Not all forecasting methods are born equal. As far as inventory optimization is concerned, classic forecasting methods tend to make no sense whatsoever. The problem does not lie in the models used to produce the forecasts, or the way those models are tuned; it lies instead in the underlying methodology which misfits the actual business needs of commerce.
Quantile forecasts, in contrast, deeply embrace the commerce viewpoint. In this introductive video, we develop two key insights related to lead time and service level. The video illustrates why those two variables need to be at the core of the forecasting methodology in order to properly address the business needs.