The Right-Sized PMO: Governance for a 50-Person Company
Say “PMO” to the owner of a fifty-person company and you can watch the shutters come down. They picture what they saw at a former employer: a department of template enforcers, a methodology binder, status meetings about status meetings. They are right to refuse that. They are wrong to conclude that governance is only for corporations.
Here is what we actually see in growing SMEs: eight initiatives running at once, nobody sure which five matter, the same three people on the critical path of everything, and projects that die quietly because no one was empowered to either finish or kill them. That is not a discipline problem. It is a missing mechanism problem, and the mechanism does not need to weigh much.
What a PMO is actually for
Strip away the corporate theatre and a PMO does three things: it makes the portfolio visible, it makes progress honest, and it makes decisions happen. At enterprise scale that takes a department. At fifty people it takes five artefacts and a rhythm, run by half a person.
The five mechanisms
1. One project list. A single page listing every initiative: owner, goal, status, next milestone. Not a tool rollout, a list. The first time an SME leadership team sees their complete portfolio on one page, someone always says “why are we doing that one?” That sentence alone pays for the exercise. Most companies discover they are running twice as many projects as they can staff, which explains why everything is half-done.
2. A one-page charter per project. Before anything starts: what are we doing, why, who owns it, what does done look like, what will it roughly cost, what are we explicitly not doing. One page, signed by whoever pays. The charter is not bureaucracy; it is the antidote to the project that mutates for a year because its purpose was never written down.
3. Honest weekly status, RAG, one page. Green, amber, red against the charter, trend arrow, decisions needed. We wrote the full rules in RAG reporting that executives trust, and they apply at fifty people exactly as at five thousand: define the colours in writing, make red mean “decide”, never punish the messenger.
4. Gates where money and scope are re-committed. Two or three per project, not twelve. At each gate the sponsor re-answers one question: knowing what we now know, would we still start this? If no, stop. A right-sized PMO kills more projects than an unmanaged portfolio does, and that is precisely its value: the killed projects release the people who make the surviving ones fast. Re-estimation belongs here too, as we argued in estimation that survives contact with reality.
5. A decision log. One running list of decisions made, by whom, when. Sounds trivial. Prevents the SME disease of re-litigating the same decision every six weeks because nobody remembers deciding it.
Who runs it
Not a hire, at first. Half of one respected person’s time: someone senior enough to chase an owner for status and be answered, organised enough to keep the list true. The moment it becomes a full department with its own templates team, you have rebuilt the thing your people fled from. The PMO serves delivery; the ratio of governance effort to project effort should embarrass no one.
External help has one legitimate role here: setting the machinery up, calibrating it to your size, and coaching the internal owner for the first quarter. That is a bounded engagement, not a dependency, and it is exactly what we do inside our IT project management services. Most of the failed projects we dissected in why IT projects fail did not lack talent. They lacked these five pages.
Frequently Asked Questions
Does a small company need a PMO?
It needs the functions of one: portfolio visibility, honest status and decision-making. At around 50 people that takes five lightweight artefacts and roughly half a person’s time, not a department.
What is a right-sized PMO?
Governance proportional to the organisation: a single project list, one-page charters, weekly RAG status, two or three gates per project and a decision log. Everything else is added only when a real problem demands it.
What is the first step to setting one up?
Write the complete project list. In our experience the list alone triggers the most valuable conversation a growing company can have: which of these should we stop?
Editorial note: This article was researched and drafted with the assistance of Claude (Anthropic), and reviewed and approved by Amazing Projects before publication.
