The Evolution of Electronic B2B Marketplaces
Susan S. DeSanti*
Director, Policy Planning
Federal Trade Commission
Remarks Before the Federal Trade Commission Public Workshop:
Competition Policy in the World of B2B Electronic Marketplaces
June 29, 2000
*/ The views I express are my own and do not necessarily reflect those of the Commission or any Commissioner. I would like to thank Bill Cohen, Deputy Director, and Hillary Greene of Policy Planning for their substantial assistance in the preparation of these remarks.
Thank you for joining us for this workshop. At the very outset I would like to thank the countless people who have helped all of us at the Federal Trade Commission ("FTC") begin to understand the still-developing phenomenon of business-to-business electronic marketplaces ("B2B e-marketplaces"). I am particularly grateful to the panelists who will be here today and tomorrow, as well as the many other business people with whom we have begun discussing these issues. We appreciate all of their time and energy, and we look forward to an ongoing dialog as B2Bs continue to develop and evolve. I also want to thank the many people within the FTC, as well as at our host site, the Department of Agriculture, who have worked so hard to put this workshop together. I especially want to thank my staff in Policy Planning, who have worked with creativity and dedication to put this workshop together, and Mike Antalics, Deputy Director for the Bureau of Competition at the FTC, who first had the idea to hold this workshop.
So why are we here? Our goal for this workshop is to lay the foundation for understanding how best to answer traditional antitrust questions in the context of new B2B technology that is driving reinvention of business on a global basis.
B2B electronic marketplaces - just like more traditional marketplaces - have the potential to raise traditional antitrust questions. For example, some B2B e-marketplaces allow information exchanges among competitors; some involve collaboration among competitors. Information sharing and collaboration among competitors may be pro- or anticompetitive, depending on the factual circumstances. But that fact alone does not answer the question of why we are here. Literally hundreds - perhaps thousands - of joint ventures among competitors are formed each year under the guidance of experienced antitrust counsel. The federal antitrust agencies in April of this year issued new Antitrust Guidelines for Collaborations among Competitors, which provide additional guidance in this area. So why have a workshop about B2Bs - what is new here?
The answer is: the technology. The technology that underlies B2B e-marketplaces and its potential for reinventing business processes are relatively new. All of what it can and cannot do is not yet clear, nor is it clear what all of its capabilities will mean for business transactions. Many of the businesspeople with whom we have spoken have started by giving the history of their company - dating all the way back to 1999. Many have described the development of B2B e-marketplaces as still "in the top half of the first inning" or, perhaps more accurately reflecting Internet time, "in the first few yards of the 100-yard dash."
What is clear is that B2B electronic marketplaces have the potential to generate remarkable savings. Such savings can enable businesses to operate much more efficiently and to lower prices to and provide greater innovation for consumers. Antitrust policy takes such savings into account, so it is important to understand them. This is especially true in those cases where savings arise because the new technology underlying B2B e-marketplaces can make markets more, not less, competitive.
So, we are here because we need to develop the best understanding possible of how to answer the long-standing, traditional antitrust questions in the context of the new technology that is driving a global re-invention of business systems. The process of advancing that understanding must be ongoing, just as B2B e-marketplaces themselves continue to develop. We hope this two-day workshop will lay the foundation for an evolving understanding of B2B e-marketplaces - the wide variety of models of these marketplaces, how and what kinds of efficiencies they generate, and, ultimately, their possible effects on competition. Only through such an understanding can we do the best job possible of ensuring that B2B e-marketplaces realize their promise for businesses and consumers.
Before we begin, I'd like to briefly set the stage for the workshop by outlining a few of the basic facts and concepts that may be helpful to keep in mind as we begin to understand B2B e-marketplaces. I do this with some trepidation, since many of these concepts are subject to varying views. But, with that caveat, let me proceed. My written remarks also highlight a few of the competition questions that B2B e-marketplaces may raise. I emphasize that these are questions; it would be premature (and unwise in the absence of particular facts) to attempt to provide answers.
First, a couple of historical precedents. Before B2Bs, there was "ERP," or enterprise resource planning, and the related process called "MRP," or materials requirements planning. These computer systems have been used by some companies internally to keep track of the products a company must buy in order to meet production schedules. Electronic Data Interchange ("EDI") is a system that has provided computerized documents through which businesses could exchange the considerable information that business-to-business transactions require.(1) For example, EDI is currently used by some large companies to transmit electronic forms such as purchase orders between buyers and sellers.
