"X" marks the spot for buried treasure, and I’ve been trying to find "X" since I was a youngster. No, I haven’t been walking around in a pirate hat with a map in one hand and a metal detector in the other. But I have had a piece of paper in one hand and a pen in the other – or since the mid-90s, a mouse and keyboard.
When I say, I’m trying to find "X," I’m really looking for the contingency factors in budgets and schedules that account for the unknown. And, of course, "X’ is always present in the IT world. From concept design thru detailed specifications, it is a way dealing with what we don’t know in our plans.
For example, early in my career many eons ago, my job was to provide initial estimate modification costs to support sales interactions for a CMMS vendor. Generally, the information I received to create the estimate came from a sales guy who provided a high-level description of a potential client’s need, usually based on a 10-minute conversation with the prospect.
At the time, my contingency approach was to estimate the development hours based on the stated need. I then doubled this estimate and applied a 20% bump on top of this total. While I was occasionally on target, some of my estimates eventually turned out too low – understandable since I really didn’t have detailed requirements.
My approach may seem quaint in retrospect, but luckily, I wasn’t responsible for the economic impact of my estimates on making the sale. Someone else would decide on the price quoted to the prospect.
Since those times so long ago, I have had to account for contingency on a wide variety of projects. I’ve also seen how numerous IT departments deal with the issue. Some employ a very structured approach concerning contingency, starting with a 50% plus or minus factor during initial concept design and driving down to 10% plus or minus, with detailed technical specifications. However, many still end up overrunning their budget plus contingency on supply chain projects.
This isn’t too surprising on supply chain execution system projects in which considerable process re-engineering occurs. The complexity inherent in these projects is fertile ground for unknowns. But there are other factors that present challenges when accounting for contingencies in these types of projects.
First, we have a vested interest in the resulting number, as approval of a project or modification is dependent on costs. Most folks believe they approach the process honestly. But the process tends to make optimists out of us.
Next, whenever a contingency number is placed on the table, there is an inevitable drive to reduce it. As we do our necessary homework, it is hard not to say that we have significantly reduced the unknowns. Contingencies can end up in the crosshairs when trying to make a business case work.
Finally, many approach contingency as accounting only for unknown requirements and hidden development complexity. But people aren’t perfect, and they tend to make mistakes. These mistakes can go beyond programming bugs touching all aspects of a project.
So how much contingency is enough? I’m not suggesting a double-the-number-plus-20% approach. But we need to step back and look at how honest we are with our processes. We need to evaluate the contingency factors that we use and resist the pressures to eviscerate them as we finalize our budgets and specifications. We need to learn to appreciate that we don’t know what we don’t know.
Do you have a better approach? How do you find "X"?
-- Tom
Photo credit: ShadBolling
E-mail This | Tweet This | Share on LinkedIn
Tags: budget, costs, time and cost