What Does a Project Management Office Do for a Business? A Straight Answer With Real Numbers

The short version: a project management office sets the rules for how projects get planned, tracked, and reported across a business, so leadership stops finding out about problems six weeks too late. Done well, it saves money by killing bad projects early and finishing good ones on time. Done badly, it becomes a reporting factory that adds meetings and slows everyone down without changing a single outcome.

What a PMO is, in plain terms

A project management office, or PMO, is the team (sometimes one person, sometimes forty) whose job is to make sure projects across a business follow the same standards, use the same templates, get tracked the same way, and get reported to leadership in a form that means something. It is not the people doing the project work. It is the structure that sits above them, watching timelines, budgets, risks, and dependencies across everything running at once.

Think of it as the difference between five departments each inventing their own way to run a project and one shared playbook everyone uses. Without a PMO, marketing tracks projects in a spreadsheet, IT uses a ticketing tool, and operations runs everything from memory and a WhatsApp group. Nobody at board level can answer a simple question like "which of our twelve active projects are behind schedule" without chasing five people for five different answers.

The three levels of PMO, and why the label matters

Most people who ask what a PMO does are picturing one thing, but there are three distinct versions and they behave very differently.

• Supportive PMO. Provides templates, training, and a shared repository. Advisory only. No power to enforce anything. Good for smaller businesses testing the idea.
• Controlling PMO. Requires projects to use specific methods and tools, and reviews them at set checkpoints. This is the most common setup in mid-size businesses of 50 to 500 staff.
• Directive PMO. runs the projects. Project managers report into the PMO, not into department heads. Common in construction, engineering, and large government contracts.

The mistake I see constantly is a business building a directive PMO's worth of process and reporting while only having the authority of a supportive one. That mismatch is where PMOs go to die.

What a PMO does day to day

Strip away the jargon and a functioning PMO does five concrete things, every single week:

• Keeps one master view of every active project, its budget, its deadline, and its actual status, not the status someone hopes is true.
• Runs the intake process so new project requests get scored and prioritised instead of just starting because someone senior asked for it.
• Owns the standard project management processes such as risk logs, change requests, and stage gates, so a project in finance and a project in sales get managed the same way.
• Chases resource conflicts before they become crises, flagging when three projects all need the same senior developer in the same week.
• Reports up to leadership in plain numbers: percentage of projects on time, percentage over budget, and which ones are at real risk of failing.

That last point is the one businesses skip, and it is the one that justifies the PMO's cost.

A real example: what happened when a 40-person agency built one

I worked with a marketing agency in the Midlands with around 40 staff that was running roughly 15 client projects at any given time. Before their PMO existed, project status lived in the heads of three account directors. When one of them left in the middle of a busy quarter, the business could not tell a client with any confidence when their campaign would launch.

They built a controlling PMO with one full-time PMO manager on around £48,000 a year, plus a shared project tool costing about £3,000 annually across the team. Within the first two quarters, on-time project delivery went from 41 percent to 68 percent. By month twelve it sat at 78 percent, and the number of projects that had to be discounted or written off due to late delivery dropped from six in the previous year to one.

The PMO manager did not do the client work. She sat in weekly checkpoints, flagged when a designer was double booked across two accounts, and made sure every project had a one page status update that leadership could read in ninety seconds instead of a forty-minute meeting. The cost of the role paid for itself inside four months just from the write-offs it prevented.

What it costs to run one

In the UK, a PMO analyst typically earns £32,000 to £45,000. A PMO manager sits around £45,000 to £65,000. A head of PMO or PMO director in a larger organisation can be £75,000 to £120,000, sometimes more in London or in regulated industries like financial services and pharma. Add software, which ranges from a few hundred pounds a year for a lightweight tool up to six figures for enterprise portfolio platforms used by large corporates.

For a business under 100 staff, a single PMO analyst or manager, well supported by good templates and a mid-tier tool, is usually enough. You do not need a department. You need one person with the authority to say "no" to a badly scoped project and the discipline to update the master tracker every single week without fail.

