An application of DEA to measure branch cross selling efficiency

Abstract

The paper presents a case study in which the DEA methodology was applied to measure bank branches' cross selling effectiveness. The aim of the campaign was for branches to cross sell 4 financial products, and not just to become good at selling 1 product. Therefore, the additive DEA model was better suited to the application than the CCR and BCC models. A straightforward application of the DEA algorithm was not viable due to the presence of zeros in the data set which distorted the business meaning of the analysis. The introduction of an additional output in the form of a count for non zero entries and a subsequent amendment of the DEA additive model reduced the impact of the zeros on the analysis and produced results which reflected the business case.