Skip to main content

Faculty Insights

Managing the Project Pipeline

By Wisconsin School of Business

October 31, 2013

A problem I often hear about in our project management courses is the struggle firms have to get their projects done on time. I expect this problem when companies first embark on becoming more project-minded, but it’s surprising to hear this symptom when firms are more mature at managing projects.  After all, the entire field of project management was developed to help reduce time delays and cost overruns while improving upon delivering solutions to stakeholders. If a firm is following sound project management principles, why would their pipeline of all projects get delayed? This seem especially odd if they have already taken steps to develop competent project managers and leaders and have seen success in delivering single projects on time and budget. What’s going on down there?

Anytime you have arrivals, departures, and a pipeline in which service is performed, you can see this time-delay problem occur. We call it a queuing system problem, and research originally performed by John Little in the 1960s helps not only to model the behavior but also to come up with solutions to the problem. In simple terms, Little’s Law states that as you increase arrivals into the pipeline, the average time to complete activities slows down. At first, when the pipeline is uncongested, the slowdown is small, but as the pipeline becomes more clogged, delays increase exponentially. We see this phenomenon occur everywhere there is a pipeline, not just in the field of project management. Let’s use an example I experience almost every day to illustrate Little’s Law: daily commuter traffic.  

I drive to work each day in a little VW Beetle. It’s not a fast car, but it can certainly go at traffic limit speeds. If I drive to work early in the morning, say 5:00 a.m., there’s little traffic on the road and it only takes me a few minutes to travel through the six-lane beltline that takes me into campus. I can drive the speed limit, and I experience no delays. If I take that same road (a pipeline) at 8:00 a.m. or 5:00 p.m., it’s a different story. Even without any sort of road construction or car breakdowns, I’m slowed down.  Sometimes when the road is really clogged, the slowdown is severe; it can take me 10-20 times longer to travel the same stretch of road that I had no problems getting through when there was no traffic. When there’s a little more traffic than my early-morning commute, delays occur, but they are small. When there’s a lot more traffic, delays are enormous.

Now back to the project pipeline discussion. In our case, arrivals are new projects; departures are completed projects; and the project pipeline is made up of the staff, resources, and project management framework used by the organization to complete the project deliverables. Why do project-minded firms see project delays? Often it’s because they have seen success at the individual project level. This success leads to more projects being introduced, which leads to more success, and eventually you’ve got a clogged pipeline, and all of the projects begin to slow down, just like road traffic example earlier. At the macro-level, it’s obvious that the problem is due to pipeline issues that Little so eloquently described in his mathematical formulas and that we experienced in the road traffic example.  However, at the individual project level, the delays aren’t as obvious. Often the delay is misdiagnosed.  Examples include:

  • Individual level: “Why is Mary/John so bad at estimating this project? He/she used to be good.”
  • Work-group level: “What are those people in IT/Marketing/Admininstration doing down there? They always botch our projects.”
  • Organizational level: “We scrapped the project management methods that used to be used, and we’re going to try something different.”

Solving the pipeline problem is simple: understand the variables that affect pipeline and queue performance and build solutions that reduce the effect those variables have.   

One of the most effective ways to solve the project pipeline problem is to closely manage what goes in—and what stays in—the pipeline. Instituting specific hurdles for projects to clear before they are even assigned resources is one way to manage the front end of the pipeline. It’s similar to the rules implemented in road traffic problems (no slow-moving or small vehicles allowed on the pipeline; signal lights that control when vehicles can get in the pipeline during stress periods). Another technique is to prune your project pipeline. Examples include periodically reviewing and removing projects that are in the pipeline but aren’t valuable to the organization or customers, and removing or re-planning projects that have stalled (similar to sending out tow trucks to clear vehicles that have stalled on the road).

There are other variables that affect project pipeline performance; all we have to do is examine more closely the Little’s Law formula. We’ll save that discussion for my next blog entry. Until then, here’s a sneak peak at what’s to come and resources to help you with this common problem.