Take a manufacturer of industrial equipment that is replacing the ERP software in its five plants. The programme brings together a common solution, a data migration, five plant roll-outs and support for the teams. Before committing the budget, the executive committee asks the sponsor to present the programme governance.
Programme governance commonly refers to the set of rules that determine who decides what in a programme, at what level, on what information and within what timescale. We propose writing it down before the teams are mobilised, in a charter signed by the sponsor, and allocating the decisions in it among four levels, from the project manager to the body that funded the programme. The charter also sets the thresholds that escalate a decision, and provides for two cases that it is tempting to leave unsaid at launch: disputes between projects and closure. Each section takes up a question that such a committee might ask.
Which decisions should programme governance assign?
Describing governance by the list of its bodies does not say what is decided in them. We propose starting from the inventory of decisions that the programme will have to take, and opening a body only if one of them needs it.
The decisions specific to the programme are those that no project can take for the others:
- the order and dates of the stages, and moving from one to the next;
- assigning work that belongs to no project;
- using the programme’s budget reserve and schedule margins;
- the date of a common milestone, such as a cutover;
- arbitration between two projects in disagreement;
- revising the expected benefits;
- closing a project or the programme.
The rest (organising a team, choosing a solution within a scope, detailed schedule) remains with the project. Our page on programme management describes the roles between which this boundary runs.
Who decides, and up to what threshold?
Each decision in the inventory is given a level, and a threshold beyond which it escalates. The table gives a possible allocation, which we propose as a starting point.
| Level | Decides alone | Escalates when |
|---|---|---|
| Project manager | What remains within their scope, budget and committed dates | A date expected by another project, or their cost or schedule tolerance, is threatened |
| Programme director | Arbitration between projects, use of the programme’s margins and reserve | The date of a stage, the programme envelope or an expected benefit is affected |
| Programme board, chaired by the sponsor | Moving to the next stage, revising the trajectory, closing a project | The envelope or the programme’s reason for being is called into question |
| Body that funded the programme | The envelope, continuation or closure of the programme | It is the last level |
The threshold is expressed through its consequences. A project manager may postpone an internal delivery; as soon as another project is waiting for it on a fixed date, the decision is no longer theirs. Management by exception in the PRINCE2 method, which deals with the project, rests on this mechanism: the project manager acts within tolerances and raises an alert when they foresee going outside them. We propose applying it at each tier of the programme.
Two clarifications make the table usable: each threshold carries a value, set in the charter, and each level has a response time, at the end of which a request left unanswered escalates one level.
On what information does the programme board rule?
The programme board rules on a file that it does not have to piece together during the meeting. We propose bringing together in it:
- the overall schedule, reduced to decision milestones and dependencies, as described in our article on the master schedule;
- dependencies between projects whose need date is threatened;
- risks common to several projects;
- the status of the expected benefits and the assumptions on which they rest;
- requests for arbitration, each presented as a choice between options whose cost and consequences are written down.
Governance also settles how this file is produced. The programme office consolidates it from what the project managers report, according to common definitions; our page on setting up a PMO explains how to build this office. The file reaches members on a fixed date, before the meeting, the running of which is the subject of our article on running a steering committee. And a project manager who reports a difficulty early should not be treated more harshly than one who keeps it quiet.
How is a dispute between two projects resolved?
Two projects in the same programme may each defend a well-founded position: one wants to freeze an interface to meet its date, the other needs to change it to deliver its scope.
In the path we propose, the dispute is first dealt with between the two project managers, with a deadline. After that date, the programme director decides. Their criterion has been written since launch: the order of priority that the sponsor has set between the dates of the stages, the scope and the cost of the programme. Without this order, every arbitration can reopen the debate on what the programme is trying to achieve.
The decision is written down, with its consequences for the project that gives way. Its baseline dates and budget are revised, so that it is not counted as late because of a decision taken for the benefit of the whole.
Some disputes go beyond the programme. When two projects compete for a skill that other projects in the organisation are also waiting for, the programme director has no authority over it, and the request goes to the body that allocates capacity, that of project portfolio management.
How is the closure of a project, or of the programme, decided?
A charter that is silent on closure leaves this decision with neither level nor criterion on the day it arises. We propose preparing closure from launch, through three entries in the charter. The moments when the question is raised automatically: the end of each stage, when the board decides to continue, adjust or stop. The facts that raise it outside these moments, such as an expected benefit that disappears or a structuring assumption that is disproved. The level that decides: the programme board for a project, the body that funded the programme for the programme itself, on the sponsor’s proposal.
The file compares what remains to be spent with what can still be obtained; sums already committed do not enter into it. It also says what becomes of what has been achieved: what is put into service as it stands, what goes back to operations, what happens to contracts in progress. If the board doubts the information presented, an independent review can precede the decision, as described on our page on project audit.
Normal closure is also a dated decision, which transfers to the business departments the tracking of the benefits still to be realised.
Where are these rules written, and when should they be reread?
They are written in the programme charter, which each project manager receives: inventory of decisions, table of levels and thresholds, dispute path, closure rules, calendar of meetings. The form is proportionate, and a programme of three projects can make all its decisions in a single body.
The charter is reread at each stage, and as soon as practice departs from it. The gap shows in facts: a programme-level decision taken in a project committee, an arbitration obtained by going directly to the sponsor, a threshold exceeded without anyone having been notified.
Faced with each of these facts, the programme board has a choice to make: restate the rule, or rewrite it in the charter.