Increasing uncertainty –
planning parameters undergo change

If we ever have to look back and name a single event that has had a lasting effect on how we conduct planning, then it would surely be COVID-19. However, we wouldn’t be looking back far enough; it is rather the increased number of dependencies in a globally networked, digital world which has made the issue of uncertainty a decisive factor in planning.

Consumer behavior is becoming more difficult to predict, demand is less cyclical, it has also become more volatile in general and geo-political events or even natural catastrophes have an impact on sales markets and the flow of goods. The VUCA world, particularly the issue of uncertainty, has entered the planning process and is here to stay. The question is: how do we deal with it?

As it always was: a number of scenarios against uncertainty

The classic approach towards dealing with uncertainty is to compare a small number of scenarios. Here, each scenario reflects the assumed development of various influential factors. Such factors are, for instance, “customer demands”, “supply chain costs” and “storage costs”. In order to analyze extreme developments, the scenarios are mostly divided into “best case” and “worst case”scenarios as well as into a “normal” scenario. Past values and expert estimates form the basis for creating the scenario.  Frequently, the objective here is to safeguard against the worst case scenario.

This all works well as long as the number of influential factors remains transparent and a sound development of these factors can be assumed for the scenarios. However, this is seldom the case. Unforeseen events are occurring more and more, the quality of the scenarios which have been created is decreasing to the same extent.

A known procedure, upgraded with mathematical optimization.

Mathematical optimization builds upon this tried-and-tested principle in order to safeguard against worst case scenarios. In this case, we do not make any assumptions about the development of influential factors but determine ranges, such as price ranges for “storage costs per square meter”  instead of an assumed specific value. A scenario becomes even more precise when the ranges are provided with the corresponding probabilities of their event. In this way, the sample space becomes very large.

Mathematical optimization can now play out its full potential. Based on the data available, it is able to calculate the optimal decision and make the solution approach more transparent. The optimal solution also takes into account that particularly the most unfavorable combinations can also occur within the ranges. For this reason, all potentially conceivable scenarios as well as the worst cases in particular, are considered automatically. Therefore, if we follow the proposed solution resulting from the mathematical calculation, we can be sure that the company can still work profitably, even in a worst case situation. This is on condition that profitability was the objective of mathematical optimization which allows more than one objective to be optimized.

Are you interested in our factsheet What are the benefits of mathematical optimization?