making the right call, Part 2
by Steven Pascale, PMP, Jim Carland, and Carlton Lorenz
SPIRITED PROJECT MANAGERS with outstanding creative dash and never-ending drive to bring their projects in on time and under budget usually find themselves at work during the Saturday morning cartoons, thereby missing a great project management parable: the saga of Wile E. Coyote's attempts to succeed in his plan to catch the elusive Road Runner. The cartoon reminds one of computer new product development, where the chase after improved technology and speed to market is analogous to Wile E. Coyote's never-ending chase to catch the Road Runner. Despite the clever schemes of the “genius extraordinaire” Coyote, and the technical support of the Acme Company, the Road Runner ends each cartoon as risk-free as he began it.
Project managers in new product development can collectively identify and agree on one common purpose, which they share with Mr. Coyote: to try to capture a constantly moving market target, and to do so by delivering a feasible plan that incorporates risk decision and potential outcomes. Corporate strategy should include risk analysis in all of elements of planning, as well as other factors like associated production capability, budgets, and resources. The trick is to mitigate internal and external schedule risk to ensure an accurate prognostication inclusive of desired corporate profits. This article discusses how a planning team takes the data generated by a risk analysis program and uses it to make the best possible decisions for a project plan and schedule.
Risk as a Project Variable. As the technology race for computer hardware development strives to keep pace with unprecedented changing consumer demands, risk becomes a component in a simulation of the variability in activity durations, resources, and cost. In developing an optimized schedule, management must learn to utilize risk in decisions for concept approval and component supplier selection.
As the chase for the Road Runner became a never-ending cycle, viewers began to regard the Coyote as a resource bottleneck, destined to forever pursue the Road Runner without any probability of ever catching him. However, in one episode, Wile E. Coyote somehow becomes productive enough to catch his prey. As he held the Road Runner by the throat, Wile E. turned to the audience with a puzzled look and held up a sign asking: “What do I do now?” He had not planned for this improbable occurrence, and was left with an unclear path to the next logical activity sequence. Project managers need to heed that warning. Although their pursuit of optimization may seem like a chase for the Road Runner, they must plan for the course of action that will follow capturing the prey.
Simulation. Computer simulation is an effective method of analyzing risk in a project schedule. For project managers interested in trying risk analysis for the first time, the first basic steps in using a model include keeping it simple, using good data, and verifying and validating the model.
The process of risk analysis often seems complex and involved. When the modeling is completed, everyone appears to be holding up a sign asking, “What do we do now?” It doesn't have to be that way.
First Iteration of a Model. When using a computer model for the first time a project manager may be disappointed with the output. Models are so powerful that project managers sometimes expect answers complete with trumpets and lyrics: “Your project will be completed on 30 June 1998.” Instead of divine revelation, the output is really more like a riddle: a pile of numbers in the form of a probability distribution.
The truth is that once a simulation is complete the real work of using a model is only beginning. Before the data can be useful, it must be interpreted.
Compare the Probablistic and Deterministic Outputs. Computer models provide by far the best information for project scheduling. That does not mean that the Critical Path Method is a useless tool. Most project managers are familiar with CPM output, but they may be unfamiliar with a plan modeled using PERT estimates as data inputs for duration and Monte Carlo risk simulation results. The CPM schedule can act as a baseline score or a familiar reference point for the novice model-user attempting to understand what the probabilistic model has produced.
The Critical Path Method is a deterministic scheduling technique; it produces a single project completion date. Using single estimates for every task in the project, CPM arrives at a date by essentially adding up the tasks. On the other hand, the PERT planning method combined with Monte Carlo computer simulation is a probabilistic scheduling method. It produces a range of potential completion dates for the project and a probability of each one occurring. By relying on a range for the completion time of each task, the model is able to generate this range of dates. The actual modeling involves calculating the possible combinations of task lengths and summing them to produce completion times. The strength of a model is that it allows the planners to choose an appropriate level of risk and determine a corresponding completion date for that level of risk.
Focus on the Critical Path. The critical path is made up of those tasks, which if delayed, will slow down the progress of an entire project. Studying the data related to critical path tasks may reveal those tasks that pose a high risk of delaying the project. Once identified, these bottlenecks may be subjected to resource leveling, a process of determining how resources and their availability affects the project timing and cost. Leveling provides a solution to resource conflicts by rescheduling activities when resources are available. Tasks representing a high potential to improve the schedule can be addressed, further maximizing project resources. The critical path will likely change as the project progresses. As resources are leveled and areas of work completed, tasks that were previously low-risk may suddenly become much greater risk factors. Careful monitoring of the critical path is crucial throughout the life of the project.
Is There More Than Analysis? Statistical analysis is often mistakenly thought of as the solution to problems, when what it really does is identify potential problems. Think of analysis as a map: it shows where you need to go, but choosing the best route is left to you. Having located problem areas in a project plan, determine the best course of action by considering the options.
Reader Service Number 5095
When problems appear in the critical path analysis, the reflex reaction is to address them with the first available solution. Although a reflex reaction may solve some problems easily enough, for others it will only make things worse. It is better to survey the project landscape at critical points, looking for every potential solution. Since the idea is to come up with as many ways of correcting each problem as possible, a brainstorming session may be an optimal approach.