Now enter B2B e-marketplaces, which potentially represent a quantum leap beyond these existing technologies. This is the point at which one is tempted to say, "It's the technology, stupid." At its most basic level, business-to-business e-commerce refers to "on-line transactions between one business, institution, or government agency and another."(2) B2B e-marketplaces are software systems that allow multiple buyers and sellers to carry out sales and procurement activities over the Internet. In their first iteration, B2B electronic marketplaces are basically about taking one or more standard business practices - such as searching for, identifying, negotiating with, ordering and receiving from, and then paying an input supplier - and putting that process online. As we will hear today, B2Bs have the potential to expand to other roles as well, such as coordinating the design of a product between a supplier and customer.
Estimates place the business-to-business commerce that all of these operations represent as accounting for "more than 70% of the regular economy."(3) Given the size of the U.S. economy, it is easy to see that even small percentages of savings for that 70% could have significant efficiency implications. And most observers predict that ultimately B2Bs will yield large percentage savings.
To understand how those savings may come about, it helps to understand the types of goods that businesses buy and sell. Commerce between businesses involves two broad categories of goods and services: operating and manufacturing inputs. Operating inputs (also known as indirect materials) are used for maintenance, repair, or operation ("MRO") and do not become part of the finished product.(4) Purchases of indirect materials typically account for a large number of transactions, but of a relatively lesser dollar value.(5)
By contrast, manufacturing inputs (also known as direct materials) are raw materials or components used directly in the manufacturing process.(6) They typically account for fewer transactions, but the dollar value of each transaction tends to be much greater than for indirect materials. Moreover, direct material purchasing tends to be a specialized function whereas the purchasing of indirect materials may be fairly widespread within an organization.(7)
Whether a purchase involves direct or indirect materials may have implications for the type of B2B solution that is the best fit. Likewise, whether materials are bought and sold through long-term contracts or on the spot market also has implications for the best "B2B fit."
For example, B2B e-marketplaces may offer one or more of several possible price determination mechanisms, with countless variations. I will briefly address three: catalogs, auctions, and exchanges.
(1) Catalog aggregators, or "metacatalogs" normalize product data from multiple vendors so that buyers can easily compare it.(8) Once on-line, the pricing in catalogs can be revised with relative ease. Similarly, sellers can customize price lists to reflect agreements reached with specific buyers.
(2) Auctions can take many forms. A forward auction "let[s] multiple buyers bid competitively for products from individual suppliers."(9)
In forward auctions, prices move up with the bidding. A reverse auction lets multiple sellers bid competitively to provide product to individual buyers. Prices move down.
(3) Lastly, an exchange is a "two-sided marketplace where buyers and suppliers negotiate prices, usually with a bid and ask system, and where prices move both up and down."(10)
Each of these systems may make more or less business sense, depending on the types of goods involved, the nature of the industry, etc.
In sum, business-to-business commerce can assume as many forms in the online world as they already do in the offline world. And quite often characterizations used in the offline world remain applicable in the online world. I would like to bring up the one most important example where that is not the case, however. B2Bs are sometimes characterized as "horizontal" or "vertical" marketplaces, but those categories do not exactly track how that language is used in antitrust analysis, so let me provide some rough definitions. Typically, horizontal marketplaces help buyers purchase a wide variety of products. They can serve many different industries and span across markets.(11) By contrast, "vertical" marketplaces tend to serve particular industries and provide product expertise and in-depth content knowledge for that industry.(12)
Where are we now? There are approximately 600 B2B e-marketplaces currently running - or at least announced - with scores more being announced almost daily.(13)
The one constant is the promise of reduced costs. Just one example: During the months of February and March 2000, a small steel supplier used a B2B e-marketplace in the metals industry to "hook[ ] up with more than 50 new customers, 90% of whom he had never heard of before."(14) That same steel supplier participating through the same B2B e-marketplace - though this time as a buyer - also had the opportunity to purchase products from some of the largest manufacturers in the United States. As the CFO of that company recounted, "[he] used to have to make 20 phone calls to get one coil of steel."(15)
Now, not only have the transaction costs been reduced, but he now has purchasing relationships with large suppliers who would not have even "notice[d]" this small buyer pre-B2B e-marketplace.(16)
Who will actually participate in these B2B e-marketplaces remains to be seen. There are many more B2B e-marketplaces in the planning stage than on the ground running, and business executives who have established B2B e-marketplaces that are running have indicated that a fair amount of time and effort is necessary for businesses to implement them. We offer no conclusions before their time. And, at least for the short-term, we have some time. With the assistance of the business community, we plan to use that time for continued study, identification of key issues, and consideration at an early stage of critical factors that may help avoid problems later on.
