Sequential market basket analysis
- PDF / 149,553 Bytes
- 12 Pages / 439.37 x 666.142 pts Page_size
- 15 Downloads / 208 Views
Sequential market basket analysis Wagner A. Kamakura
Published online: 22 May 2012 # Springer Science+Business Media, LLC 2012
Abstract Market basket analysis (MBA) is a powerful and common practice in modern retailing that has some limitations stemming from the fact that it infers purchase sequence from joint-purchasing data. However, internet retailers automatically collect purchase-sequence data from their shoppers, and new technology is available for traditional (bricks and mortar) retailers to do the same, making it possible to analyze purchase sequences, rather than inferring them from joint purchases. This study first compares and contrasts traditional market basket analysis with a sequential extension, and then proposes a framework for purchase-sequence analysis, which is illustrated utilizing shopping trip data from one grocery store. Keywords Market basket analysis . Shopping behavior
1 Introduction Market basket analysis (MBA), also known as product affinity analysis, is already widely known and utilized by traditional (bricks and mortar) and Internet retailers. A Google search on the keyword set “market basket analysis” reveals over 40,000 hits, providing evidence of its prevalence in modern retailing. The basic idea behind Market basket analysis is to find pairs or sets of products that are jointly observed in large samples of baskets, based on the assumption that purchase of one or more of the products within a set would lead to purchase of the remaining ones, thereby providing leads for cross-selling, bundling, product positioning, etc. The underlying assumption in market basket analysis is that joint occurrence of two or more products in most baskets imply that these products are complements in purchase (if not in consumption), and therefore, purchase of one will lead to purchase of others. As we will discuss later, market basket analysis makes
W. A. Kamakura (*) Fuqua School of Business, Duke University, One Towerview Road, Durham, NC 27708-0120, USA e-mail: [email protected]
506
Mark Lett (2012) 23:505–516
inferences about purchase sequence from data on joint purchases, which are potentially misleading. In this study, we argue that with the advent of new technology, it is not necessary to make potentially misleading inferences about purchase sequences from joint purchases. Most internet retailers already collect data on the sequence items are added to shopping basket. Traditional retailers can also obtain the same sequence data using RFID technology (Sorensen 2003). Similarly, most services organizations know exactly the order by which their customers adopt new services. Given that data on purchase sequences are becoming more widely available, we propose an extension of market basket analysis into sequential market basket analysis, which takes advantage of these data, and develop a methodology for analyzing purchase sequences. The rest of this study proceeds as follows: first, we describe traditional market basket analysis based on joint purchases. We then present our proposed sequential mar
Data Loading...
