J. Brian Atwater, CPIM, and Paul Pittman, PhD, CFPIM, CSCP, Jonah | September/October 2012 | 12 | 5
Using visual tools to better understand supply chains
There are many critical factors to consider when making supply chain decisions. Some are not easily quantifiable, such as the risk and potential cost of a disruption, which can be extensive. When you consider the volcanic eruption in Iceland, earthquakes in Japan and New Zealand, pirate attacks off the coast of Somalia, and ongoing economic and political crises around the world, it is easy to see how many supply chain and operations management professionals can feel they are the victims of bad luck.
Defining supply chain complexity
Anticipating all the problems that can occur in a supply chain seems impossible. Too often, strategy consists of identifying a low-cost supplier and simply hoping that disruptions and unanticipated costs don’t offset the savings generated from the new provider. However, instead of stumbling blindly through major decisions, professionals must begin to see the causal links in their environments and become better able to recognize and even anticipate future opportunities and threats. They must be ready to address and meet the challenges of supply chain complexity.
Many of us are familiar with the following scenario: A buyer finds a new source in a distant country for a particular component, and the price is only $5 per unit. The company’s current supplier charges twice that, and the lowest domestic bid is $9.25. Even considering transportation costs and duties, the total price with the new supplier still equals only $8.50 per unit. Coupled with the source’s reputation for reliability, the decision to switch is an apparent no-brainer.
But then, additional expenses materialize out of nowhere. After factoring in unanticipated fees, costs associated with correcting quality problems, and travel for relationship development, the actual landed cost for the item is a whopping $13.75.
What makes supply chains so complex? One aspect is the sheer number of variables. There are many parties involved, including suppliers, logistics providers, and customers—and they often are geographically dispersed. Additionally, there are many intermediate facilities where material is checked in, stored, and rerouted, each with its own personnel. Further, there are different methods of transport to consider, as well as the related costs, benefits, and drawbacks. International and domestic regulatory compliance issues also must be addressed. Even a relatively simple supply chain of 10 suppliers, 10 customers, 3 process points, and 4 transport modes results in a total of 1,200 possible configurations (10 × 10 × 3 × 4).
This phenomenon is known as detail complexity—intricacy that stems from the multiplicative nature of the variables. For example, simply adding one more process point to the scenario results in 1,600 possible combinations (10 × 10 × 4 × 4). Until recently, finding the optimal configuration in this type of system was time consuming and virtually impossible. Computers have made the process faster and easier: Today, professionals can use landed-cost calculators and other software packages to quickly assess their options.
However, this does not mean the complexity problem is solved. After all, the only truly reliable principle is Murphy’s law—the old adage that states whatever can go wrong, will. Put another way, the behavior of the variables changes over time, often in response to actions taken by supply chain stakeholders. Suppliers switch pricing; competitors enter the market; governments alter regulations; the local economy shifts—this is known as dynamic complexity. It manifests when, for example, you offer a large contract to one supplier and another lowers its prices in an effort to entice you back or when disadvantaged people in coastal countries decide to hijack and ransom the shipments they see floating near their shores. Natural disasters also fall into this category and change the ways variables interact.
Supply chain and operations management professionals cannot control all aspects of their environments. However, the key to mastering dynamic complexity is not in controlling the environment; rather, it lies in the ability to understand and anticipate possible events. There is clear need for a method to identify these so-called predictable surprises—a way to see how all the different pieces fit together.
Are you in the loop?
On some level, we all are familiar with thinking in terms of cause and effect. We learn many of life’s early lessons this way. We also make use of formal methods for documenting causal relationships and organizing thoughts, including cause-and-effect illustrations, also known as fishbone or Ishikawa diagrams.
However, while valuable, these systems have a drawback in that they are linear. In this aspect, they imply a singular relationship of cause and effect, and there is a beginning and ending point. One nonlinear system for creating visual representations of supply chain dynamics is the causal loop diagram. This is a logical tool for demonstrating how variables interact and reach outcomes. Often, drawing a causal loop diagram will challenge both the status quo and intuition, yielding significant insights into complexity. As the tools are further refined, supply chain and operations management professionals gain further insights into their systems, identifying high-leverage points that can bring about the most influence. Additionally, as the diagrams are tangible entities, they are easy to share with others and enable the exchange of knowledge.
