PMO is dead — development teams don't need control from outside

PMO is dead. I'm writing it as bluntly as that on purpose, because the topic is too important to wrap diplomatically. Development teams no longer need an external control body that prescribes how they should work. Good teams don't need it anyway, and PMO won't make the bad ones any better. And there's barely any space in between.

I have sat through more meetings with project managers, product owners and project offices than is healthy. I have seen teams that flourished thanks to good management, and I have seen teams that suffocated under the standard apparatus of a classic PMO. The difference was never the name of the role. The difference was whether the people filling that role were taking work off the team or piling it on.

What PMO originally wanted

The classic Project Management Office grew out of a sensible idea. In large companies with many parallel projects there was chaos. Budgets ran out of control. Dependencies were unclear. Priorities were not set. So a body was created to define standards, weigh up risks, plan portfolios and consolidate reports. In the world of the 1990s and early 2000s, this was often the only possible answer.

The problem is: we no longer live in that world. Development teams today have tools that deliver transparency almost for free. Pull requests are traceable, deployments are documented, issue trackers show progress in real time. A manager whose job is to create transparency walks, in many companies, into a room where all the windows are already open. They then produce second-order transparency — transparency about the transparency — and that is precisely the moment in which bureaucracy is born.

Where PMO actually does damage today

I see three concrete patterns in which PMO does damage. First: status theatre. Teams have to build slides every week explaining what is already in their tickets. That ties up developer time that would otherwise be productive, and it forces everyone involved into a language that was never designed for building.

Second: decision logjam. If three committees sit between the team and the actual decision, decisions become slow and defensive. That is the opposite of what a good engineering team needs. Engineers solve problems best when they are close to the problem and bear the consequences. A decoupled PMO decouples precisely that chain.

Third: diffusion of responsibility. In teams with a classic PMO, it eventually becomes unclear who is responsible for what. The team says the PM is responsible. The PM says the business is responsible. The business says the team has to decide. The result is that nobody really decides any more, and projects slide into a state where a lot is discussed and little is built.

What helps instead

I have no time for railing against something without naming an alternative. What I have seen in functioning teams over recent years is a different kind of structure. Self-organising teams with clear outcome ownership. One or two people in engineering management whose job is not to control, but to clear obstacles. A thin but binding layer that brings in company context and protects teams from themselves when it has to.

This is less of a 'PMO' in the old sense and more of an 'enablement office'. Its job isn't to tell the team what to do. Its job is to make sure the team can do what it does. To defend the budget. To resolve blockages outside the team. To translate company strategy into a language the team can connect to.

That is a different mindset. And yes, it requires people with different skills than the classic PMO. It requires calmness, trust and a very clean handling of power. That is rare, and that is why good engineering management is rare too.

My position

For many years I have worked in structures where teams have as much autonomy as possible. That is not laissez-faire. That is strict discipline in other places. Clear goals, clear roles, clear expectations. But within that frame the team decides. And the team delivers.

When I support companies that want to move in this direction, the first step is often not 'we're introducing a new framework'. The first step is to see how much time, energy and morale is currently being burnt on rituals nobody needs any more. Clearing that out is the most underestimated leadership achievement of our time.

The term PMO will still be around. Some people will fill it with sensible life. For me, the classic version is dead. And I don't find that sad. I find it overdue.