequency: monthly, quarterly, weekly, near-real-time.
- Latency: when data will be available after the period ends.
In the business case, you do not need to list every technical detail of the data pipeline, but you must show that:
- Every primary KPI has a trusted, accessible data source.
- You know who will provide the data for reviews.
If a KPI’s data source does not yet exist, you should explicitly include creation of that data capability in your implementation plan and cost/resourcing.
16.3 Dashboards and Performance Cadence
Measurement works when it is visible, regular, and tied to decisions. That’s what dashboards and cadence do.
What dashboards should (and shouldn’t) do
Dashboards for a business case should:
- Show a small number of primary KPIs and benefit measures, not every metric in the program.
- Separate leading indicators (adoption, early performance) from lagging outcomes (revenue, cost, risk).
- Allow drill-down by relevant dimensions (region, product, segment, channel) without overwhelming the main view.
- Highlight variance vs baseline and vs plan, not just current levels.
They should not become:
- A comprehensive enterprise performance system.
- A dumping ground for every metric anyone asked for “just in case.”
Think of them as decision dashboards: would a BU head or ExCo member get useful signals from one or two screens?
Cadence
Set a clear rhythm for reviewing KPIs and benefits:
- For active implementation phases:
- Monthly is usually right (sometimes even bi-weekly for high-risk changes).
- Focus on leading indicators: adoption, cycle times, error rates, early customer feedback.
- For post-implementation / BAU:
- Quarterly is often enough for benefits realization and value tracking.
- Focus on lagging indicators: realized savings, sustained revenue uplift, risk incident trends.
Tie cadence to existing forums:
- Monthly initiative steering committee.
- Quarterly business reviews for P&L owners.
- Risk committee or control forums for risk/ESG metrics.
Avoid creating entirely new meetings if existing ones can absorb the topic; instead, secure a standing agenda slot.
Roles in performance review
At each review, ensure:
- Benefit owners explain KPI movements and actions they are taking.
- Finance validates financial translation of KPI changes (e.g., how much savings have actually hit the P&L).
- The initiative lead tracks implementation milestones and remaining risks.
- The sponsor arbitrates trade-offs and escalates structural issues.
The business case should specify “who will be in the room” and “which forum” owns performance once the project team starts to demobilize.
16.4 Step-by-Step Measurement Plan Guide
Here’s a practical sequence to create a measurement and tracking plan that you can drop directly into your case.
Step 1 – Start from the benefits register
Take your list of benefits (from Chapter 10 / your benefits register) and, for each:
- Group benefits into 3–7 value themes (e.g., “front-office labor efficiency,” “platform cost savings,” “churn reduction,” “incident reduction”).
- Note the financial value associated with each theme and the main business owner.
This gives you the backbone of your measurement plan.
Step 2 – Define primary KPIs per theme
For each value theme, choose:
- One primary KPI that directly reflects value for that theme.
- Optionally 1–2 supporting indicators (e.g., adoption, quality) that explain the primary KPI.
Write a short explanation:
- “This KPI tracks the X effect; an improvement of Y units per period corresponds to $Z benefit.”
If you cannot make that link, revisit the KPI or the benefit logic.
Step 3 – Set baselines and targets
For each primary KPI:
- Document the baseline level with dates and data source.
- Define the target trajectory over time (e.g., Year 1: 10% improvement, Year 2: 20%, Year 3: 25%).
- Ensure the aggregate impact of those improvements matches (or intentionally revises) your value case.
Align this with benefit owners and finance. Resolve differences now, not after approval.
Step 4 – Specify data sources and ownership
Create a line per KPI:
- Data source/system (e.g., CRM report X, warehouse table Y).
- Data owner (name or role).
- Frequency and lag (e.g., monthly, available by day 5).
- Any known data quality issues and remediation plan.
If a KPI requires new data collection (e.g., adding a field, changing process to capture data), add that into your implementation roadmap.
Step 5 – Design the dashboard view
Sketch (literally on paper is fine) the main dashboard:
- Section A: Overall outcomes vs plan (e.g., NPV realization, big KPIs vs target).
