I've sat in the boardroom when a transformation programme is declared a success. Milestone met, go-live achieved, champagne ordered. And I've returned to that same organisation eighteen months later to find the expected revenue lift hasn't materialised, the new system is being worked around rather than used, and the PMO that delivered the whole thing has quietly been disbanded.
This happens more often than anyone in our industry likes to admit.
According to Gartner, 60% of PMOs are shut down within three years of being established. Forrester puts the number even higher, at up to 75%. These aren't small organisations with limited budgets. They're enterprises with access to top talent, premium tooling, and experienced consultants. And they still fail, repeatedly, for the same handful of reasons.
I've spent my career recovering programmes that were supposed to have already succeeded. What I've found is that PMO failure is rarely a mystery. It's a pattern. And once you know what to look for, you can see it forming long before the shutdown notice arrives.
Here's what that pattern looks like, and what it actually takes to break it.
The Mistake Most PMOs Make Before They've Even Started
The most common failure I see isn't a delivery problem. It's a design problem.
PMOs are often stood up to solve an immediate pain point: a programme running over budget, a board that wants more visibility, a regulator asking for tighter controls. So the PMO gets built around that pain point. It becomes a reporting function, a governance layer, a compliance mechanism.
And then, six months in, the delivery teams start routing around it.
The part most coverage misses: 68% of organisations have no formal way to prioritise projects or link them to strategy, according to research by Business Improvement Architects. The PMO isn't failing at delivery. It's delivering the wrong things, and nobody built in a mechanism to catch that.
Built for Control, Not for Outcomes
When a PMO is designed primarily as a control function, it creates a predictable dynamic. Delivery teams see it as overhead. Executives see it as a cost centre. And when the business comes under pressure, it's the first thing cut.
The PMOs I've seen survive and thrive are built differently. They're positioned as strategic enablers, not process police. Their value is measured in outcomes delivered, not reports produced.
The difference in results is stark. According to PMI's research, organisations with a strategic PMO are more than 50% more likely to have their projects finish on time and within budget compared to those without one. But that only holds if the PMO is actually connected to strategy in the first place.
The Reporting Trap
Here's a pattern I encounter constantly. A PMO gets stood up and immediately builds a dashboard. Red, amber, green statuses. Weekly reports. Executive packs. It looks like governance. It feels like control.
But a dashboard that tells you what happened last week is not the same as a governance structure that influences what happens next week. When I ask PMO leaders "what decision did your last report enable?" the answer is usually silence.
A PMO that only reports status becomes irrelevant. A PMO that interprets data and drives decisions becomes indispensable.
The Four Failure Modes I See on Every Recovery Engagement
When I'm called in to recover a stalled programme or restructure a failing PMO, I run a rapid diagnostic in the first two weeks. Across dozens of engagements, the same failure modes appear with remarkable consistency. Not all four are present every time, but at least two almost always are.
1. Junior Staff Running Senior Decisions
This one is more common than it should be. A PMO is resourced with capable coordinators and analysts, but nobody with the authority or experience to push back on a sponsor, challenge a vendor's delivery claim, or make a call when two workstreams conflict.
The result: escalations pile up, decisions get deferred, and the programme loses weeks to things that should have been resolved in a single conversation. On large IT programmes, McKinsey and Oxford research found that every additional year a large project runs increases cost overruns by 15%. Slow decisions are expensive decisions.
The fix: Senior-led delivery. Not senior oversight from a distance, but senior practitioners embedded in the programme, accountable for outcomes.
2. Governance That Exists on Paper Only
Most failing programmes have governance frameworks. They have RACI matrices, steering committees, change control boards. What they don't have is governance that actually functions.
I've reviewed steering committee packs that were 80 slides long and answered no questions of substance. I've sat in change control meetings where every request was approved because nobody wanted to be the person who said no.
Real governance is lean, decisive, and uncomfortable. It means saying no to scope creep when the business is pushing for it. It means escalating to the board when a vendor is underdelivering. It means having the conversation nobody wants to have, three months before it becomes a crisis.
