Bidder behaviours on eBay: collectibles and commodities

In the past decade, online auctioning has become one of the most successful business innovations. eBay is the global leader in this marketspace. This paper focuses on understanding bidder behaviour in order to provide insight into the design of information systems to support online auctions. In contrast to previous work that uses discrete bidding information, we implement a functional viewpoint of the bidding path. The similarity between paths is measured in the functional space using L 2 distance, which allows us to utilize K-means and PAM clustering methods. We apply this method to a large sample of collectible (1968 Camaro) and commodity (Digital Camera) eBay auction data to explore current bidding behaviours. We identify two different clusters of bidding behavior across both items which is associated with a specific winning percentage. Sniping exists as the dominant behaviour, but a hybrid approach has emerged (Sentry + Sniping).