In my experience it's almost always a relationship issue dressed up as a governance issue. If people bypass the PMO, it's usually because they don't see the value, or worse, they see the PMO as a gatekeeper that slows things down.

The most effective thing I've done is make the PMO genuinely useful to those stakeholders first. Once they start coming to you because you help them, the bypassing largely stops on its own.

Governance without teeth is just paperwork. If there are no consequences for bypassing the PMO, not punitive, but structural, people will keep doing it.

We tied PMO sign-off to budget release. Not as a power play, but as a way of ensuring resource conflicts get caught early. It changed behaviour quickly. The key was framing it as protection for the stakeholder, not a hurdle.

Both of the above are true simultaneously, which is what makes it hard. You need the governance structure AND the relationships.

One practical thing: we introduced a lightweight "project registration" step that takes five minutes and gives stakeholders something in return, visibility on resource availability, a risk flag, a named contact. Make the value immediate and tangible and most people will comply.

There's a difference between a PMO that's bypassed because it's perceived as a bureaucratic hurdle and one that's bypassed because it simply doesn't have organisational authority. The solutions are completely different.

The Bellevue University PMC published a good piece on this shift from PMO as governance enforcer to strategic integrator, the argument being that PMOs that try to enforce compliance will always be fought, while those that position as strategic partners earn their seat at the table. Worth a read.

Joost, fair point on shadow budgets. What I'd say is that the mechanism matters less than the principle. Whether it's budget gates, resource allocation, or something else, there has to be some structural reason for stakeholders to engage with the PMO. Pure persuasion has a ceiling.

That said, you're right that a badly designed gate creates worse behaviour than no gate at all. The design has to make engaging with the PMO genuinely easier than going around it.

We tried the "make the PMO useful" approach for two years and it helped at the margin, but some stakeholders just don't care how useful you are. They have their own agenda, their own timelines, and the PMO is a variable in an equation they've already solved without you.

The honest answer for some organisations is that stakeholder bypass is a symptom of a weak PMO mandate, and no amount of helpfulness fixes a mandate problem. That requires a conversation at board level about what the PMO is actually for.

I'd push back slightly on Bas's point about tying PMO sign-off to budget release. I've seen that approach backfire, when people feel controlled, they find creative ways around the process. Shadow budgets, "innovation funds," favours from other departments. You haven't fixed the bypass, you've just made it invisible.

Governance only works when people believe it adds value to them specifically. If it doesn't, you've just built a wall and people will go around it.

One practical thing that helped us: a very simple RACI for project initiation that made the PMO the "consulted" rather than the "approver" on most decisions. Stakeholders stopped feeling like they needed permission and started treating us as a resource. The bypass rate dropped significantly within a quarter.

The psychology of RACI placement is underrated. Nobody likes asking for permission. Everyone's happy to consult an expert.

The language of "approver" vs "consulted" changes the whole dynamic. We use a similar approach now; we call our intake process a "sense check" rather than an approval gate, which sounds trivial but has made a real difference to how people engage with it.

Quick question on the RACI approach: how do you handle situations where the PMO identifies a serious risk during the consultation but the stakeholder proceeds anyway? Do you have any formal escalation path or is it just noted and moved on?

Good question Fleur. We document it, flag it to the sponsor, and note it in the project record. We don't have a veto, and I'm not sure we'd want one; that would bring us back to the approver dynamic. But having it on record means that when things go wrong, the conversation about what to do next is based on facts rather than memory. Over time, stakeholders have learned that ignoring PMO flags tends to create problems they'd rather have avoided.

Joost, this is one of the older threads but the topic never gets old.

Something that helped us in our organisation in Bogota was making the bypass visible without making it punitive. When a stakeholder goes directly to a project team, we started sending a short friendly note acknowledging the request and asking if they wanted us to log it formally. No drama, no reprimand. Just a paper trail.

Most stakeholders are not malicious, they bypass the PMO because its faster or because they have a relationship with someone on the team. When you make the informal route slightly more visible, some of them voluntarily come through the proper channel because they do not want the attention. The ones who keep bypassing are usually the ones with enough seniority that you need to manage the relationship at a different level anyway.

The key is that the note has to be warm and helpful, not passive aggressive. Tone is everything here.

Very interesting discussion here. From my experience in fintech in Kyiv, I see this tension quite often between PMO governance and agile delivery teams. One thing that helped us was separating two different types of governance.

The first type is what I call 'outcome governance': are we building the right things, are we allocating budget to the right priorities, are risks visible at portfolio level? This is genuinely valuable and agile teams do not object to it.

The second type is 'process governance': did you fill in the status report, did you follow the template, did you attend the steering committee? This is the type that agile teams experience as the 'process police' problem. And honestly, when I look at it objectively, they are often right to resist it. When we stopped enforcing process governance rigidly and focused almost entirely on outcome governance, the bypass problem largely disappeared. The teams were not trying to avoid accountability. They were trying to avoid bureaucracy.

I want to add a dimension that is maybe less common in Western PMO discussions: the value of informal influence as a PMO metric.

In the organisations I have worked in, in West Africa and in France, the PMO's most important value is often not visible in any dashboard. It is the conversations you have, the trust you build with senior leaders, the early warning you provide before a problem becomes a crisis.

Of course this is very difficult to measure. But I think it is worth trying. In my current organisation we track something we call 'early escalations': situations where the PMO identified and raised a risk before it became an issue. We count these and estimate the cost that was avoided. It is not perfect but it gives leadership a sense of the value we provide beyond the standard reporting metrics. And it changes the conversation from 'what does the PMO do' to 'what would have happened without it.'