Competition Analysis and B2B e-Marketplaces
What are some of the antitrust questions that might arise in the context of B2B e-marketplaces? I offer some preliminary views. The antitrust issues that will be relevant will vary depending on how specific marketplaces are organized, the extent and types of information exchanged, whether the marketplace involves joint selling or joint purchasing, whether there are multiple B2B e-marketplaces in the relevant market, and other factors. Given the diversity of B2B e-marketplaces and the extent to which they are continuing to evolve and develop, it would be premature to attempt to lay out an antitrust analysis of B2B e-marketplaces. However, the Antitrust Guidelines for Collaborations among Competitors(17) ("Competitor Collaboration Guidelines") recently issued jointly by the Federal Trade Commission and the Antitrust Division of the Department of Justice provide a useful starting point for questions that should be considered when a B2B e-marketplace is organized. I will identify some of those questions, as well as some questions in other areas not covered by the Competitor Collaboration Guidelines.
Rule of reason analysis(18) provides for consideration of efficiencies; given the significant potential of B2B electronic marketplaces to create efficiencies, this is important. As a general matter, the Guidelines recognize that "[c]ompetitor collaborations have the potential to generate significant efficiencies in a variety of ways. For example, a competitor collaboration may enable firms to offer goods or services that are cheaper, more valuable to consumers, or brought to market faster than would otherwise be possible."(19) In the B2B e-marketplace context, through real-time exchanges about the price, quality, and availability of products, for instance, "B2Bs can reduce the costs that buyers and sellers would otherwise expend to locate and negotiate with each other."(20) In fact, reduced transaction costs appear to be the one constant in predictions regarding the efficiencies of B2B e-marketplaces.
Under the rule of reason analysis in the Competitor Collaboration Guidelines, however, efficiencies only need to be considered "if the Agencies conclude that [the agreement being analyzed] has caused, or is likely to cause, anticompetitive harm." There are many questions that must be asked before the Agencies reach any conclusion of a likelihood of anticompetitive harm that then triggers an in-depth look at efficiencies. It is to those questions that I will turn next.
Nature of the Conduct
To ask questions about possible anticompetitive harm requires a focus on the conduct at issue: what is the nature of the agreement that may raise antitrust issues? I will separate out some of the types of joint conduct by competitors - whether in the context of an online B2B e-marketplace or offline - that have the potential to raise antitrust issues.
Joint Marketing and Purchasing
One of the first questions is whether a particular B2B e-marketplace involves joint selling or joint purchasing. Although many B2B e-marketplaces may provide opportunities for joint selling or joint buying in the future, a relatively small proportion of such marketplaces now entail such activities. For those that do, the Guidelines outline the basic issues.