- Section B: Revenue, cost, risk value themes with simple charts.
- Section C: Leading indicators (adoption, usage, quality).
Limit this to what can fit on 1–2 screens. Decide which breakdowns are most important (e.g., by region, channel).
Document in the case:
- “We will implement a simple dashboard in [tool] showing the following KPIs… owned by [data team].”
Step 6 – Fix cadence and forums
Agree:
- Which forum will review which subset of KPIs (e.g., steering group vs business review vs risk committee).
- Frequency of each review (monthly vs quarterly).
- Which roles must attend (sponsor, benefit owners, finance, risk, data).
In the business case, this might be a short section:
- “KPIs A, B, C will be reviewed monthly in the initiative steering. KPIs D, E, F will be reviewed quarterly in the BU performance review, with finance presenting realized vs planned benefits.”
Step 7 – Define variance rules and actions
Make clear what happens when numbers deviate:
- Thresholds: e.g., “If KPI falls more than 10% below target for two consecutive quarters, the sponsor must present a remediation or re-scope proposal.”
- Actions: additional analysis, targeted interventions, re-prioritization, or if needed, pause/stop decisions.
This connects measurement to management, not just reporting.
Step 8 – Integrate into the business case and implementation plan
Summarize your measurement plan in the case:
- 1 page: KPI overview table (theme, KPI, baseline, target, owner, review forum).
- ½ page: dashboard concept and data sourcing summary.
- ½ page: cadence and variance rules.
Then ensure your implementation roadmap includes:
- Building or adjusting data pipelines.
- Designing and deploying the dashboard.
- Aligning performance routines and scorecards with the new KPIs.
16.5 KPI & Tracking Checklist
Use this checklist as your last gate before you claim your case is “measurement ready.”
KPI design
- Each major benefit theme has at least one primary KPI.
- Each primary KPI is clearly defined (what it measures, unit, scope, time bucket).
- KPIs are aligned with case objectives, not just generic business-as-usual metrics.
- Number of primary KPIs is manageable (typically ≤10).
Baselines and targets
- Baseline values for all primary KPIs are documented with dates and data sources.
- Targets are defined over time and are consistent with the financial value case.
- Benefit owners and finance have explicitly agreed to the baselines and targets.
- Any weak or proxy baselines are acknowledged, with a plan to refine early.
Data and systems
- Each KPI has a clear system of record or data source.
- Data owners are identified and accept responsibility for data quality.
- Measurement frequency and lag are defined, and realistic (e.g., monthly with data by day 5).
- Any required new data collection or pipeline changes are included in the implementation plan and cost.
Dashboards and reporting
- A simple, high-signal dashboard design is described (layout, key views, drill-downs).
- The dashboard focuses on value themes and key KPIs, not an exhaustive list.
- Leading indicators (adoption, usage, early quality) and lagging outcomes (revenue, cost, risk) are both represented.
- It is clear who will build and maintain the dashboard.
Cadence and forums
- Measurement and review cadence is defined (monthly vs quarterly) and tied to existing forums where possible.
- Each review forum has named roles: sponsor, benefit owners, finance, risk/controls, initiative lead.
- Implementation-phase reviews emphasize leading indicators; BAU-phase reviews emphasize realized benefits.
- Timing of reviews aligns with financial reporting cycles where relevant.
Ownership and accountability
- Each KPI has a benefit owner (role/title) accountable for its movement.
- Finance has a defined role in validating the financial translation of KPIs (e.g., savings that hit P&L).
- Benefit KPIs are, or will be, integrated into relevant scorecards and performance dialogues, not only project reports.
Variance management
- Thresholds for acceptable variance vs target are defined (e.g., ±10% for two quarters).
- There is a clear process for escalation and decision-making when KPIs deviate materially (who proposes actions, who decides).
- The case explains how benefit shortfalls may lead to course corrections, re-phasing, or stopping the initiative.
If you can tick these items with a straight face, your KPIs and tracking plan are not just a nice appendix—they are a real management system for value. That is what separates a business case that looks good on paper from one that reliably delivers the outcomes you promised.