3. Vendor Relationships Without Accountability
Enterprise transformations almost always involve multiple vendors. System integrators, software providers, managed service partners. And almost always, the accountability between them is blurry.
When something goes wrong, each vendor points to the others. The PMO, if it's not set up to manage this, becomes a referee with no authority. Delays compound. Costs escalate. The client organisation absorbs the impact.
What good looks like: Clear contractual accountability, integrated delivery planning across vendors, and a PMO with the commercial awareness to know when a vendor is managing their contract rather than managing delivery.
4. Change Management Treated as an Afterthought
According to Prosci's research, 65% of organisations cite poor change management as the top reason for project failure. In my experience, that's not because organisations don't know change management matters. It's because they treat it as something that happens after delivery, not something that runs alongside it.
I've seen ERP implementations go live on time and on budget, only to find adoption rates below 30% six months later. The system worked. The change didn't land. And the business benefits never materialised.
Change management isn't a communications plan. It's a structured approach to shifting behaviour across an organisation, and it needs to start on day one of the programme, not day one of go-live.
What a PMO Recovery Actually Looks Like
When a programme is stalled or a PMO has lost the confidence of its stakeholders, the instinct is often to add more process. More governance meetings. More reporting. More consultants producing more frameworks.
That's almost always the wrong call.
Recovery is about removing friction, not adding structure.
The first thing I do when I take on a recovery engagement is a rapid diagnostic: two weeks, no deliverables, just conversations and observation. I'm looking for where decisions are getting stuck, which relationships are broken, and what the real blockers are versus the stated ones.
The Rapid Recovery Framework
Based on that diagnostic, recovery typically moves through three phases:
Stabilise: Stop the bleeding. Identify the two or three issues causing the most damage and resolve them immediately. This might mean replacing a vendor relationship, restructuring the steering committee, or simply getting the right people in the same room for the first time.
Restructure: Rebuild the PMO operating model around outcomes, not outputs. This means redefining what the PMO is accountable for, who it reports to, and how its value is measured. PMI research consistently shows that PMOs with executive-level reporting lines deliver materially better portfolio outcomes than those buried in IT or Finance.
Sustain: Embed the changes so they outlast the recovery engagement. The goal is never to create dependency on external support. It's to build internal capability and governance that holds when we leave.
The Numbers Behind Recovery
The case for investing in proper PMO recovery isn't difficult to make.
| The cost of doing nothing | The data |
|---|---|
| Large IT projects run an average of 45% over budget | McKinsey / Oxford, 5,400+ projects analysed |
| 70% of digital transformations fail to meet their objectives | McKinsey enterprise transformation research |
| Every additional year a large project runs adds 15% to cost overruns | McKinsey / Oxford |
| Organisations waste 9.9% of every dollar invested in poorly run projects | PMI Pulse of the Profession |
For an enterprise running a £200 million project portfolio, that last figure alone represents £20 million walking out the door annually. A recovery engagement that costs a fraction of that and stops the bleed is not a cost. It's a return.
The Question Worth Asking Right Now
If you're reading this because something feels off with your programme or your PMO, trust that instinct.
The warning signs are usually visible well before a programme formally stalls. Decisions that take longer than they should. Steering committees that feel performative. Delivery teams that have stopped bringing problems forward because they've learned nothing gets resolved. A growing gap between what the programme reports and what people privately believe.
These aren't soft concerns. They're early indicators of structural failure.
The honest question to ask: Is your PMO set up to deliver the business outcome you were promised, or is it set up to deliver the programme on paper?
If you're not certain of the answer, that's worth a conversation. At Prime Project Management, we work exclusively with senior-led delivery, which means every engagement is run by a practitioner who has done this before, not a team of analysts working from a methodology deck.
If your programme is stalled, your PMO is underperforming, or you're about to embark on a large-scale transformation and want to get the structure right from the start, get in touch with our team. We'll tell you honestly what we see and what we'd do about it.