The Guidelines recognize that agreements jointly to sell, distribute, or promote goods or services "may be procompetitive, for example, where a combination of complementary assets enables products more quickly and efficiently to reach the marketplace."(21) On the other hand, "marketing collaborations may involve agreements on price, output, or other competitively significant variables, or on the use of competitively significant assets, such as an extensive distribution network, that can result in anticompetitive harm" by creating, increasing, or facilitating the exercise of market power.(22)
The Guidelines similarly recognize the procompetitive and anticompetitive potential of joint purchasing. "Many [buying collaborations] do not raise antitrust concerns and indeed may be procompetitive."(23) Joint purchasing can help reduce transaction costs through scale economies in purchasing, reduce manufacturing costs, and produce other efficiencies as well. However, such agreements also can create, increase, or facilitate the exercise of market power, which in the case of buyers is termed "monopsony power." Monopsony raises the concern that by depressing price to suppliers, the joint purchasing agreement will reduce output below what would have prevailed in the absence of the agreement. Where a marketplace aggregates the purchase orders of multiple purchasers, it may be important to carefully distinguish between price reductions resulting from the buyers' and sellers' cost savings and price reductions that instead may result from the exertion of monopsony power by the buyers.(24)
Of course, it is important to identify the appropriate relevant market for purposes of determining the possibility of buying market power. For example, if all the high-tech widget manufacturers combine their purchases of pencils, they still may not have any market power in the market for pencil sales, since presumably many businesses worldwide purchase large quantities of pencils. If, on the other hand, all of the high-tech widget manufacturers combine their purchases of an input useful only in the manufacture of high-tech widgets, that would appear much more likely to raise significant antitrust concerns about joint buying power.(25)
The Guidelines make explicit the Agencies' recognition "that the sharing of information among competitors may be procompetitive and is often reasonably necessary to achieve the procompetitive benefits of certain collaborations."(26)
Nevertheless, the Guidelines also caution that in some circumstances the sharing of information "may increase the likelihood of collusion"and indicate that, other things being equal, greater concern is likely when the information that is shared relates to price, output, costs, or strategic planning.(27) The sharing of such information may facilitate collusion by, among other things, enhancing the ability to predict a competitor's prices from knowledge of its costs or to project or monitor a competitor's output level.
Experience demonstrates that information-sharing using modern information technology can facilitate complex collusion practices. In United States v. Airline Tariff Publishing Co.,(28) the Department of Justice charged that the defendant airlines colluded to increase prices, eliminate discounts, and establish other fare restrictions through a combination of advanced announcements of price changes and use of "footnote designators,"each of which was published by a computerized fare dissemination service.(29) The defendant airlines entered a consent agreement that prohibited price signaling and the exchange of certain information.
Depending on how they are structured, B2B e-marketplaces may also have potential to raise issues regarding information-sharing. There are several questions to be considered. Are the markets at issue otherwise susceptible to collusive activity?(30) Could the specific information divulged facilitate collusion? Is that information already available through other sources or might the ease or speed with which it could be transmitted through a B2B e-marketplace render it of greater concern? What firewalls or other devices might be employed to limit the exchange of competitively significant information without interfering with the efficient functioning of the e-marketplace? It all depends on the facts, but I suspect that many potential problems in the information-sharing area can be solved and that a little attention to these issues beforehand might smooth most of the bumps down the road.
Membership and Network Effects
Finally, B2B e-marketplaces raise the issue of membership. Two distinct concerns arise: exclusion, which involves restrictions permitting only select businesses to participate in a given B2B e-marketplace, and exclusivity, which involves rules prohibiting firms that do participate in a given marketplace from joining other B2B e-marketplaces or sending them much business.
For analyzing exclusionary conduct, a good place to begin is the guidance offered by the Supreme Court in Northwest Wholesaler Stationers, Inc. v. Pacific Stationary and Printing Co. There the Court observed that "not every cooperative activity involving a restraint or exclusion will share with the per se forbidden [group] boycotts the likelihood of predominantly anticompetitive consequences."(31) In so doing, the Court established that when a joint venture does not possess "market power or exclusive access to an element essential to effective competition" the rule of reason should be applied to membership claims.(32) The Court also emphasized that some exclusionary conduct may be supported by business justifications showing that the practices "were intended to enhance overall efficiency and make markets more competitive."(33)
Similarly, our concern regarding membership is its effect on competition. The analysis is likely to ask whether denial of membership significantly raises rivals' costs by denying or limiting access to a key input. Then it asks if this is likely to harm competition downstream. Because the analysis would be fact-specific, we might question how denial of membership affects firms in a particular industry and then what this means for competition in a particular market. Ultimately, the focus is on whether the challenged conduct harms competition, not whether it harms competitors.
For example, assume that some of the independent companies in a particular industry form a B2B e-marketplace for procurement. Through volume purchasing, they are able to secure savings that would be unavailable individually. The e-marketplace members then deny the application of a competitor who seeks to trade through that marketplace. That applicant complains to the Commission that exclusion from the B2B e-marketplace was an unreasonable restraint of trade.
