Invited ReviewEmerging multiple issue e-auctions
Introduction
As pointed out by Geoffrion and Krishnan (2001), the Internet-driven networked economy is creating ample opportunities for operations research (OR) practitioners and theoreticians to work together with information systems and computer science specialists on WWW-based applications in many areas of business. Notable advances have already been made in financial services, electronic markets, supply-chain management, and travel-related services, among others.
The following quote from Bapna et al.'s article (2001) provides the basis for our paper:
“The Internet-driven networked economy is evolving to the point where firms are beginning to realize the enormous business value possible from a richly connected, global network of consumers and producers. However, full exploitation of these seemingly limitless opportunities will depend largely on the design of efficient mercantile processes that facilitate a wide variety of mechanisms for the exchange of assets, goods, and services. … Online auctions, brought about by the synergetic combination of Internet technology and traditional auction mechanisms, present a significant new dimension for mercantile processes, many of which are not yet fully understood.” (Bapna et al., 2001).
One such synergetic combination is the implementation of various auction types for procurement purposes in the supply chain. The procurement function is of great significance for many industrial corporations and public sector organizations, not to mention retail and wholesale businesses. Corporations and public sector organizations routinely arrange Request for Quotes (RFQs).1 In an RFQ the buyer normally specifies the merchandise, including desired quality, the quantity demanded, terms and time of delivery, terms of payment, and so forth, and requests the bidders to specify the price and quantity supplied if less than quantity demanded. If the buyer is willing to consider different quality levels and standards, and different terms of delivery and payment, she can indicate that in the RFQ. Once bids have been delivered, the buyer must make an intelligent decision about whom to award the contract. If the buyer knows precisely what she wants, including all terms of delivery and payment, and fully trusts the suppliers, the comparison boils down to comparing costs and choosing the supplier(s) offering best prices. Such an ideal situation rarely exists in the real world and the buyer has to make complex tradeoffs between price, different quality levels, and different terms of delivery and payment. Obviously, arranging RFQs is costly and time-consuming. Furthermore, the buyer may not be aware of all potential suppliers. In short, there are a number of potential complications and inefficiencies in the procurement process, some of which may be alleviated by the use of web-based electronic markets. In particular, electronic markets are expected to reach more qualified bidders, generate cost savings for buyers, create new means to increase sales for suppliers, and lower transaction costs for both, to name a few (Kambil and van Heck, 2002).
The earliest WWW-based trading sites date back to 1995, although the history of electronic markets/auctions is older. Lucking-Reiley (1999) provides an interesting discussion of the earliest WWW-based trading sites. Kambil and van Heck (2002) go further back in time and provide an insightful discussion of the history of electronic markets and the role that information and communication technologies have played in this process. They also describe how firms can design and profit from online auctions, particularly by placing auctions properly in their broader context of exchange processes and taking all stakeholders in a market into account. Today, WWW-based trading sites are plentiful. The largest B2C sites list hundreds of thousands of goods/services for sale. As pointed out by Bapna et al. (2001) and Gupta and Bapna (2001), until recently the electronic trading sites have been simple, concentrating on the price of the merchandize. The Internet, however, has allowed the inclusion of many novel and useful features into trading sites. We will discuss several such features below.
The roots of electronic auction and negotiation mechanisms are in the auction and negotiation theory. See, for instance, Raiffa (1982), Teich et al. (1994) (negotiations); or Kagel and Roth (1995), Milgrom (1989), and Rothkopf and Harstad (1994a) (auctions). Economists have in-depth investigated, isolated, single good auctions. The ascending price English auction, the descending price Dutch auction, and the Vickrey second-price sealed bid auction have been the most commonly studied auction mechanisms. The theoretical properties, particularly their revenue-generating properties, of the various auction mechanisms have been the subject of intense research. However, as pointed out by Rothkopf and Harstad (1994a) in their critical essay, it would be useful to expand the focus, because isolated, single-good auctions are not the most common and interesting ones from the practical perspective. Hence, there have been several extensions to the traditional auction paradigm in recent years. One active field of research has been multiple unit auctions. In the simplest case, the bidders are allowed to buy only one unit of the merchandise. In the more realistic case, such restrictions are not imposed. As Bapna et al. (2001), Rothkopf and Harstad (1994a), and Tenorio (1999), among others, have pointed out, the strong theoretical results obtained for isolated single good auctions, are not necessarily transferable to the more complicated multiple unit situation. Another extension is the development of combinatorial auctions, in which bidders desire to buy or sell bundles of goods rather than one single good (Rothkopf et al., 1998). Such combinatorial auctions have been of considerable interest to operations researchers and computer scientists, especially the computational issues associated with determining the winner and the final allocation (Kelly and Steinberg, 2000; Sandholm, 2002; Jones and Koehler, 2002).
In this paper, we focus on another extension to the traditional auction paradigm, one that is particularly relevant for the RFQ-setting, namely the multidimensional auction2 (Che (1993); Branco (1997); Koppius and van Heck (2003); De Smet (in press); note that some authors refer to this auction as a multi-attribute auction, e.g. Bichler (2000); Beil and Wein (2001)). In such a setting, there are multiple dimensions to a transaction, such as quality, delivery time and warranty terms that all need to be incorporated through the auction mechanism. If we have a multidimensional auction where quantity is also a relevant variable, we refer to it as a multiple issue auction. Perhaps since multidimensional/multiple issue auctions hold great promise for the improvement of B2B transactions, their development has largely been practice driven.
