Product bundling for ‘Efficient’ vs ‘Non-Efficient’ customers: Market Basket Analysis employing Genetic Algorithm

Abstract

ABSTRACT Achieving profit maximization by retailers is based on a comprehensive understanding of their customers’ brand preferences as well as their purchasing patterns. Segmentation of customers merely leads to grouping the customers into segmented cluster according to their buying behaviour and demographics. However, an in-depth knowledge of purchase pattern over a period of time is necessary to tailor customers’ needs according to their worth to the retailers. It would be more beneficial for the retailers if they are able to map the product choice of more efficient customers (having higher average CLV) vis-a vis those of less-efficient customer (having lower average CLV). This work leverages the concept that customers having higher average CLV are likely to have more meaningful purchase pattern, thus aiding retailers in tailoring their offerings in a seasoned manner. The methodology involving the use of genetic algorithm for market basket analysis results in better mapping of product choices of more efficient compared to less efficient customers, thus aiding the retailers in better merchandizing, planogram management and category management for enhanced profitability.