Why the PMO Keeps Dying — and What Should Take Its Place
The office had counted everything it did and measured nothing it was for.
The meeting where the office dies
The programme office rarely dies in a crisis. It dies in a budget review, on an ordinary Tuesday, when someone with a spreadsheet asks what the people in it actually do.
I have sat in that meeting. The head of the office answered the only way the office had taught him to answer — with its outputs. Template compliance across the portfolio stood at ninety-four per cent. The weekly status pack ran to forty-two pages and reached the executive every Friday without fail. Every project had a risk log, a RAG rating, a resource plan. By its own measures the office was an unqualified success.
Then the finance director asked the question no volume of reporting survives. He did not ask whether the reports were produced. He asked which decision, in the past year, any of them had changed. The room went quiet. The office had counted everything it did and measured nothing it was for.
This is how the programme management office usually ends: not condemned, simply unable to justify itself. And it is worth understanding why, because the same office is already being built again somewhere else — with the same good intentions, and heading for the same afternoon in front of the same spreadsheet.
A short and repeating life
The story is familiar enough to be a genre. An organisation is burned by a programme that overran, or a regulator asks how decisions are controlled, or a portfolio grows past the point where any one person can hold it in their head. Out of that discomfort comes a decision to impose order, and the office is born — usually with real energy and a genuine mandate.
In its first year it earns its keep handsomely. It gives a common language to projects that had none. It makes the portfolio visible for the first time, so that a director who was flying blind can finally see the shape of what he owns. Nobody argues with it, because everybody remembers the mess it was created to end.
Then the predictable thing happens. Having been rewarded for imposing standards, the office keeps imposing them. Three people become fourteen over four years. The template pack grows because every past failure leaves behind a new field to complete. The reports lengthen, because length reads as diligence. A gate that once forced a hard conversation becomes a form to be countersigned. Slowly the office stops being the thing that helps the work and becomes the thing the work must satisfy. By the time the budget review comes round, it has become overhead that defends itself in the language of overhead — and that is a losing argument in any room where money is short.
The death is rarely dramatic. Underneath it, though, the same handful of causes recur:
- It measures its own activity, not the decisions it improves. Reports produced, templates completed, meetings held — all countable, none of them the point.
- It becomes the keeper of the method rather than the servant of the mission. Compliance with the process turns into the goal, and the process outlives the reason it existed.
- It sits furthest from the work and closest to the rules. The people writing the standards are the people least exposed to the consequences of following them.
- It cannot show the counterfactual. An office that prevents bad decisions has, by definition, nothing to point to — and so, when it is asked to prove its worth, it points to its paperwork.
An office that measures itself by the reports it produces will always look busy and never look necessary. The two are not the same, and the gap between them is exactly where the budget cut lands.
The reporting factory
Nowhere is the drift clearer than in what most offices spend most of their time doing: reporting. It is worth being blunt about how little of it informs anything.
Consider the weekly pack — the forty-two pages, two analysts, three days to assemble. Ask what decision it is built to support and the honest answer is usually none. It is built to be complete, to withstand the question “why isn’t this in here?”, and completeness is the enemy of attention. An executive given forty-two pages reads the summary and trusts the colour coding, which brings us to the RAG status: the most quietly dangerous instrument in the whole apparatus.
A status that is green on the outside and red the moment you cut into it is not an accident; it is what the system rewards. A manager who reports amber invites scrutiny, a recovery plan, an uncomfortable meeting. A manager who holds at green is left alone. So status drifts green and stays there — until the week before a slip that everyone close to the work had seen coming for a month. I have watched a programme report green for eleven consecutive weeks and then move to red with a six-week delay already baked in. The reporting had not failed to detect the problem. It had been carefully arranged not to.
“A report that no decision depends on is not information; it is insurance, bought with other people’s attention.”
None of this is the fault of the analysts, who are diligent people doing exactly what the office asks. It is the fault of an office that has confused producing a record with exercising judgement.
The case for the defence
It would be easy, and wrong, to conclude from all this that governance is the problem and that delivery should be left to look after itself. The strongest case for the standing office is a serious one, and it deserves to be met head on.
It runs roughly as follows. Complex portfolios need independent assurance — someone whose loyalty is to the whole, not to any single project fighting for its life. They need a single, trusted version of the truth, so that two directors are not arguing from two different sets of numbers. They need standards, so that a manager moving between programmes is not relearning the basics each time. And in a climate of tighter control and sharper questions from auditors since Sarbanes-Oxley, they need a defensible trail showing that decisions were controlled and not merely made. Every one of those needs is real. I have seen programmes fail for the lack of each of them.
The reply is not that the needs are imaginary. It is that we have met real needs with the wrong kind of thing. We took a set of genuine functions — assurance, a portfolio view, common standards, an auditable trail — and we built an institution around them: a headcount, a permanent box on the organisation chart, a budget line to defend. And once a thing is an institution, it measures itself by the only things institutions can easily measure about themselves — how much process they maintain and how fully it is complied with. The needs were real. The monument was the mistake.
A service, not an institution
If the office should not be a monument, what should it be? The answer is not another box on the chart with a better mandate. It is a different kind of thing altogether — a service to the organisation’s judgement, held to the standard of whether that judgement improved. Five principles separate the one from the other.
- Own it close to the work. The people providing assurance should be near enough to delivery to smell when something is wrong, and senior enough that their questions carry weight. Governance exercised from a distance becomes governance by template; governance exercised up close becomes conversation, which is the only form of it that ever changed a decision.
- Measure it by decisions, not documents. The right question for any part of the function is the finance director’s question: which decision did this improve, and how would we know if it had not? A report, a gate, a review that cannot answer it should be stopped, not refined.
- Make it temporary by design. A portfolio has seasons; the support it needs in the thick of a transformation is not what it needs a year later. Constitute the function for the need in front of you and be willing to stand it down when that need passes. An office built to be permanent will find reasons to remain long after its reasons have gone.
- Staff it with people who have delivered. The credibility of assurance rests entirely on whether the person giving it has ever carried the thing they are assuring. A function staffed by administrators of the method will enforce the method; a function staffed by people who have delivered will use the method where it helps and set it aside where it does not.
- Let each report serve a decision or die. Start from the decisions the leadership actually has to make and build the smallest reporting that serves them. Everything that survives that test earns its place. Everything that does not is the forty-two pages.
The distinction is easier to see set side by side:
| The office as institution | The function as service |
|---|---|
| Measured by process maintained and complied with | Measured by decisions improved |
| Permanent box on the organisation chart | Constituted for a need, stood down when it passes |
| Staffed by administrators of the method | Staffed by people who have delivered |
| Reports built to be complete | Reports built to serve a decision |
| Sits above the work, enforcing | Sits alongside the work, helping |
None of this abolishes assurance, standards, or a portfolio view. It keeps every function the standing office was created to provide and discards only the institution that grew up around them. That is the whole move: keep the service, lose the monument.
The good ones were always temporary
The programme office does not need saving. It needs to stop being built as a monument to control and start being run as a service to judgement — and to be held, without sentiment, to whether that judgement is any better for its existence.
The best offices I have worked alongside understood this instinctively. They were small, senior, and slightly impatient with their own machinery. They measured their success not by how indispensable they had become but by how little they were needed by the end — by how much good judgement they had left behind in the teams rather than hoarded in the centre. They were, in the best sense, temporary. When the next budget review came, nobody had to ask what they were for, because everyone in the room could name a decision that had gone better because they were there.
That is the office worth building. Not the one that survives the spreadsheet, but the one that never has to face it.