Project Monitoring, Analytics, and Control

This chapter provides a detailed overview of the processes involved in monitoring and reporting project performance.

The Illusion of Linearity

The best monitoring data in the world is useless if you lack the ability to interpret it correctly. One of the most common interpretation errors is assuming the relationship between two things is linear when it is in fact nonlinear. Numerous studies in cognitive psychology have shown that humans have a hard time grasping nonlinear systems, where the relationship between cause and effect is uncertain. A cognitive bias in favor of linearity makes us naturally predisposed to perceive simple, direct relationships between things, when in reality more complex forces are at play.

For example, marketing forecasts often assume a linear relationship between consumer attitudes and behavior, when in fact things are much more complicated. One study focused on the relationship between consumers' stated preference for organic products and the same consumers' actual behavior. You might think that someone with a strong preference for organic products would buy more organic vegetables than someone with a less strong preference for organic products. You might be surprised to learn that this is not the case, because the relationship between consumer attitudes and behavior is nonlinear.

Project managers fall prey to the linearity bias frequently, especially when it comes to the relationship between time and the many elements of a project. Because time is shown on the x-axis in Microsoft Project, we make the mistake of thinking that individual tasks will be completed in one linear stream of accomplishment. In reality, however, the relationship graph may take the form of a curve or a step function. Failure to grasp this means that any attempts to monitor and control a project are founded on incorrect assumptions, and therefore doomed to failure.

In addition to muddying your understanding of cause and effect, the linearity bias can cause you to confuse activity with accomplishment. But just because people are bustling around the office does not mean they are actually getting anything done. Think of the kind of unfocused activity that often occurs as you're getting ready to move from one home to the next. You might spend some time sorting kitchen utensils until you get distracted by alphabetizing your CD collection before you pack it away in boxes. Then, suddenly, the movers show up, and you kick into gear. In one hour, you might accomplish more than in the previous three days. A graph of your accomplishments during the move might look like the step function shown in Figure 11-2, with very little of importance actually being accomplished, followed by a great deal being accomplished.

Figure 11-2: Productivity often takes the form of a step function; here, the process of packing up to move begins with very little being accomplished, followed by the movers showing up, at which point a great deal is accomplished.

As a project manager, you need to make your monitoring measures factor in the nonlinearity of resource use. Resource expenditures are often low at first. As a result, an inexperienced project manager might be lulled into thinking she is working with a linear system, in which resource expenditures will continue at the same rate throughout the project. In most projects, however, most of the resources are used up near the end of the project. Suddenly, the slope of the graph illustrating resource use over time takes a vertical turn, as in Figure 11-3.

Figure 11-3: Resource consumption can seem linear and then change dramatically; here, resources are consumed at a linear rate of 1% per week through week 10, followed by a sudden uptick in weeks 10-13

Note that the step function model of productivity applies to most Agile projects. Productivity is zero until the team can demonstrate that they have created a working feature, at which point the productivity graph takes a step up. Ideally, each sprint causes another step up, but if the client is not satisfied with the outcome of a particular sprint, productivity stays flat until the end of the next sprint.