| Hi-index | 0.00 |
Though the Pareto/NBD (developed by Schmittlein et al. 1987) is a powerful model for customer base analysis, it is difficult to implement especially in terms of parameter estimation. In this paper, the authors propose a MCMC algorithm for model estimation, and a Monte Carlo simulative approach to calculate key results of the model. The outcome of the method is a measure in which value is operationalized as a probability distribution, in contrast to previous studies has actually computed a spot estimation. The algorithm is applied into two direct marketing datasets and gets close parameter estimates with MLE. By implementing MC simulation, the study also shows a good interval predictive performance of the Pareto/NBD. Further more, the authors propose three possible extensions to the Pareto/NBD model and derive a GG/NBD model as a generalization to the Pareto/NBD.