Protecting service levels with forecasts

Forecasts provide essential data for operating in today’s volatile environments by offering valuable insights into future events. However, in a VUCA world, contingencies will inevitably alter our predictions. The question is not if disruptions will occur, but when they will happen. Companies must be prepared for these challenges and use their tools accordingly.

Many companies base their supply and operations planning on creating connections and dependencies. Holding large amounts of stock is detrimental due to the financial costs and the negative impact on overall organizational performance. Therefore, their goal is to order only what is needed, when it is needed.

However, the effectiveness of this planning approach is threatened by one factor: variability. This method intrinsically hinders the ability to adapt to demand and supply fluctuations that can, and eventually will, occur.

The implementation of forecasts cannot be approached in the same manner, as they also face variability and its accuracy can be impacted by unforeseen events. Therefore, companies must be prepared for these scenarios. Instead of generating replenishment orders exactly to cover for the most probable forecasted demand number, it is a better practice to consider the range of possibilities forecasts provide - what is called "confidence interval" in statistics.

With that information, a stock buffer can be created. The inventory should contain enough units to avoid a stock breakage if the best-case scenario finally occurs. On the other extreme, if demand follows the expected minimum, the only remaining inventory will be the difference of the amounts between the best and worst-case scenarios.