Market Segment Derivation and Profiling Via a Finite Mixture Model Framework
Marketing Letters, Volume 13, Issue 1, pp 17-25, 2002
Posted: 13 Jun 2016
Date Written: February 2002
The Marketing literature has shown how difficult it is to profile market segments derived with finite mixture models, especially using traditional descriptor variables (e.g., demographics). Such profiling is critical for the proper implementation of segmentation strategy. We propose a new finite mixture modelling approach that provides a variety of model specifications to address this segmentation dilemma. Our proposed approach allows for a large number of nested models (special cases) and associated tests of (local) independence to distinguish amongst them. A commercial application to customer satisfaction is provided where a variety of different model specifications are tested and compared.
Keywords: finite mixture models, market segmentation, concomitant variables, customer satisfaction
Suggested Citation: Suggested Citation