Application of Hidden Markov Model to Analyze Enthusiasts' Dynamics of a Lifestyle Brand

摘要

This study investigates the dynamics of customers' buying behavior of a prominent lifestyle brand in Japan with an unique brand community. The analysis examines the lifestyle of its customers, from dimensions of customers' enthusiasm towards the brand to purchase preference factors of where, when, and how purchases. We construct a Hidden Markov Model (HMM) to relate the customers' profile based on the multiple purchase preference factors to observe buying behavior over four different periods. Using this model, we have identified six relationship states and classified customers into corresponding behavioral sets over four periods. Finally, the hidden customer states are compared among male, female, mature, and variety of shop visits behavior for identifying key behavioral differences. The findings on customer behavioral dynamics are useful to retailers for the purpose of enhancing services and customer experiences through more appropriate product recommendations and for new product development.