2014 Marketing Science Conference, June 2014, Atlanta, GA
Nian Wang. Co-author: Joseph Pancras, Hongju Liu, and Malcolm Houtz
Many consumer goods such as books, DVDs, magazines are sold through catalogs with ‘buy now, pay later’ offers. Firms in such markets face the challenge of predicting what types of customers will not pay for the product, termed ‘bad debt’ customers, since such bad debts pose significant costs for firms. Such firms also need to understand consumer’s propensity to return products, since this is also costly for firms. This paper proposes a new model considering both the bad debt and return simultaneously, and applies the model to the data obtained from a Spanish book company. We find that the extra mailings have a greater influence on bad debt, and that historical bad debt behavior influences the probability of return. These findings help improve the firm’s targeting policy, and also point to the potential for more elaborate modeling of the tradeoffs between the bad debt and return.