When a GPS calculates your destination arrival time, it’s taking into account the remaining route, the speed limits and even the current traffic when adding up the individual routes and providing you a single answer. That would be great for projects too, right?
Projects, however, are more difficult to forecast as tasks contains greater uncertainly than the segments of your trip. Most project software reinforces this predicament by IGNORING the uncertainty, requiring you to enter a single precise number into the task duration field. (And if you’re NOT estimating task durations, you’re missing out on something important that’s discussed in this blog.)
Consider this statement: A duration “range” will always be more accurate than a discrete single number. Can we agree on that? The time it takes to do anything “ranges” from x to y. Therefore, to be most accurate in estimating tasks, we should use a range of time. But (most) software can’t deal with ranges. So a project planner (and team) feels compelled to make a choice.
Simplified, you have three choices. High-end of the range, low-end of the range, or somewhere in the middle. The most sophisticated model I’ve seen suggests the “middle” should be a calculation to dertermine a “safe” estimate: (low-end + 4 x mean + high-end) / 6. Given those three choices of high, middle and low, which do you think is used most of the time?
High seems most prudent. A high-end estimate will help account for uncertainty in completing a task. It prepares the project for “bad surprises.” This seems to be the safest way to go. However, there are two significant downsides to this approach.
First, using high-end estimates pushes the project completion deep into the future. This drives executives/managers nuts. They marvel “how did a six-month project grow to twelve months with detailed planning?!” Project timelines expanding by 2-3 times is common with this planning method. Unwittingly, this practice creates a toxic environment where managers suspect project teams of sandbagging and teams think managers are incompetently setting unrealistic expectations.
Second, when performing the task, teams almost always use up the entirety of the planned duration. The adage “work expands to fill the time allotted” is largely true. Most work is amazingly completed ON the planned target date. In practice, the “safety” embedded in these task estimates is used whether it’s needed or not. Delivering early, when it happens, fails to benefit projects because people working downstream tasks are typically not prepared to start the next step early — they’re working according to the project schedule.
These two downsides are sinister. One diminishes the trust between managers and project teams. The second consumes the safety allotted for project uncertainty whether it’s needed or not, thereby making projects longer then they need to be. Using high-end estimates is NOT the way to go.
Well, this leaves two alternatives — the low-end or the middle. You’re going to be shocked by the answer. With that “tease,” check out Part Two of this blog as to why low-end is the best alternative… with a caveat.
