Um modelo quantitativo para o valor do cliente

• Ano de defesa: 2007
• Autor(a) principal: Ferreira, Eduardo Carlos
• Orientador(a): Bussab, Wilton de Oliveira
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Valor do cliente; Valor vitalício do cliente; Modelos lineares hierárquicos; Modelo de risco proporcional de Cox; Indústria de cartões de crédito
• Link de acesso: https://hdl.handle.net/10438/2595

Application of statistical and financial techniques in decision-making models for marketing investments is an area of study with considerable potential that, as yet, is largely unexplored. The Customer Equity model and Customer Lifetime Value Analysis are gaining recognition in the literature in this area, with discussions among academics and practitioners regarding how to project the Net Present Value of future cash flow generated by a client during the lifetime of his commercial relationship with a firm. Despite the recent appearance of several articles in the academic literature describing customer equity calculations, few suggest statistical techniques or estimators for the parameters that could be used by a firm for its own projections. The model used as a reference in this study proposes an original combination involving survival analysis techniques and hierarchical linear models. In this thesis, we will demonstrate how hierarchical linear models can be used as an important tool to understand, explain and predict the gross margin generated by a client. Use of this technique in business administration is a recent development and it has not yet been applied to the calculation of customer lifetime value. This technique, in addition to other advantages, takes into consideration both the individual characteristics of the client and the trend of the margin he generates over time. Survival analysis has been used to predict the probability that a client will maintain a commercial relationship with the firm. In order to make a further contribution in this area, in this study we develop an application based on Cox’s model (1972), using a methodology that also predicts client behavior on the basis of his individual characteristics. As an illustration, the model will be applied in one of the five largest credit card issuers in Brazil. We note that the same model could be applied in industries with similar characteristics, such as telecommunications, internet service providers, banks and finance companies. Given the exploratory nature of this study, its conclusions are limited to the analysis of this sample, however they provide a starting point for further research.