Regularities in aggregated consumer behaviour and prevention of stock-outs in retailing

In retailing supply-side stock outs (empty inventories) and demand-side stock outs (empty shelves) have to be discerned, because both can occur independently from each other. They entail different kinds of damages and call for different kinds of countermeasures. The aim of this paper ist to present one such countermeasure, a method that, based on a permanent analysis of incoming scanner data (e.g. market basket data), generates hints at shelves that might be empty. For this purpose selling probabilities for specific time periods are uses. The method is discussed theoretically and tried empirically with real market basket data provided by a German retailer....