In analyzing whether anticompetitive effects resulted from exclusion from the e-marketplace, many factors could come under consideration. Can the excluded firm acquire the same inputs or adequate substitutes at a comparable cost through other means? For example, if the excluded firm could readily form or join an alternative e-marketplace, or if the evidence demonstrated that participation in an e-marketplace was not necessary for acquiring comparably-priced inputs, it is less likely that there would be a competitive harm. Assuming that exclusion from the B2B e-marketplace precluded the firm from accessing necessary inputs at comparable cost, an examination of the competitive consequences in the relevant retail market might then become warranted. For example, what is the nature of the competition to be experienced among the non-excluded firms in the relevant market? How, if at all, did the excluded firm uniquely contribute to the market's competitive dynamics? At bottom, only exclusionary conduct that harmed competition rather than harming the excluded competitor alone would be challenged.
All of the competitive effects I have discussed so far implicate the markets for the goods bought and sold through B2B e-marketplaces. Yet the advent of electronic B2B commerce affects another market, too: the market for marketplaces themselves. If one B2B or a few e-marketplaces were to grow so large as to crowd out all others (online & off), it likely would warrant antitrust attention. Of course, there may be reasons for such consolidation. Perhaps network effects dictate that only a few marketplaces dominate the industry. It is also possible, however, that network effects may diminish after some volume level or that efficient, competing e-marketplaces may thrive given adequate interoperability.
As always, only recourse to the specific facts of the case can tell us what the result will be. But the sort of questions we tend to ask in such circumstances are these: Must a minimum level or percentage of sales or purchases be made through a B2B e-marketplace in which a seller or buyer participates? Do B2B e-marketplaces impose any other requirements affecting participants' outside purchases or sales? What consequences can be expected to follow from a decision to join, or not to join, a B2B e- marketplace? Can a given market efficiently support more than one B2B e-marketplace? These and similar questions can help us determine the antitrust implications in the market for marketplaces.
In sum, B2B e-marketplaces offer opportunities for substantial productivity-increasing effects, but also possibly raise concerns. Collusion, monopsony, and exclusionary conduct preceded the e-commerce revolution and will no doubt succeed it as well. The issues are not necessarily novel, but much of the factual context is new. Bearing that in mind, I have intentionally asked many more questions than I have answered and enter this workshop very much in a learning mode.
Once again, I appreciate this opportunity to speak with you about B2B e-marketplaces, and I look forward to what I hope will be a stimulating and informative discussion of the issues that they raise.
1. Bear Sterns, The Internet Business-to-Business Report, September 1999, at 9.
2. Glossary (visited June 20, 2000)<http://www.netmarketmakers.com/glossary>.
3. Erick Schonfeld, Corporations of the World, Unite!, E-Company, June 2000, at 125.
4. Steven Kaplan and Mohanbir Sawhney, E-Hubs: The New B2B Marketplaces, Harv. Bus. Rev., May-June 2000, at 98.
5. Sam Kinney, An Overview of B2B and Purchasing Technology: Response to Call for Submissions, June 2000, at 12.
6. 6Steven Kaplan and Mohanbir Sawhney, E-Hubs: The New B2B Marketplaces, Harv. Bus. Rev., May-June 2000, at 98.
7. Sam Kinney, An Overview of B2B and Purchasing Technology: Response to Call for Submissions, June 2000, at 12.
8. 8Glossary (visited June 20, 2000)<http://www.netmarketmakers.com/glossary>.
11. Morgan Stanley Dean Witter, The B2B Internet Report: Collaborative Commerce, April 2000, at 57.
13. 13Erick Schonfeld, Corporations of the World, Unite!, E-Company, June 2000, 123, 125.
14. 14Eryn Brown, Is the Internet Stronger than Steel?, Fortune, May 15, 2000, at 162.
16. 16Id. There are also B2Bs developing to pool the purchasing power of smaller businesses. Hundreds of school districts are joining B2Bs to assist in their purchasing of school supplies and equipment. The B2Bs may offer increased opportunities for the schools to achieve volume discounts through combining their orders in addition to other savings. Electronic School, E-Commerce Comes to School, June 2000 (visited June 28, 2000)<http.//www.electronic-school.com/2000/06/0600ewire.html>. As one senior business director for a school district said, "It's highly unlikely for [us] to call around town on a $1,200 purchase just to save $50 or $60. . . . It's just not efficient." Id. Even small individual savings could add up, given that in 1998-99 public school expenditures for supplies and equipment totaled $16-$26 billion. Id.
