Authors: Natsuki Sano
Addresses: Department of Industrial Administration, Faculty of Science and Technology, Tokyo University of Science, 2641 Yamazaki, Noda, Chiba, Japan
Abstract: Many studies have been devoted to grasping the customer's purchase decision process by using point-of-sales data. A few recent studies gathered customer shopping path data by using radio frequency identification technology in addition to point-of-sales data in an attempt to grasp customers' in-store behaviour in more detail. However, customer shopping path data only provides coordinate information in a store, whereas an estimation of the state of customer behaviour is needed. This paper proposes a customer behaviour model based on a hidden Markov model, which is used to estimate two states of customer behaviour in a sales area, namely, 'pass by' and 'stop'. In addition, we propose three evaluation measures of sales areas: non-purchase rate after stop, purchase rate after pass by, and stop rate based on the estimated state of customer behaviour. These measures are useful for the assessment of sales areas. In addition, shopping momentum is known as a state of mind that induces a subsequent purchase while shopping. We identify the customers who are experiencing shopping momentum by using point-of-sales and customer shopping path data and compare their purchase results with those of other customers.
Keywords: radio frequency identification; RFID; shopping path data; hidden Markov model; HMM; shopping momentum; marketing; behaviour estimation; customer behaviour; sales areas; supermarkets; purchase behaviour; purchase decisions; decision making; in-store behaviour.
International Journal of Knowledge Engineering and Soft Data Paradigms, 2016 Vol.5 No.2, pp.135 - 145
Received: 21 Jul 2015
Accepted: 29 Sep 2015
Published online: 20 Apr 2016 *