Exploiting sensitivity analysis in Bayesian networks for customer satisfaction study

Abstract

The paper presents an application of Bayesian network technology in a empirical customer satisfaction study. The findings of the study should provide insight as to the importance of product/service dimensions in terms of the strength of their influence on overall satisfaction. To this end we apply a sensitivity analysis of the model's probabilistic parameters, which enables us to classify the dimensions with respect to their (non) linear and synergy effects on low and high overall satisfaction judgments. Selected results from a real-world case study are shown to demonstrate the usefulness of the approach.