I recently heard through the grapevine that a large bank (I won’t tell you which one) is cancelling 25% of its IT projects.
These projects all went through some sort of cost/benefit analysis. Business cases were written, and approved. The funding was granted.
They’re being cancelled, often before they even start, because the IT department can’t manage to spend all the money allocated to projects for the year. Apparently they routinely only spend 75% of the money they get allocated for projects, so they’re cutting 25% of their project budget by simply killing projects.
A bold choice.
Right, or Wrong?
Is this the right, or the wrong choice?
Well, it depends on your point of view.
If you’re in finance, the value and timing of cash flows are important, and right now they’re all wrong by 25%. That’s huge, particularly with a bank’s IT budget.
If you’re a manager, and you know that you can only achieve 75% of the current workload, you may as well stop trying to do the other 25%. It’s a needless distraction.
But simply culling projects isn’t really the answer, either. Someone obviously thought the project was beneficial, and would achieve a positive return. That’s what a business case is for: figuring out if the project will make (or save) more money than it costs.
So by culling projects, you’re deliberately making (or saving) less money.
Measuring PMO Effectiveness
One positive observation I have is that someone was looking at the effectiveness of the Project Management Office, or PMO. The PMO is a centralised function found in most large organisations that handles all the internal project management.
The PMO is responsible for coordinating all the project managers, and managing the project budgets, and dealing with status reporting and so on.
So in this case, they’re not effective to the tune of 25% of budget.
Which is pretty poor.
But at least it’s being measured. Imagine if you were only 75% effective and you didn’t even know. It’s not hard to imagine, because many companies don’t even look at this.
Precious few measure much beyond budget, and “is it late?”
I was doing some statistics homework last night, and @simalam reminded me of operations theory. You’ve probably heard of six-sigma, or lean manufacturing, or Scientific Management.
They’re old concepts, and are part of the reason why factories can churn out millions, billions!, of products that are all the same.
It still astounds me how few of these ideas, well proven in real-life situations, have made it into the management of PMOs.
At least the bank here was looking at what the expected spend rate was, and noticed that it was too low, and then did something about it.
But a better situation might be to gather more statistics on projects, and then see how things can be improved.
After all, the alternative isn’t really a great plan.
Definition of Insanity
Attributed to Albert Einstein is the saying that insanity is “doing the same thing over and over again and expecting different results.”
Do you know how late your projects are on average? How much over, or under budget they are?
Do you know if a particular project manager tends to bring projects in on time, or over time?
Do you check to see how successful projects are six months after go-live? Does a particular PM rush everything through at the last second and then throw a dead-cat over the fence to operations?
Imagine you know that, on average, projects run for 4.5 weeks longer than originally estimated, plus/minus a week. When a new project comes to you, and their business plan says they’ll take 18 weeks to finish, do you ask them what they’re going to do differently than all the other projects so that they actually run to time?
Couldn’t our mysterious bank look at all the projects to see if there’s a particular type of project that is dropping the average? Maybe bigger ones don’t spend their money on time. Maybe all the small ones get delayed by months because they’re not big and important enough to get executive signoff on time.
Ask Better Questions
Having more, and better, information doesn’t mean you suddenly get magical answers to things.
But it does mean you can start asking better questions.
Do you have experience with a well run PMO? Or maybe you have some interesting war stories of PMO #fail. Share them in the comments!
Very interesting article Justin.
I think the main issue with metric based decision making, especially statistically based, is that most peoples level of numeracy let alone statistical knowledge is surprisingly low.
So the introduction of the “black art” can face lots of resistance primarily because people fundamentally don’t understand the base concepts. In my opinion for it to be successful there has to be a very good communicator or communication tool to deliver the information in order to “start asking better questions”.
You don’t necessarily need to go the full multiple-regression analysis route. Just a little more empiricism is all I’m asking for.
Because if you’re not making decisions based on evidence, what are you basing them on?
And how can you tell if they’re any good?
I fully agree that some basic empiricism is required.
My concern is that I’m not convinced that many decisions are actually based on any sort of evidence. :)
Pingback: BlogFest2: Dimension Data | eigenmagic