17. United States Department of Justice and Federal Trade Commission, Guidelines for Collaborations Among Competitors (April 7, 2000), reprinted in 4 Trade Reg. Rep. (CCH) ¶ 13,160.
18. The analysis of collaborations among competitors traditionally has been conducted in two ways - rule of reason and per se analysis. Under per se analysis, certain conduct is deemed so likely to be harmful that it is quickly condemned. The Competitor Collaboration Guidelines suggest that it would be a rare case in which setting up a B2B e-marketplace itself would be viewed as per se illegal, especially since the Guidelines specifically recognize that an efficiency-enhancing integration may be accomplished by contract - that is, the integration of assets takes place not in physical space but "virtually," as is typically the case with B2B e-marketplaces. Competitor Collaboration Guidelines, § 3.2. Of course, for on-line B2B e-marketplaces, just as with off-line competitor collaborations, particular agreements among competitors on matters such as price or output might merit per se treatment if not reasonably necessary to the efficiency-enhancing integration. But so long as the underlying formation and operation of the B2B e-marketplace appears procompetitive, and associated agreements appear plausibly to represent legitimate efforts to achieve procompetitive benefits, per se treatment appears unlikely.
19. 19Competitor Collaboration Guidelines, § 3.36.
20. Charles F. Rule, Mark E. Plotkin, and Michael J. Fanelli, B2B or Collusion?, LegalTimes.com, April 3, 2000, http:www//5.law.com/dc-shl/display.cfm?id=2972. See also, Jonathan B. Baker, Identifying Horizontal Price Fixing in the Electronic Marketplace, 65 Antitrust L.J. 41, 44-46 (1996).
21. 21Competitor Collaboration Guidelines, § 3.31(a).
22. As the Guidelines explain, "[m]arket share and market concentration affect the likelihood that [a particular agreement] will create or increase market power or facilitate its exercise." § 3.33. However, when evaluating market share for a collaboration, aggregating shares might overstate the market power of the collaboration. For example, if the participants in one B2B e-marketplace also sell or buy through other B2B e-marketplaces, it would overstate the market power of the first marketplace simply to aggregate all of the shares of the participants and attribute them to that marketplace. The Guidelines provide for adjustments to more accurately assess market shares and the likelihood of anticompetitive harm. Questions to ask in this connection include, among others: To what extent is the collaboration non-exclusive such that participants have the ability and incentive to continue to make purchases or sales either independently or through competing B2B e-marketplaces? What is each participant's financial interest in the collaboration and to what extent will that affect its incentives to make transactions independently or through other B2B e-marketplaces? To what extent does the B2B e-marketplace's organization or governance structure affect independent competition? See Competitor Collaboration Guidelines, § 3.34.
23. 23Competitor Collaboration Guidelines, § 3.31(a).
24. See, e.g., Marius Schwartz, Buyer Power Concerns and the Aetna-Prudential Merger, November 1999 at 3 (visited June 27, 2000) <http//www.usdoj.gov/atr/public/speeches/3924.htm>.
25. The Guidelines also identify potential concerns with anticompetitive effects in downstream markets: "Buying collaborations also may facilitate collusion [in downstream markets] by standardizing participants' costs or by enhancing the ability to project or monitor a participant's output level through knowledge of its input purchases." Competitor Collaboration Guidelines, § 3.31(a).
26. 26Id. at § 3.31(b).
28. 58 Fed. Reg. 3971 (1993).
29. Id. at 3976.
30. The Competitor Collaboration Guidelines in this regard look to market concentration as well as the various factors - such as product or firm homogeneity and the gains from deviating from a collusive arrangement - familiar from the Horizontal Merger Guidelines' discussion of coordinated interaction. Competitor Collaboration Guidelines, § 3.33 (incorporating factors discussed in the Horizontal Merger Guidelines, § 2.1).
31. 472 U.S. 284, 295 (1985).
32. Id. at 296.
33. Id. at 293.