Applying System Dynamics to Business: An Expense Management Example

If you’re new to system dynamics, it may be hard to see a pragmatic application to your business needs. This companion article to “Pipeline Inventory: The Missing Factor in Organizational Expense Management” [1] describes a seemingly simple problem and how a common approach may make things worse, not better. System dynamics offers an effective, easy solution. The associated model can be downloaded for experimentation. Is system dynamics for you? Perhaps you’ve read of system dynamics in books such as The Fifth Discipline [2] and concluded it doesn’t apply to you. You may have seen how system dynamics can solve big problems for companies with deep pockets: forecasting demand for a new car design [3] or estimating the effects of policy decisions on a regional electric power system [4]. Something applied to such big problems probably won’t affect you; you certainly can’t afford it. This short article will demonstrate by example that system dynamics can solve problems of the magnitude seen by many middle managers, and such approaches don’t need to have prohibitive costs. It provides a short introduction to system dynamics, describes a real management problem, presents a model that can be used to solve the problem, and closes with a summary of the lessons that can be drawn. The model [5] is also available to examine in closer detail. What is system dynamics? System dynamics is a method of solving problems by computer simulation. Like many simulation methods, it offers the promise of less expensive learning—it’s cheaper and faster to experiment with the effect of new policies on a computer model than on a real system with real people, equipment, and processes. Bill Harris c ©2000 by Facilitated Systems System dynamics [6] incorporates features that make it especially useful for certain problems: • System dynamics can model the effect of information feedback on the future direction of the model. For example, if the cycle time in a production process becomes excessive, management would typically intervene to make changes, while many process simulations would simply report that cycle time was continuing to grow. System dynamics handles this easily. • To facilitate the use of information feedback, system dynamics carefully separates information accumulated about the state of the system and information fed back from that state to control the evolution of future states. These two key features enable a system dynamics model to address situations that evolve over time. Research indicates that people are often poor at solving such problems mentally [7]. The language of system dynamics is simple (figure 1). Levels (also called