There’s a reason IT PMOs fail

Only a third of IT PMOs will ever work, and the rest of ’em won’t. For the mathematically challenged, such as myself, that means that two-thirds of all the IT project management offices (or program management offices, or whatever you want to call them) will fail. Nothing scientific in these numbers, mind you, and I didn’t do an extensive survey, this is just what I’ve seen myself, and you’re going to have to take it (or not) on faith.

In my line of business, I see lots and lots of PMOs – in fact, I’d say that they’ve become almost fashionable in the last couple of years; you can’t seem to turn around without someone setting up, re-energizing or re-engineering an IT PMO. There’s even a conference or two dedicated exclusively to IT PMOs; one recently wrapped up in Savannah, Ga. And I’ll bet some expert in Savannah told the attendees: “Two-thirds of PMOs will fail, present company excepted, of course.”

Think about the implications: that’s a significant amount of time and money that major organizations are putting into the idea of improving their project management practices (that’s good) and two-thirds of it is going down the drain (that’s bad). That’s two-thirds of the IS organizations who are ticked off with their PMOs and the people in it. That’s two-thirds of a whole bunch of good opportunities being wasted.

Why the dismal number? That’s easy: ’cause most PMOs are set up for the wrong reasons. Period. End of column. That’s it.

Most of the PMOs that I see are set up as a response to a recurring set of problems in their host organizations: “We keep screwing up on our projects (blowing budgets, blowing schedules, failing to deliver what we said we would) – we’ve got to get some better controls in place” or “We can’t seem to get the information that managers wants into their hands” or “We’re trying to do way too much with too few people – how do we choose the projects we want to stick with and which ones we should drop?”

Legitimate problems all, and all noble reasons to pursue improved project and program management practices, but ultimately wrongheaded if they’re what your PMO is all about.

And what makes it even worse, is that just by the nature of the thing, PMOs, and the people in them, aren’t going to be popular with their IT colleagues. You think IT in general has a popularity problem with the businesses it supports? PMOs have the same problem inside IT departments. Think about those episodes of NYPD Blue where the IAD guys (I can’t remember what IAD it stands for, but they’re the internal police guys who investigate transgressions by other police guys) turn up – no one likes ’em, and no one wants to work with them. Why? Because they’re not seen as being there to help – they’re seen as being there to uncover mistakes and squeal back to management.

From a bad start, many PMOs take the next step in being annoying: they try to tell project teams “how they must report to meet management’s expectations,” they harangue the teams to use “standards,” and they ask for additional reports the teams hadn’t planned to provide. In short order, they become the “Project Nazis.”

Now to the minority: the one-third of IT PMOs that actually work – how do they do it? As the weathered old cowboy Curly said to Billy Crystal in City Slickers: “It’s just one thing.”

And that one thing is looking at the PMO from the perspective of the project teams first, and management second.

Successful PMOs don’t start by dictating and/or reporting and/or controling and/or tracking standards. They don’t issue orders, and they never say: “you need to do this ’cause management wants it.” If you’re in the PMO, and you force/cajole/bully the teams into filling out paperwork, your IT colleagues will avoid you in the halls, they’ll disparage you behind your back and they won’t like you much either.

As John F. Kennedy said, ask not what the project teams can do for the PMO, but what the PMO can do for the project teams (I’m sure he said something like that), and if the PMO doesn’t do anything else in its first six months of existence but help the teams execute better, it’ll be wildly successful.

PMO guys – repeat after me: “Hey! I hear you’ve stakeholders up the whazoo on your project. Bet it’s tough to get them all aligned. We’ve got this great tool that attacks just that problem…can we help?”

“Hey! I hear you’ve got your first project steering committee meeting tomorrow. Those guys drive me nuts when they can’t agree on priorities…we’ve got this project triangle thing that we think might help…”

Dig in, do the grunt work and don’t expect the teams to produce anything for the PMO that you wouldn’t want to produce yourself if you were in their position, and don’t ask for anything that doesn’t have a direct benefit to the project itself. In fact, for the first few times the project teams use a tool or technique the PMO suggests, the PMO staff had better be prepared to do the work for teams. If you force them to fill out paperwork, they’ll avoid you in the halls and disparage you behind your back.

The trick is giving the people who are actually doing the project tools that works on the ground and in the trenches right from day one. Do that, and the project teams will beat a path to the PMO door.

You can worry about putting in the control and reporting and tracking mechanisms later – in fact, those mechanisms should be a natural by-product of a well managed project, well managed partly because the (hugely helpful and supportive) PMO has provided an excellent set of tools, practices and competencies that make the project teams better. More on this subject later.

Hanley is an IS professional in Calgary. He can be reached at