I am preparing to speak to a Project Management group in Austin this afternoon. My preparation included reading a few PMI articles, one of which was quite interesting to the Lean Six Sigma side of me. What Causes Bad Estimates… and What You Can Do About It by James T. Brown. It was found on the Project Management Institute website.
The article spoke of research that showed that 90% of all projects (out of 258) were found to have cost overruns. The reasons for the overruns were documented as estimation issues, since the project did not estimate the proper time or effort to get it completed as planned.
The two primary reasons discussed are: Not including sufficient time to deal with uncertainties and creating an artificially low estimate in order to get the project approved. I propose that these two reasons are be driven by a similar motive, to get approval to execute the project.
Acknowledging the uncertainties in estimating time and cost are easy for Lean Six Sigma practitioners, because we deal with means and variability every day. In the project management world, most work is built on using averages. It is assumed that if you average enough steps or activities together, that the final project will be good. If this article is correct that 90% are late and over cost, buiding a plan using averages is not adequate.
Using a Lean Six Sigma view, the project plan should plan a mean and a confidence interval for true completion date. I have been put in a position to provide project estimates at my last employer and we used nominal estimates at first, but we were wrong a bit over 50% of the time. As you would expect, after getting metaphorically beat up enough times, my Lean Six Sigma group began providing the 95% upper bound for time and the 95% lower bound for the projected improvement and cost savings to the company executives. Using this model we were always early and we always exceeded our savings goals. During each project, we still worked with milestones based on the average expectations to stay on track, we just avoided sharing that plan with the leadership.
The second issue, under-scoping the effort to get it approved, is very risky. At one level, you risk your credibility every time you take this path. If it is an internal project, your career could be limited because of this type of plan. If you are under-scoping bids on external work, your company may risk fines or other problems if you complete the project late and over-budget. I would rather not commit to the work without a realistic time and cost plan. Now if the management accepts my plan but sets an artificial deadline that is different from my estimate, I will still do the work, but I will make my plan public and let them understand it is a risk to meet their artificial deadline.
Consider reading this article, it may enlighten you to the project management views that are not always included into Lean Six Sigma training.