How to determine a confidence level in your project schedule – Part Two
I had planned to post the last part of my short series on schedule confidence today. And I will do that shortly after this brief digression. But a gentleman named Tony Welsh, president of a company named Barbecana made an insightful comment in regards to my last post that you may not have seen. His basic contention was that you could not measure schedule confidence accurately unless you used Monte Carlo simulation. (For the record, his company makes such a tool).
And having taught risk at universities, I wholeheartedly agree. Monte Carlo simulation, for those who are not familiar with it, is software used to simulate a project over and over again many times. The result is a probability distribution of possible outcomes. So it’s a way of finding the risk in your project that can’t be detected by just looking at your schedule with the naked eye. There’s a good article about it here
The only problem I have with Tony’s assessment is that in 20 years of practicing project management at a variety of companies, I have not run across one yet that uses Monte Carlo. So, curious, I did a mini-survey of some of my colleagues, all of whom have at least 15 years in the business. Most of them have never seen Monte Carlo outside of the classroom. (All these people are also teachers). Interestingly, one colleague – who works at a defense contractor – basically came back and said, Yeah we use it all the time. He says it’s very important in his culture. And so now I’m wondering if Monte Carlo is one of those best practice tools so well adopted by defense and military (like Earned Value) that seem to have almost completely been overlooked in the private sector. (Where I’ve spent most of my PM time).
And speaking of the private sector, I may be just one small player, but sometimes I’m fortunate enough to get into some pretty large companies. So for the past 9 years I’ve been brought in to pharmaceutical companies to help them create schedules and risk registers. And you’d think that, given that their issues are literally life and death, they’d use something like Monte Carlo religiously. But no. Most of them, except for PMP’s, have never heard of it. Hell, I’m lucky to get them to even spend time building a risk register.
So I will definitely concede the point that using Monte Carlo increases your ability to measure confidence. No doubt about it. And further, that it may well have gained a foothold in certain segments where projects are being done. But I also have to live in the commercial world. And in that world, use of Monte Carlo is as rare as a Yeti sighting. And so if a customer tells me they’re flat out not going to invest in this software, then we’re back to measuring confidence by other methods. Flawed though they may be.
(Stay tuned for Part Three)