PLS Path Modeling in Marketing and Genetic Algorithm Segmentation

This paper presents the PLS genetic algorithm segmentation methodology which uses directed random searches to detect an optimum solution in the complex search space that underlies data partitioning tasks in PLS path modeling. The results of a simulation study allow a primary assessment of this novel approach and reveal its capabilities and effectiveness. Furthermore, applying the approach to the American Customer Satisfaction Index model allows unobserved heterogeneity and different consumer segments to be uncovered.