Why the PMO Grew but the Value Didn’t
Fourteen people, forty tabs, and the decision that governed the programme's fate was being taken on a number nobody in the room believed.
The Office That Grew While the Work Stood Still
Walk the floor of almost any large programme just now and you will find the same quiet monument to good intentions: the programme office. Two years ago it was a planner and a part-time administrator working out of a corner. Today it is fourteen people. It owns a governance calendar, a risk-and-issue process, a benefits-tracking spreadsheet with more tabs than anyone has opened this quarter, and a weekly status pack that lands in a hundred inboxes on a Friday afternoon and is read, in any meaningful sense, by almost none of them. Every one of those additions was justified when it was made. And yet if you ask the plain question — is the programme delivering faster, or more surely, because the office grew from two people to fourteen? — the honest answer, in a great many cases, is no.
That gap is the subject worth dwelling on. Not the failed PMO — those are easy to see and easy to blame — but the successful one: well run, well liked, staffed by conscientious people, growing steadily, and adding less value with every increment of its growth. It is one of the more expensive patterns in change work today precisely because it does not look like a problem. It looks like control.
What the Office Is Supposed to Do
A programme office earns its keep by doing for delivery what delivery cannot easily do for itself. It holds the single coherent plan when a dozen workstreams each hold their own. It makes risk and dependency visible across boundaries no individual manager can see over. It gives the sponsor an honest, aggregated picture so that decisions rest on something firmer than the loudest voice in the room. Done well, that is a genuine multiplier: a small, sharp office can make a large programme materially more likely to land.
The word that matters in all of that is service. A good office exists to serve the delivery of the thing. Its customer is the programme, and everything it produces should be traceable to a decision the programme needs to make or a risk the programme needs to see. Hold on to that test — traceable to a decision — because almost everything that goes wrong with programme offices is a slow loss of it.
A programme office is a service to delivery, not a layer above it. The moment its customer becomes the reporting line rather than the programme, its growth stops tracking its value.
Why It Grows Past Its Usefulness
The trouble is that a programme office grows by a logic almost entirely disconnected from the value it adds. Demand for reassurance is effectively infinite, and the office is the organisation’s instrument for producing it. So it accretes.
- Every incident spawns a permanent report. A problem surfaces that a report would have caught, so a new report is commissioned. The problem passes; the report remains, produced forever after by someone whose week now includes it. Reports are added far more readily than they are ever retired.
- Every framework brings a new role. As each method and standard is adopted in turn, each arrives with its own artefacts, its own board, its own custodian. The office becomes the place where all of these live, and each one is another desk.
- The office cannot say no. When a senior stakeholder asks for a new cut of the data, the office has neither the standing nor the appetite to refuse. Saying yes is safe and immediate; questioning whether the cut will change any decision is awkward and slow. Multiply that small asymmetry across two years and you have fourteen people.
- It is measured by its activity, not its effect. The office reports on the programme, but rarely is anyone reporting on the office. Its health is judged by whether the pack went out on time and the templates were complete — by throughput, not by whether any of it moved a decision. What is measured by activity will grow by activity.
Let me make it concrete. A transformation programme I have in mind — a composite, but true to the pattern — ran a PMO that grew from three people to fourteen over about two years. Its flagship artefact was a status workbook of some forty tabs, refreshed weekly, consuming the better part of two full-time roles just to assemble. It was a genuinely impressive object. Yet the one number the steering committee actually needed to steer by — a trustworthy forecast of cost to complete — was still wrong, because it was rolled up from workstream figures that each carried their own optimism, and no amount of formatting corrected the optimism underneath. The office had scaled the production of information almost perfectly while leaving the reliability of the one figure that mattered untouched. Fourteen people, forty tabs, and the decision that governed the programme’s fate was being taken on a number nobody in the room believed.
That is the shape of it. The office was not lazy or incompetent. It was busy, and its busyness had become the point.
The Case for the Big Office — Taken Seriously
There is a real counter-argument here, and it deserves better than to be waved away, because the people who make it are usually right about the danger they fear.
Strip the office back too far, they say, and you lose exactly the things it exists to protect. On a genuinely large programme, coordination is not a luxury; someone has to hold the master plan, reconcile the dependencies, and give the sponsor a single version of the truth, or a dozen workstreams will each optimise locally and the whole will drift. Assurance is not bureaucracy; it is how an organisation catches the failure before it becomes a headline. And in regulated work, where an external examiner may one day ask to see the audit trail, the documentation the lean enthusiast sneers at is not overhead at all — it is the evidence that keeps the programme, and sometimes the organisation, out of trouble. All of this is true.
But notice what it argues for. It argues for coordination, assurance, and a reliable trail — every one of which is a service to a decision. It does not argue for fourteen people, forty tabs, or a hundred-inbox distribution list. The strong case for the office is a case for what the office does when it is doing its job. It is not a case for the office growing without limit, and the two are constantly and conveniently confused. The answer to “we clearly need a PMO” is yes; the answer to “therefore a larger PMO is a safer one” is no. Almost every bloated office was built by treating the first yes as though it settled the second question too.
“The strong case for a programme office is a case for what it does, never for how big it is. We keep mistaking the second for the first.”
Keeping the Office Honest
The discipline that keeps a programme office valuable is not, in the end, about size. It is about keeping every part of it tied to a decision. A few habits do most of the work.
- Name the customer, and let the customer refuse. Every report and artefact should have a named recipient who uses it to make a decision — and who is allowed to say “I don’t need this”. A report no one will decline is a report no one is really using. The right to refuse is what turns a distribution list back into a service.
- Retire an artefact for every one you add. Reports accumulate because adding is easy and removing is nobody’s job. Make removal somebody’s job. Once a quarter, put the whole catalogue on the table and ask of each item what decision it changed; the ones that changed none come off the list.
- Kill the report that never alters a decision. The forty-tab workbook is not a sign of rigour. It is a sign that production has outrun purpose. If a cut of the data has never once changed what anyone did, its cost is pure, and its comfort is the only thing it delivers.
- Measure the office by the delivery it enables, not the paper it produces. Judge it on the accuracy of its forecasts, the speed with which decisions get made and unblocked, the risks it caught early enough to matter. An office measured on outcomes stays small and sharp. An office measured on throughput grows to fourteen.
None of this is an argument against the programme office. It is an argument for the office remembering what it is for. The best I have worked with were not the largest; they were the ones that could tell you, for every artefact they produced, exactly which decision it served — and would quietly stop producing anything that failed the test. They stayed small not because they were starved, but because they were disciplined, and the discipline was the value.
The empire is the easy path. It grows on its own, funded by everyone’s entirely reasonable appetite for reassurance, and it will keep growing until someone asks the awkward question of each new report and each new hire: what decision does this change? Ask it early and often, and the office stays a multiplier. Stop asking, and you will look up in two years to find fourteen people producing a beautiful account of a programme that is no better run than when there were two.