How to determine a confidence level in your project schedule – Part One
Recently I was asked to work on a project wherein I had to create a Master Schedule for release of a customer product. Basically, the situation was that there were 5 team leaders, not all of whom fully understood their workflow and not all of whom had ever been in a room together for more than a few hours. (In order to gather input for the schedule, we held a two-day meeting using the Work Breakdown Structure. But more on that in a subsequent post). I was later asked by someone what my confidence level was in the schedule. By that she meant, what percent confidence did I have that the schedule could be accomplished in the allotted four-month timeframe.
So I thought about it for a bit and realized that I could not just pluck a number out of the air. That in fact, my estimate of confidence level had to be based on something specific in the schedule. (Over and above my confidence in the ability of the team to execute).
So what I’m going to talk about here (and in part two) is what I believe to be criteria for determining confidence levels in a schedule. Again, these are just my thoughts. Seems to me that someone more industrious than me could ultimately convert this into metrics. Below are the parameters that I consider when judging schedule confidence, from most important to least:
Estimates. I’m thinking of time estimates here rather than cost. So in Project, we record durations or perhaps effort, sometimes both. I’ve found generally that people aren’t good at estimating work even if they’ve done it before. This is in large part because they don’t keep any historical records. And so they guesstimate based on, perhaps, a vague remembrance of how long it took them to do a similar thing last time. And then they pad it. What raises or lowers confidence here is the ability of the team member to estimate correctly. And then, what happens to this estimate as the schedule iterates, possibly through several levels of management. Is it respected? Is it dismissed and then lowered to meet some arbitrary deadline (trade show, time-to-market, review board, etc.)?
Resources. Every project I’ve either managed or for which I’ve helped create a schedule has a resource crunch. So, there are always too few resources doing too many tasks on too many projects. And so, how is this accounted for? Is it an infinite resource schedule? In other words is the schedule set up as if all resources can work on it all the time? If so, low confidence. Do they use the schedule to accurately account for people’s time on each task? Every Project schedule I’ve ever seen shows at least half the resources in red (over-allocated) on the resource view. This often goes overlooked and the schedule gets baselined anyway. And everybody wonders why the dates were not met. Meanwhile, it turns out that poor Joe was scheduled at 24 hours/day for four weeks. To raise confidence in the schedule, you must level those resources. Resources are not machines. They cannot work at 150% capacity. At least not for the long haul.
(Stay tuned for Part Two later this week)