In the search for options, consider that the approaches to dealing with risk fall into three groups: avoidance, mitigation, retention. Avoidance is the avoiding of risk altogether. Avoidance tactics can range from overloading resources on a task as a means of all but eliminating risk to removing a task from the project entirely. Mitigation involves acting to reduce project risk to acceptable levels. The majority of risk-management strategies are forms of mitigation. Retention is the acceptance of a task risk as a part of the overall project risk. Sometimes the benefits of reducing a risk do not justify applying resources. In those cases, retention is an appropriate strategy.
Common sense requires one to generate options for dealing with risk. A solution that seems ideal on paper may not be a viable possibility at all. For example, during aircraft manufacturing assembly, reducing the time required to install wiring in the cockpit can significantly improve the assembly schedule. Two airframe electricians are usually assigned to the job. The intuitive risk-reduction tactic in this case is to dispatch more electricians to help wire the cockpit. However, as it turns out, only two electricians can fit into the workspace at a time. Any additional resources devoted to this task would be wasted. As in every aspect of project planning, you can improve the quality of the plan by including people who actually perform a job on the planning team.
Consider the Outcomes. Once the team has generated options for managing project risk, take the time to consider the potential outcomes of each. One option may indeed reduce risk as it is intended, but it may create additional risk in the process. The advantage of using simulation instead of using single-point estimates like CPM is apparent in this step of the analysis. It is possible to insert a potential change into the model, and then simulate to determine the effect the change will have on the project's outcome. Multiple solutions to the same problem can be simulated and then compared to find the most suitable level of risk that management wants to assume.
Doing nothing, while perhaps impractical, is an option that should always be considered. An evaluation of the no-action strategy may not produce any workable alternatives, but can at least serve as a baseline or worst-case-scenario outcome.
Make the Change. Books have been written imploring managers and executives to embrace change as a necessary part of risk and doing business. Risk analysis is the process of determining a clear path of decision alternatives. That is to say, a structured methodology for risk analysis is concerned ultimately with improving a project by changing those things that can be made better. Every step of the modeling and analysis is designed to help project managers make positive changes in the project plan. Use the analysis, find the best option, and when the time is right, act. Once you know what the best path is, failing to choose it in the face of change is tantamount to intentionally adding risk to the project.
Model Again. Once the bottlenecks have been worked out and the project is set to begin, it's time to put the model away, right? No! The ability to enter new data and model over and over is what makes simulation so useful. However, correcting one bottleneck may create others. Only by simulating the variability in durations, costs, calendars, and resource usage can desired project outcomes be decided upon and new problems be identified and dealt with. Also, the critical path is likely to change over time. Rerunning the simulation is the best way for the project team to be aware of the current critical path. Once under way, as areas of work are completed, actual task completion times should be recorded into the schedule. Keeping the project schedule updated produces a higher degree of accuracy, especially for those projects that are riddled with complexity.
The Limitations of Models. Simulation, when coupled with project management software tools, is extremely powerful for analyzing levels of project risk and helps upper management in its decisions regarding suppliers, allocation quantities, or speed-to-market aggressiveness. When used appropriately, it can improve a project schedule significantly. However, where risk is prevalent, a model is not the answer to every project planning opportunity. Accordingly, project managers must be aware of the strengths and limitations of using computer project scheduling models. Knowing not only what a model can do but also when they are best used makes the difference between an outstanding project manager with mastery of risk simulation and one who is a slave to managing critical issues on the fly.
Simulation Is Not Psychic. It is easy to understand how one could assume that a computer model has some ability to predict the future. Our generation believes very strongly in the power of numbers. Statistical methods and advanced scientific analysis have taken on a near-mystical quality in our culture. But the simple truth is, simulation cannot predict the future. Although it is based on reliable data, and although it may imply some predictive power, a model's output is essentially an educated guess about what will happen. In other words, no matter how confident one is with the estimated duration/cost ranges or resource distributions, or even how accurate a model seems to be, there will always be a chance that the project outcome will be significantly different than what the model has predicted. This is the nature of chance. The discrepancy between the model and reality means that there will always be error in every model.
USING COMPUTER SIMULATION to manage projects and risk can be a daunting challenge. There are a myriad of details with which to be concerned, and many ways to make small but critical errors. At the same time, simulation provides the project manager with more useful, detailed information about project risk than any other analytical method available. The importance of simulation and managing risk is intended for making the right call, and making sure that the work being performed is in line with the overall goals of the project. When working with a planning model, remember the lesson learned from the Road Runner and Wile E. Coyote. ■
Steve Pascale, PMP, is owner of Pascal Product Advancement in Boston, Mass., which specializes in new product development.
Jim Carland, Ph.D., is a business professor at Western Carolina University. He is also a Certified Management Accountant and a Certified Public Accountant. An author of more than 75 articles and papers, he is a recognized expert in small business.
Carlton Lorenz is a public relations student studying for his master's degree at the University of Georgia.
PM Network • March 1998