Análise comparativa de metodologias de cálculo para o valor do tempo de vida do cliente : estudo de caso aplicado a uma grande rede de supermercados

Abstract

More and more Brazilian retail supermarket enterprises search for information coming from screened data to increase their marketing strategies. Traditionally, heuristics based on recency, frequency and monetary value variables (RFM) are used to determine which clients should receive relationship priority. This paper addresses the production of forecasts of the customer lifetime value (CLV) through distinctive methodologies, aiming the comparison between them. The data analyzes refer to a group of 5025 clients, belonging to the same entrance harvest (fidelity programming cadastre) from retail supermarket network. By means of the CLV calculus it is possible to identify the best clients according to the individual projection of the contribution margin. Calculus proceedings were developed for application of the simplified approach, suggested by Gupta and Lehmann (2003), in enterprises with noncontractual setting with the client. The results were compared with usage of stochastic models: BG/BB, to estimate the frequency of future buying months, and Gama-Gama, for the prediction of the expected value to be spent per buying month. It was observed that the prediction of values found for a 12 month period was relatively comparable. Regarding the CLV estimates there were significant differences.