eBay Action Project Description
I have been interested in auctions ever since graduate school and
took some classes with Nobel Laureate Vernon Smith and his colleague,
Jim Cox. Specifically, I was intrigued by the decision errors made
by some bidders in the auction markets, and how a bidding "team"
or group might fare in different auction market mechanism.
This interest lead to a National Science Foundation Grant with
Jim Cox ($278,000). We examined the effect of the winner's curse
on bidding teams and that project is described here.
Clearly this lead to my interest in eBay -
online auctions! Using money from the NSF Research Experiences
for Undergraduates, we built some software to extract a few auctions
for study. To date we have over 100 million auctions in
the database... :) !!!
The papers that are coming out of this project
Smith, C.A.P., and Vijaysarathy, L. (2003),
“Who wins on eBay: An Analysis of Bidders and Their Bid Behaviours,”
Journal of Electronic Markets, 13(4):459-470.
Online auctioning is one of the more successful business innovations on the Web. The
auction format at eBay, the leading online auctioneer, has some unique
characteristics including a fixed closing time for the bids and the use of a
proxy bidding system that that is capable of acting as a bidder’s agent. These
features, coupled with third party products such as sniping software, have
introduced novel bidder behaviours that may not occur in more traditional
auction formats. In an attempt to study these behaviours, we collected and analyzed
data from over 11,000 eBay auctions. This paper presents the results of the
analyses including descriptive information about the auctions and the
classification of bids and bidders based on bid timing, frequency, and strategy
employed. The different types of bidders and their success rates offer insights
into the nature of bidder participation in eBay auctions. In addition,
chi-square analyses reveal significant differences among the bidding strategies
with respect to auction outcome. Implications of the findings and a framework
to guide future research on online auctions are presented.
Smith, C.A.P. and Vijayasarathy, L. (2002), “Predicting Sniping in
eBay Auctions,” INFORMS Conference, San Jose, CA, November.
From data gathered from thousands of completed
auctions on eBay’s auction website, we show that sniping occurs quite
frequently (we define sniping as any bidding that occurs in the last minute of
an auction). Sniping occurs on average in 25% of our sample of 16000
auctions. We hypothesize that sniping
behavior is explained by a number of factors including the number of bids, the
number of bidders, the seller and buyer ratings, the relative bid price, and
the product type. We test our hypotheses using logistic regression and offer
a predictive model of sniping with implications for both theorists and practitioners.
Smith, C.A.P. and Vijayasarathy, L. (2003), “Sniping in
eBay: A Cultural Analysis,” Working Paper
From data gathered from hundreds of completed auctions on eBay’s
country specific websites, we show that sniping occurs quite frequently
(we define sniping as any bidding that occurs in the last minute of an
auction). Sniping occurs on average in 25% of our sample of 1600 auctions.
Sniping behavior was significantly different between countries with the
United Kingdom having the lowest, at 12%, and Sweden having the
highest, at 36.5%. We analyze sniping behavior across cultures and
suggest a model based on Hofstede`s cultural dimensions as assigned
to each sample country. Uncertainty Avoidance, Individualism and Long-Term
Orientation were significant with respect to sniping.