Big Data Market Segmentation

Market segmentation is a multicriterion problem. This dissertation addresses the multicriterion nature of market segmentation with a new unified segmentation model that is derived from a multiobjective conceptual framework. The unified model elegantly solves the intrinsic antagonistic problem of market segmentation by generating a set of Pareto optimal solutions that represent different tradeoffs among multiple conflicting objectives. This dissertation develops an innovative implementation named Multicriterion Market Segmentation using Evolutionary Algorithm (MMSEA). Based on multiobjective evolutionary algorithms, MMSEA overcomes many limitations and disadvantages of existing methods by optimizing multiple objectives simultaneously, searching for globally optimal solutions and generating a set of Pareto optimal solutions. It also suggests the interesting solutions based on the geometric characteristics of Pareto front. The method was applied to customer value and benefit segmentation for the cell phone service market (a descriptive segmentation model) and customer response segmentation for a national retailer (two predictive segmentation models). The empirical evaluation shows that the proposed unified market segmentation model and solution techniques provide the decision makers with many insights and enhanced flexibility that are missing in existing market segmentation methods.