|
|
|
|
|
Reducing the Cone of Uncertainty
|
Most leaders in the federal information technology (IT) community agree that IT project performance is not as good as it should be. Overall, 2004 saw an improvement in US IT project success rates, according to the Standish Group CHAOS report. Still, however, according to the same report, 15 percent of IT projects fail outright, and 51 percent are "challenged" ( defined as 20 percent over budget or time and not fully functional). The report found that these failures and challenges represent $55 billion in project waste.
Unfortunately, federal IT projects fare worse. One figure reported in the Journal of Defense Software Engineering, following an analysis of 250 large software projects, put the failure rate at 90 percent with the challenged rate at 20 percent. As many as 70 percent of the 250 projects experienced project overruns in excess of 50 percent of the original estimates.
There are a number of factors that contribute to poor project outcomes, but careful retrospective analysis reveals that the straining points are related and controllable. We will address some of these factors and explore ways to reduce the "cone of uncertainty" in an effort to reduce project risk and increase project performance.
It's All About Planning
Both reports indicate that poor project management contributed to the poor project performance. That seems like an obvious culprit, but further data dissection shows that it is poor project planning not project management that creates an environment of uncertainty and risk. It is the high uncertainty and risk that statically cause project costs to increase resulting in outright cancellations, scope escalation, and budget overruns. After all, the cost of capital is higher for high-risk ventures than for low-risk ventures because the cone of uncertainty is larger, which leads to a broader range of potential undesirable outcomes.
The principles of the cone of uncertainty are used in everything from predicting hurricane destinations to basic everyday outcome analysis. In project speak, the cone of uncertainty presents a way of looking at a project in phases to determine how decisions made in a previous phase affect the outcomes in future phases. The cone is always greater at the beginning of the project because the uncertainty-reducing decisions have not been made yet. As decisions are made, the cone of uncertainty narrows. A project manager wants the cone to narrow faster to reduce uncertainty and risk. However, at the same time, the project manager wants the reduction of uncertainty to be narrowed into the desirable outcome.
The Cone of Uncertainty
From “10 Deadly
Sins of Software Estimation” by Steve McConnell, © 2002-2007,
available from
www.construx.com.
"One cannot calculate the precise future motion of a particle, but only a range of possibilities for the future motion of the particle."
- Heisenberg, in uncertainty principle paper 1927.
Many people would argue that project planning is part of project management. And, in some stable, mature industries that may be true. IT projects in the federal government, however, are almost always envisioned in the "plans and policy branch" and executed by the "software engineering branch." That is, the vision and requirement for a system is mission-driven as it should be, then tossed over the fence to the coders to make happen. This "fence" can be considered the delineation between project planning and project management. Project planning is where the parameters of the cone of uncertainty are formulated. Project management is were the project is executed and where that cone starts to become clear.
It is not that project management is unimportant. Pure project management is an operational science that originated many years before the software industry; a good project manager is always a good project manager, regardless of the project. IT projects are unique, however, in that they often require a more rigorous requirements process, more uncovering of the unknowns, and more "chunking," that is breaking up into smaller, less risky pieces. (This is especially true if the execution phase is contracted out.) That said, the initial planning will still determine the project risk and success probability.
The major activities associated with IT project planning are
project scoping workforce planning project measurement project milestones change management quality control cost estimation
Cost estimation for large, complex projects is difficult in the very early stages. (See the top line in the cone of uncertainty graph.) However, there is one law that applies throughout the planning process: cost, time, and functionality are intrinsically linked in a 'pyramid of estimation'.
Pyramid of Estimation
The idea is that if time and cost are the essence of a project, then functionality becomes defined and is no longer a variable. If functionality is defined, then cost and time are defined. Further, if any one of these points becomes fixed, the other two points become controlled. Think of it as the "conservation of energy" for projects. The point is this: if we manage the success of a project by adherence to planned cost, time, and functionality, the triangle retains its equilateral shape.
It is the need for this kind of balance that makes project planning so important. The process of defining the overall project scope, system requirements, and functionality starts to solidify one point in the pyramid of estimation. Workforce planning helps to understand staff and contractor requirements and overall project slippage. Further, defining the amount of project controls, such as quality assurance, project milestones, change management, and project measurement, helps understand the additional resource requirements. When those points become more understood the cost can be calculated more accurately. The more accurate the cost estimate the less likely a cost overrun will occur.
The overall project planning exercise reduces the cone of uncertainty faster and helps ensure a project's desired outcome. That said, requirements change and sometimes priorities change. Good project planning will ensure processes are in place to manage that change; a good project management team will execute those processes.
For information on how we can help you with your IT strategic management issues, visit us at www.pivotal-insight.com.
|
Free trial
Subscribe Now
|
