Cluster analysis as a rational method for pharmaceutical market segmentation on price basis

Abstract

Three possible methods for post-hoc market segmentation on price basis are compared in the paper. They are k-means clustering, hierarchical clustering and exhaustive search. The results have showed that exhaustive search is preferred for price-based segmentation in inventory management where fast and reliable method is needed and a comprehensive picture of prices distribution is not required. However, hierarchical clustering is more valuable for product positioning due to its inner capability to represent prices distribution via the specific plot – dendrogram. Both strong and weak sides of the methods are discussed and practical recommendations are given.