An analysis of online procurement auctions: What determines the price?

Auctions have been used widely in the marketplace to obtain the best possible price for relatively exclusive items for a long time.  Over the past two decades, there has been a surge of auction activities in the marketplace, especially over the internet. Rapid technological innovation has facilitated the use of online auctions not only for business to consumer (B2C) transactions but also for business to business (B2B) transactions and government procurements.  Much of this growth is taking place in electronic procurement auction. This recent boom in online business to business (B2B) auction activities has created a need for methodical and in depth scrutiny of these auctions. Organizations using online reverse auctions need to have a better understanding of the operational strategies of their suppliers in order to design a comprehensive and efficient auction format. In this paper, we examine a B2B auction market to analyze some important aspects of the bidder strategies in B2B reverse auctions by studying 6052 business to business (B2B) online procurement auctions carried out by a large multinational firm between 2005 and 2007. The data set comprises of first-price auctions where the lowest bidder wins the contract at its bid price. For each item auctioned, the purchasing agent of the firm specified the auction formats, identified the potential suppliers and invited them to place bids in the auctions. The firm employed different auction designs that offered different degree of information to the suppliers. The participating suppliers at all point in time were aware of the bidding terms, time remaining in the auction and their own bidding history but had only partial or no information about the bidding history of their competitors.

Using multiple regressions, we find that: 1) auction success depends on the initial bid price.  2) Auction designs affect the dynamics of bidder strategy and the final price. For example, sniping or last minute bidding that is related to non-competitive bidding practices is higher in auctions that provide more information to potential bidders. 3) Winning bids are higher or auction success is lower when the same set of bidders bid against each other regularly. 4) Auction length does not have significant effect on winning bid price.