We review the state-of-the-art of multiple issue auctions, designed primarily for B2B procurement situations. We provide a classification of auctions based on their characteristics, and present a generic description of multiple issue e-auction mechanisms. We also discuss the role that OR has played in their design and implementation. The emphasis is on academic research, although we devote one section to overviewing relevant commercial auction sites. We also discuss what criteria one should use in judging the performance of auctions in the WWW environment.3 In an appendix we provide a glossary of terms and definitions for readers not intimately familiar with auction research.
Section snippets
A classification of auction situations based on their characteristics
At the DEXA Conference in London, September 2000, the participants drafted a classification of negotiation protocols.4 Many of the elements in our classification can be found in this `London Classification' as well. However, in the `London Classification' negotiations and auctions were both included and treated alike. We find this confusing.
A generic description of multiple issue e-auction mechanisms
For simplicity, we consider auctions of one commodity (good/service) and assume that the buyer demands k units (positive real number) of it. With n attributes besides price and quantity, the bids become (n+2)-dimensional vectors, where the n dimensions represent quality, terms of warranty, delivery, payment, etc. If all issues and attributes, besides price and quantity, have been `priced out' prior to actual bidding, each bid is simply a two-dimensional vector: (q,p), where q⩽k is a real
Incorporating preferences over multiple issues in auctions
A central issue underlying multiple-issue auctions is eliciting the bid taker's preferences over the issues involved. This problem is far from trivial and of great practical importance.
In the economics and management science literature the decision-maker's preferences over multiple issues, attributes, or criteria are often (at least conceptually) represented in terms of a multi-attribute value function when no uncertainty regarding the consequences of outcomes is present. Elaborate schemes have
Multiple issue auction mechanisms in practice
Searching the WWW10 we have found many more than a dozen companies that offer multi-dimensional RFQ, auction or negotiation software applications. We briefly characterize some of them (see Table 2). They all incorporate multiple attributes or issues into the bidding or negotiation process. Some of the sites are pure reverse auction or auction
Judging the performance of auctions in the WWW environment
Economics literature discusses a number of criteria for comparing auction mechanisms. They include allocative efficiency, revenue to seller (in forward auctions), and cost to buyer (in reverse auctions). Economists have traditionally considered isolated single good, price-only auctions. Our focus is, however, on multiple issue, multiple unit e-auctions. Traditional economists' criteria are not necessarily directly useful in such a complicated environment.
Conclusion
We have discussed the state-of-the-art of the emerging field of multiple issue, multiple unit online procurement auctions and measures to judge their performance. Published academic research on multiple issue online auctions is relatively scarce, yet commercial online auction sites are plentiful. In more than a few cases, academic auction researchers have been involved or have at least consulted in the development work. Still, it is difficult to escape the feeling that it would be in
Acknowledgements
This paper was written while the second and third authors were on sabbatical in the Industrial Engineering Department, Arizona State University, Tempe, Arizona 85287. The authors wish to thank the Foundation for Economic Education (Finland), the Foundation of the Helsinki School of Economics, and the Academy of Finland for financial support.
Glossary
Definitions and terminology
- Allocative efficiency:
- The sum of the producer and consumer surplus resulting from the auction, divided by the sum of the producer and consumer surplus, which would be realized if the bidders with the highest valuations received the contract.
- Auction owner:
- The initiator of an auction; the buyer in a reverse auction and the seller in a forward auction.
- Bid decrement:
- (In a reverse auction) the desired reduction in price (or total cost) from one bid to the next.
- Bid status: `active', `inactive',
References (58)
An experimental analysis of multi-attribute auctions
Decision Support Systems
(2000)- et al.
Combinatorial auctions using rule-based bids
Decision Support Systems
(2002) - et al.
On the role of discrete bid levels in oral auctions
European Journal of Operational Research
(1994) Approaches to winner determination in combinatorial auctions
Decision Support Systems
(2000)Algorithm for optimal winner determination in combinatorial auctions
Artificial Intelligence
(2002)- et al.
Multiple issue auction and market algorithms for the world wide web
Decision Support Systems
(1999) Multidimensional auctions
Economics Letters
(1988)- et al.
A parametrization of the auction design space
Journal of Games and Economic Behavior
(2001) A multiple criteria method for choosing among discrete alternatives
European Journal of Operational Research
(1981)Strategic jump bidding in English auctions
Review of Economic Studies
(1998)
A theoretical and empirical investigation of multi-item on-line auctions
Information Technology and Management
Insights and analyses of on-line auctions
Communications of the ACM
Automated negotiations: A survey of the state of the art
Wirtschaftsinformatik
Combined negotiation in e-commerce: Concepts and architecture
Electronic Commerce Research
The Future of e-Markets: Multidimensional Market Mechanisms
Comparison of weighting judgments in multiattribute utility measurement
Management Science
The design of multidimensional auctions
Rand Journal of Economics
Design competition through multidimensional auctions
Rand Journal of Economics
Combinatorial auctions: A survey
INFORMS Journal of Computing
Prospects for operations research in the e-business era
Interfaces
Online auctions: A closer look
Making Markets: How Firms Can Design and Profit from Online Auctions and Exchanges
Cited by (100)
The impact of suppliers’ overconfidence on the choice of auction formats under supply disruption
2024, Expert Systems with ApplicationsQuality manipulation and limit corruption in competitive procurement
2020, European Journal of Operational ResearchTruthful multi-unit multi-attribute double auctions for perishable supply chain trading
2016, Transportation Research Part E: Logistics and Transportation ReviewProcurement auctions and supply chain performance
2015, International Journal of Production Economics