The part nobody likes to admit

Here is the uncomfortable bit. A large proportion of PMOs get quietly dismantled within two to three years of being set up, and it is rarely because project management stopped mattering. It is because the PMO measured the wrong thing. Many PMOs spend their energy proving that process was followed, that the right form was filled in, that the stage gate was signed off, rather than proving that the business made better decisions or lost less money as a result of their existence. Leadership eventually notices they are paying for paperwork, not outcomes, and the PMO gets cut in the next cost review.

A PMO that survives is one that can point to a number: projects delivered on time went up, budget overruns went down, a project was killed six months early and saved £200,000 rather than being allowed to limp on. If your PMO cannot produce that kind of sentence, it is decoration, not infrastructure.

Where AI has changed the job

The reporting side of PMO work used to eat a huge chunk of the week, chasing project managers for status updates and manually building slide decks for the board. That part of the job is shrinking fast. Automated status pulls from project tools, AI summaries of risk logs, and tools that draft the weekly leadership update from raw data are now standard in businesses that have modernised their PMO function. It is worth reading about when AI agents make sense in a business before you assume every reporting task should be automated, because some of it still needs a human judgement call, particularly around which risks get escalated to the board and which get quietly managed.

Training a PMO team to use these tools, and deciding what gets checked before it goes to a board pack, ties into a wider question a lot of businesses are wrestling with right now, which is how much AI tool use should be allowed in training and assessment for the people doing this work in the first place.

Where a PMO overlaps with money and billing

One thing that surprises people is how much PMO work involves finance, not just timelines. A controlling PMO usually tracks project budgets against actuals, flags scope creep before it becomes an unbilled cost, and in agency or consultancy settings, keeps an eye on whether project milestones have triggered invoicing. It is worth knowing, for instance, whether an invoice always has to come before payment on staged projects, because a PMO that does not understand billing triggers will happily report a project as "on track" while the business has quietly done three months of unbilled work.

Do you need a PMO, or just better habits?

Not every business needs one. If you are running two or three projects a year with a small team, a PMO is overkill; a shared tracker and a weekly fifteen-minute check-in will do the job. The signs you need a PMO look like this:

• You are running more than eight to ten projects at once and nobody has a single view of all of them.
• Projects regularly run over budget or over time and nobody can say why until it is too late.
• Different teams use completely different methods, so moving a project manager between teams means retraining them from scratch.
• Leadership finds out about a failing project from a client complaint rather than from an internal report.

If none of those apply, save the salary and fix your habits instead. If two or more apply, a PMO will pay for itself within a year, as it did with the agency above.

How to set one up without wasting a year

The businesses that get this right move fast and keep it small at first.

• Pick one person and give them real authority to enforce standards, not just suggest them.
• Choose one project tool for the whole business and retire every spreadsheet within 90 days.
• Build three templates only: a project charter, a weekly status update, and a risk log. Resist the urge to build twenty.
• Set two numbers you will track from day one: percentage of projects on time, and percentage on budget. Report both monthly.
• Review after six months and kill anything that isn't producing decisions, not just paperwork.

If you are hiring for this role, it is also worth reading honestly about whether project management still has a future as a career, because the shape of the job is shifting toward oversight and judgement and away from manual status chasing, which changes who you should be hiring for a PMO role in 2026 versus who you'd have hired five years ago.

Frequently asked questions

What is the main purpose of a project management office?

Its main purpose is to give a business one consistent, honest view of every project running at once, so leadership can spot problems early and make funding decisions based on real status rather than guesswork.

Is a PMO the same as a project manager?

No. A project manager runs a single project day to day. A PMO sets the standards and tracks the whole portfolio of projects across the business, and in a controlling or directive setup, the project managers report through it.

How much does it cost to set up a PMO in a small business?

For a business under 100 staff, expect one PMO manager at roughly £45,000 to £65,000 a year plus a project tracking tool costing a few hundred to a few thousand pounds annually. That single hire, done well, is usually enough for businesses running under 30 projects at once.

Why do so many PMOs fail or get shut down?

Most fail because they measure process compliance, like whether a form got filled in, instead of measuring outcomes, like whether projects finished on time and on budget. When leadership cannot see a clear financial benefit, the PMO gets cut in the next cost review, often within two to three years.