Life sometimes allow your business a global reset. The reset may come as an economic crisis. If that happened to your business, you have to start from scratch, and you need some inputs. You already tortured yourself with questions: ‘What if I’d held on to that contract?’, ‘What if I did that thing differently?’, ‘What if I’d known that in advance?”. Well, you didn’t, and here you are. Now it’s time to start using analytics from the beginning. To test ‘what-if’ scenarios, model outcomes, and simulate situations. It’s time to talk to people to collect the data and create your ‘Management Flight Simulator’. This is the term that experts at the MIT Sloan School of Management use to describe system dynamics modeling. Remember John Sterman’s book, ‘Business Dynamics’? Sterman, a professor at MIT Sloan, is the greatest evangelist of using system dynamics methods for what-if scenario testing in business environments. But before we review the method, I need to explain a bit about the concept of ‘cognitive maps’ and the right way to ‘talk to people’.
A cognitive map, or causal map, is a way to document your understanding of your situation on paper. These maps express your judgement that certain events or actions will lead to practical outcomes. If you do not have any understanding or judgment, then you head out, conduct interviews and draw the map. What is that map? It’s a set of nodes linked together by relationships. At first, you need to collect the information, or conduct ‘data elicitation’. A good example of data elicitation is a TV interview with a financial analyst in which he or she speaks about oil prices and speculates whether prices will go up or down. The TV host asks the analyst series of open-ended questions: ‘What can cause the oil prices to change?’, ‘What factors control oil prices?’ This unstructured data gathering – ‘elicitation’, or simply an interview – yields a richer understanding of a situation and important insights into existing knowledge. These probing open-ended questions can provide the TV host (or you, as the person who works on the what-if scenario testing), with a set of potential variables needed for the management flight simulator, like ‘crude output rate’, ‘transportation cost’, etc. Linkage between the map nodes can be derived from listening for words in the analyst’s speech, such as ‘if-then’, ‘because’, ‘so’, etc. Another way to reveal relationships between the map nodes is to list all the variables and ask an expert to draw connecting links among them.
System dynamics operates using casual loop diagrams. The loops are the links between nodes that can cause the node’s value to grow or to decrease. Reflecting on the oil price example, the nodes ‘crude output rate’ and ‘transportation cost’ are linked to the ‘oil price’ node. In general, if ‘crude output rate’ goes up, then the ‘oil price’ goes down. If an oil company manager without strategic vision wants to increase revenue during a time of decreasing prices, then he will order an increase in output in attempt to produce and sell even more oil – a decision that will cause prices to drop even lower. In system dynamics terms, this map has a reinforcing loop between the nodes. In this situation, the transportation cost can balance the oil price, since when transportation is costly then the end price is higher. The connection between these nodes on the map is called the balancing loop. Obviously, nothing can stop the TV host from asking more open-ended questions and forcing the analyst to continue speculating and adding more and more nodes and loops to the casual loop diagram. The final diagram – here, ‘final’ means the point when the analyst cannot add any more factors – can work as a model, or simulator, of the oil market – precisely the simulation the company managers need to test their decisions.
When programmed using specialized computer software, or even simply drawn on a whiteboard, the causal loop diagram will help you understand dynamics of your system and your business. When all the parameters are in place, you can study the stability of your decisions for vulnerability to external changes. You can identify scenarios where your business will operate in a self-sustaining mode, and ones where it will be highly unstable, when even a small breakage can cause a disaster. I’m not talking about business process modeling here – although you can surely use system dynamics for that. I’m talking about interactions between different units and even organizations inside your business.
An illustration is an Ericsson case. In the middle of the first decade of the 21st century, a lightning strike shut down a Phillips-owned semiconductor processing that was a key Ericsson supplier of mobile phone chips. Coincidently, Nokia also bought the same chips from Phillips, and the plant was the only plant Phillips had delegated for these particular semiconductor devices. In two days, news of the supply-chain problem reached Nokia’s top management. Almost instantly, Nokia contacted all the relevant semiconductor device vendors to ask for help. Some agreed to adjust their processes and produce the required chips. At Ericsson, the message to the top managers traveled for three weeks from department to department, from office to office. Finally, when the Ericsson leadership realized they had a problem, all the available manufacturers had been contracted by Nokia. Guess what? Ericsson does not produce cell phones anymore. There might be other reasons, but soon after the strike, Sony snapped up the broken Ericsson mobile division and opened the door to the mobile device market.
Managers in the Ericsson mobile division, although smart, ignored the rules of system dynamics and did not sufficiently test their business for stability. You are smarter. You know that testing ‘what-if’ testing scenarios is essential from the very beginning. You will not wait for the new crisis to come. You are going to build a fortress.