Customer profiling for Malaysia online retail industry using K-Means clustering and RM model

Abstract

Malaysia's online retail industry is growing sophisticated for the past years and is not expected to stop growing in the following years. Meanwhile, customers are becoming smarter about buying. Online Retailers have to identify and understand their customer needs to provide appropriate services/products to the demanding customer and attracting new customers. Customer profiling is a method that helps retailers to understand their customers. This study examines the usefulness of the LRFMP model (Length, Recency, Frequency, Monetary, and Periodicity), the models that comprised part of its variables, and its predecessor RFM model using the Silhouette Index test. Furthermore, an automated Elbow Method was employed and its usefulness was compared against the conventional visual analytics. As result, the RM model was selected as the finest model in performing K-Means Clustering in the given context. Despite the unusefulness of the LRFMP model in K-Means Clustering, some of its variables remained useful in the customer profiling process by providing extra information on cluster characteristics. Moreover, the effect of sample size on cluster validity was investigated. Lastly, the limitations and future research recommendations are discussed alongside the discussion to bridge for future works.