Two PMOs I have been, and the one I will not be again.

I have been two kinds of programme office in my career, and there is a third I keep getting hired to undo.

I want to write about which one I would build again, and which one I would not, because the answers are not the ones I expected at twenty-five.

The first one. The scorekeeper.

The first PMO I ran was a beautiful piece of machinery. Spreadsheets that fed dashboards that fed exec packs. Weekly status, monthly steering, quarterly review. RAID logs that were updated on schedule. A RACI matrix that any auditor would smile at.

The team called it the scorekeeper. They were right.

It did not make anything go faster. It made everything go visible. The throughput of the programmes I ran was not higher than the throughput of the unmanaged ones around me. The reporting cadence was much, much higher.

I was promoted on the strength of those dashboards. I would do it differently now.

The second one. The air-traffic controller.

A few years later I ran a programme that was actually under pressure. We had a vendor SLA penalty that was triggered weekly. The team was 22 people across four geographies. The work was harder. The room was smaller.

So I rebuilt the office around a single question: what is in motion, and what is stuck?

We dropped the weekly steering. It became an exception-only meeting, called within four business hours of anything stuck-for-more-than-48-hours. The RAID register became a thing that lived in the same place the team worked, not a separate ritual. The status update became a Slack message, written by the engineer who finished the thing, not by me. I wrote one paragraph per week, for the sponsor, and it was always specific and almost always under a hundred words.

The programme ran 18% under budget and a month early. I do not think any of those numbers had anything to do with my dashboards.

That is the PMO I would build again.

The third one. The dismantling.

The third PMO is the one I keep getting hired to undo.

It is the scorekeeper grown to monstrous size. Twenty-eight templates. Three governance committees. A separate “PMO of the PMO” doing meta-reporting. The organisation has confused “amount of process” for “quality of delivery” and has built a small bureaucracy to defend the confusion.

When I am called in, two things have usually already happened. Engineering velocity has dropped to a number people are embarrassed about. And the PMO has produced a report explaining why the drop is not their problem.

The work I do then is not technical. It is not strategic. It is a slow, polite, well-documented dismantling. I remove templates one at a time, and I make sure the dashboard that visualises their removal is the only dashboard the leadership sees.

In every one of these engagements, the organisation’s delivery throughput improves within a quarter.

What I would tell my twenty-five-year-old self

If I could send a note back, it would say something like this. The org will reward you for building the scorekeeper, but the org will need you to be the air-traffic controller, and the only thing in either job that matters is the one question. What is in motion, and what is stuck?

The bureaucracy is the consolation prize for being unable to answer it.

The good PMO is the one that does.