The short answer
A board needs a small number of measures that answer three questions: are we on track, what will it cost to finish, and what needs a decision from us? Around eight KPIs is usually enough. Keep the measures that show whether the PMO itself is performing in a separate, less frequent report.
Fewer measures, better decisions
“Boards don't need thirty measures. The PMO reports I've seen work best answered three questions: are we on track, what will it cost to finish, and what needs a decision from us?”
A long list of KPIs feels thorough, but it hides the few numbers that matter. Each measure on the board report should lead to an action when it moves. If it wouldn't, it belongs somewhere else.
The board KPIs, grouped by question
Are we on track?
- Projects on schedule. The share of active projects at green, amber and red, with the reasons for anything red.
- Key milestones met. Milestones due this period that were hit, and those that slipped.
- Forecast finish against plan. For each major project, the expected completion date compared with the approved one.
What will it cost to finish?
- Spend against budget. Actual cost to date compared with the budget for the same point.
- Forecast cost to completion. What each project is now expected to cost in total, compared with what was approved. This is the earliest warning of an overspend.
- Benefits on track. Whether the projects that have finished are delivering what their business case promised.
What needs a decision from us?
- Decisions and escalations waiting. What the board needs to decide, and how long each item has been waiting.
- Top risks and issues. The few portfolio-level risks and issues that need board attention, each with an owner.
- People over-committed. Key people booked beyond their capacity, which is often the hidden cause of slipping dates.
Measures that show the PMO is working
These matter, but they don't need to go to the board every month. Review them quarterly:
- Reports on time. The share of projects reporting by the deadline, using the agreed template.
- Standards followed. The share of projects with an approved business case, initiation document and current RAID log.
- Time to approve. How long a proposed project takes to get a go or no-go decision.
- PMO maturity. Progress measured against a structured health check, repeated once or twice a year.
Making the KPIs useful
- Set thresholds that trigger action. Agree in advance what turns a measure amber or red, so the discussion is about what to do, not whether there's a problem.
- Prefer forward-looking measures. Forecast cost and forecast finish dates warn you early. Spend to date only tells you what has already happened.
- Take the numbers from one source. If the KPIs are rebuilt by hand each month, they'll be late and argued over.
About the author
Stuart Lester
Founder and CEO
Reviewed by Martin Brodie-Lynch