Abstract:

This paper begins by estimating a composite measure (index) of attitudinal loyalty for customers of a major telecom supplier's ISP services. The index is constructed in such a way as to maximize differentiation amongst respondents. Next it employs finite mixture methodology for identifying latent customer groups based on the index. It then assigns intuitively meaningful, robust information theoretic measures of importance to both drivers' of within group dependent measures, as well as drivers' of between group membership.