Design and evaluation of feedback schemes for multiattribute procurement auctions

Multiattribute auctions, which allow bids on multiple dimensions of the product, are IT-enabled sourcing mechanisms that increase the efficiency of procurement for configurable goods and services compared to price-only auctions. Given the strategic nature of procurement auctions, the amount of information concerning the buyer's preferences that is disclosed to the suppliers has implications on the profits of the buyer and suppliers and, consequently, on the long-term relationship between them. This study develops novel feedback schemes for multiattribute auctions that protect buyer's preference information from the supplier and suppliers' cost schedule from the buyer. We conduct a laboratory experiment to study bidder behavior and profit implications under three different feedback regimes. Our results indicate that bidders are able to extract more profit with more information regarding the state of the auction in terms of provisional allocation and prices. Furthermore, bidding behavior is substantially influenced by the nature and type of feedback.