Researchers at Carnegie Mellon University (CMU) have been developing a software solution designed to recognize fraud in online auctions. Specifically, it will analyze bidding patterns to look for shill bidding. What will the impact of this new software being developed by researchers at CMU be on eBay?
If released to the public, members of the eBay community could potentially see if sellers are driving the bid price up with shill bids. Shill bids are bids that the seller enters, or that are entered on the seller's behalf, in order to drive up the final bid price. When shill bidding is successful, the shill bids are overbid by a legitimate buyer. Shill bidding schemes have become increasing complex on eBay, with sellers using multiple accounts to enter shill bids on their items to increase their profits. A practice unfair to buyers, shill bidding, can be hard to spot, but now this software may make it easier for buyers to steer clear (and report) auctions from sellers that are involved in shill bidding and fraud.
Of course, the greatest utility of this software would occur if eBay were to adopt it and implement it on the enterprise level, weeding out shill bidders and sellers that fraudulently inflate their final values. However, it is because eBay takes a percent of the final value of an item that it seems unlikely that it would be in eBay's favor to adopt this solution. Should the software prove successful in reducing shill bidding, it would ultimately drive down the final values for many transactions, and subsequently reduce eBay's profit on those transactions.
Another potential issue with adoption of an anti-bid-fraud solution is eBay's existing difficulty providing quality customer service. While one would hope that should an anti-fraud solution be implemented, it would be one that works, nothing is 100% and it should be expected that there will be some mistakes. With numerous complaints already from people within the community who feel eBay's Customer Service is a bit lacking, what will this added influx of reported individuals, some of whom may be legitimate sellers, do to an already taxed communication system? It would not seem implausible that innocents would be caught in such a dragnet.
It is reassuring to know that there are developers out there who are looking into a solution for this problem. But, like most other security solutions, it will need to be adopted at the enterprise level to be truly effective.