Executive Summary
The recent proliferation of auction sites on the Internet and the growing importance of online auctions as exchange mechanisms have attracted the attention of academic researchers who have studied issues such as the effect of auction formats, the extent of the winner’s curse, the last-minute bidding phenomenon, and the value of seller reputation. However, understanding is limited such that it is difficult to explain bidding behavior over the entire sequence of bids as opposed to simply summary outcomes in an auction.
For example, while an auction is in progress, participants in the auction will be influenced by various types of value signals (e.g., minimum bid, seller reputation, other participants’ bids, number of bids submitted up to that point), which in turn can affect their decision dynamics for the auctioned item. Recently, the standard assumption of bidder rationality in online bidding behavior has been questioned in a variety of empirical settings. In particular, researchers report evidence of herd behavior bias and find an effect of minimum bid on the final auction price.
In this research, the authors develop a dynamic parametric stochastic model of bidding behavior in Internet auctions. They incorporate and model simultaneously four key components of the bidding process under an integrated framework: whether people will bid on an auction, (if so) who will bid, when they will bid, and how much they will bid over the entire sequence of bids in an auction. This integrated framework is based on a single latent time-varying construct of consumer willingness to bid (WTB), which bidders have and update for a particular auction item over the course of the auction duration. The modeling approach is also based on a simple yet general bidding premise: The observed bidder’s latent WTB at a specific bid is greater than the outstanding bid, yet WTB is unconstrained for all other potential bidders. Thus, the authors impose no structural assumption on bidder rationality or equilibrium behavior; instead, they derive their model using a probabilistic modeling paradigm. This may be the first attempt to model behavioral aspects of bidding behavior formally for the entire sequence of bids in Internet auctions.
Using a database of notebook auctions from one of the largest Internet auction sites in Korea, the authors demonstrate that this general (yet parsimonious) model captures the key behavioral aspects of bidding behavior. They find that most product characteristics matter in the expected ways. Their other primary findings are as follows: First, the estimates of most auction design variables are significantly associated with WTB. Second, negative reputation has a significant negative effect on WTB, whereas positive reputation is not statistically significant. Third, the estimates of most time-varying variables are significantly associated with WTB, which indicates that the information in the other participants’ bids has an impact on the bidding process. Finally, although an individual bidder’s prior winning experience is not statistically significant, an his or her losses, both in number and amount, are significantly and negatively associated with WTB.
Through a data-windowing procedure that assesses the set of potential bidders for a given auctioned item, the authors provide a valuable tool for managers at auction sites to conduct their customer relationship management efforts, which require them to evaluate the “goodness” of the listed auction items (i.e., whether people bid) and the goodness of the potential bidders in their Internet auctions (i.e., who bids, when they bid, and how much they bid). Because the model can infer whether people bid, who bids, when they bid, and how much they bid at each time over the course of the auction duration among the potential bidders, this approach can be considered for the development of contact (communication) strategies at auction sites.
Biography
Young-Hoon Park is Assistant Professor of Marketing in the Johnson Graduate School of Management at Cornell University. He earned a doctoral degree in Marketing and a master’s degree in Operations and Information Management from the Wharton School at the University of Pennsylvania. Previously, he received a bachelor’s degree from Sogang University and a master’s degree from Korea Advanced Institute of Science and Technology. His research interests include customer relationship management, database marketing, and the marketing–operations interface. His research has appeared in Marketing Science.
Eric T. Bradlow is currently Professor of Marketing and Statistics and Academic Director of the Wharton Small Business Development Center, the Wharton School, at the University of Pennsylvania. He earned a Bachelor of Science in Economics from the Wharton School in 1988, a master’s degree in Mathematical Statistics in 1990, and a doctoral degree in Mathematical Statistics in 1994 from Harvard University. He has recently published articles in Journal of the American Statistical Association, Psychometrika, Statistica Sinica, Chance, Marketing Science, and Journal of Marketing Research. He also serves as Associate Editor for Journal of Computational and Graphical Statistics, Journal of Educational and Behavioral Statistics, and Psychometrika. He is on the editorial board of Marketing Letters, Marketing Science, Journal of Marketing Research, Quantitative Marketing and Economics, and Quarterly Journal of Electronic Commerce. His research interests include Bayesian modeling, statistical computing, and developing new methodology for unique data structures with application to business problems.
J Marketing Research, Volume 42, Number 4, November 2005
View Table of Contents.