Consider Little’s law, originally used to model customer behavior in a store. It commonly is expressed as L = λ × W, where L is the average number of customers in the system, λ is the average arrival rate, and W is the average time a customer spends shopping. More recently, the principle has been adopted in lean manufacturing to calculate inventory levels, lead times, demand uncertainty, and customer service response times.
In Figure 1, Little’s law in a manufacturing setting is depicted using a causal loop diagram. The situation can be seen as self-reinforcing either positively or negatively—the phenomena of virtuous or vicious cycles. For example, in order to improve customer service, inventory is increased. This eventually leads to even longer lead times and more uncertainty in demand planning, ultimately having the effect of decreasing customer service. This vicious cycle demonstrates a common idea in lean: that it is never as easy as increasing inventory to protect against disruptions. Another takeaway is that line length, demand fluctuations, and arrival rates often are not within our ability to control; however, customer service time is, making it a critical variable to leverage.
Causal loop diagrams illustrate the reinforcing dynamic of Little’s law more clearly than the original equation. It is a representation of dynamic complexity rather than detail complexity. The next step is to apply the same causal methodology to supply chains. As demand for a company’s products grows, there also is a need for increased supply. As supplies increase, so do inventory and the ability to offer products at a competitive price. Figure 2 captures this basic supply chain dynamic.
Unfortunately, the system does not always behave in such a virtuous manner. In fact, over time it can rapidly spiral negatively when other factors come into play, such as a sudden drop in demand or a disruption to a key supplier. One of the benefits of causal loop diagrams is that they can be expanded to capture these additional causal connections.
Figure 3 shows a causal loop diagram that incorporates additional dynamics found in many supply chains. While it may not capture all aspects of every supply chain, it does illustrate how dynamically complex they can be. For example, it’s clear from the diagram that, when suppliers become spread out geographically, transportation distances increase, thus negatively affecting lead times and vulnerability to supply disruptions. The diagram also shows how transportation distance affects costs.
Answering the tough questions
While some connections are easily intuited, causal loop diagrams expose those that are less obvious. They also help people visualize the relationships of many variables simultaneously. As other factors are identified, a diagram can be expanded to capture their behavior, as well. In fact, the more you understand your supply chain, the more complete the diagram becomes. The value of supply chain causal loop diagrams is realized when they help users understand the less-obvious supply chain questions and consequences.
Some examples of the sort of questions causal loop diagrams can help answer include:
- What will be the result of a major disruption to the supply chain?
- Why have many organizations recently nearshored or localized suppliers?
- Why do new sources of supply sometimes lead to inventory reductions?
- How can new sources of supply result in the need for even more supply?
- If purchasing staff are measured by unit price and the transportation and distribution staff are measured by total logistics cost, what are the impacts on the supply chain?
How can causal loop diagrams go about answering these questions? Let’s examine the last one as a test case. It refers to a potential conflict situation: As purchasers pursue lower unit prices, they seek the lowest-cost suppliers without regard for any other costs. This may increase the geographical spread of suppliers, which increases the total logistics costs, potentially offsetting any of the savings in unit price and creating conflict between the logistics and purchasing groups. Examining the situation using a causal loop diagram can help explain how a $5 part can wind up with a landed cost of $13.75.
Causal diagrams for your supply chain
Causal loop diagrams have been around for many years. They are excellent tools for visualizing, documenting, sharing, and ultimately understanding how various factors affect the supply chain. There is no single diagram that can match all supply chains—the power of the tool lies in its ability to be customized.
J. Brian Atwater, CPIM, is an associate professor of production and operations management in the Jon M. Huntsman School of Business at Utah State University. He has worked as an examiner for the Shingo Prize for Excellence in Manufacturing and the Shingo Prize for Research. Atwater may be contacted at email@example.com.
Paul Pittman, PhD, CFPIM, CSCP, Jonah, is a professor of operations management at Indiana University Southeast. He also is a partner of The LAMP Group. He may be contacted at firstname.lastname@example.org.
To comment on this article, send a message to email@example.com.