Setting up a PMO: the decisions that matter in the first 90 days

Most PMOs that fail do not fail on templates. They fail because nobody agreed what the PMO was for. These are the decisions I make first when I set one up, in the order I make them.

Agree the mandate

A PMO can be supportive (templates, coaching, reporting help), controlling (standards that projects must follow) or directive (it manages the projects itself). Each is legitimate. Choosing one with the sponsor, in writing, prevents months of friction.

Build the inventory

List every project and initiative that is running or approved, with owner, budget, dates and status. In most organizations this list does not exist in one place. Producing it is often the PMO's first piece of real value.

Set decision rights

Define who approves a new project, who can change scope or budget, and who can stop work. Light stage gates with clear criteria do more than a thick methodology document.

Choose one reporting rhythm

One cycle, one format and a short set of measures: schedule, cost, risk, and benefit where it can be tracked. Leaders should get the same view every time, so the trend becomes the message.

Standardize only what saves effort

Introduce a template when it removes work, such as a status report that feeds the dashboard automatically. Skip templates that only add work.

Invest in people

Coaching project managers through their first stage gates and certifications builds the capability that keeps the PMO alive after its first sponsor moves on.

Deliver a visible win early

Pick a problem leadership already feels, such as a portfolio-wide dashboard or a recovery plan for a struggling project, and solve it within the quarter. Credibility comes before standards.