Only a few years ago, B2B exchanges were expected to completely alter conventional buyer-supplier relationships. Some e-business proponents foresaw a sweeping away of industry and national differences in purchasing patterns as Old Economy firms were displaced by the business models of their New Economy rivals. The reality, however, has proven to be otherwise, as evidenced by the number of independent public B2B exchanges that are no longer in operation.

Yet even as public exchanges have run into trouble, large industry wide exchanges are continuing to crop up and smaller, private exchanges are flourishing. “There has been a distinct shift away from the specialized public exchanges such as Ventro or VerticalNet, and towards large industry-wide exchanges run by consortia of incumbent firms, such as Covisint in the automotive industry and Transora in the consumer products sectors,” note Wharton management professor John Paul MacDuffie and Susan Helper, a professor of economics at Case Western Reserve University, in a new paper entitled B2B and Modes of Exchange: Evolutionary and Transformative Effects. “Firms are also drawn to private exchanges, even though these offer fewer potential benefits than the industry-level exchanges that promote standardization throughout a larger network.”

In their paper, part of a forthcoming book called The Global Internet Economy, edited by Wharton management professor Bruce Kogut under the auspices of Wharton’s Reginald H. Jones Center, the two researchers argue that B2B will be “evolutionary” rather than “revolutionary,” that it will continue to be used, but in a way that enhances, not replaces, individual companies’ business strategies. Companies and nations, MacDuffie and Helper predict, will develop e-business tools that reinforce old paradigms for purchasing and supplier relations.

“There are patterns of social interaction that are deeply imbedded in systems of procurement,” says MacDuffie. “The patterns differ by company history, country, laws, institutions, geography and resources. They aren’t easily overturned in some wonderful convergent vision, and their role in influencing the adoption of a new technology is often underestimated. They influence the way any given firm would want to use the Internet.”

Increasingly, the authors say, firms will choose among different forms of B2B depending on their strategy for managing the value chain. Rarely will one form of B2B address all aspects of that strategy, so companies are likely to use a mix of different exchanges. What’s more, says MacDuffie, “a lot of what the Internet and B2B do is reduce information and coordination costs. That’s a benefit that can absorb whatever kind of past approach to procurement and supplier relations you’ve had. So B2B doesn’t give you a strong incentive to move away from your traditional capability; in fact, it lets you realize some gains right away from sticking to your knitting.”

Focusing on the auto industry as a window into the potential impact of B2B on global industries and national economies, Helper and MacDuffie describe ways in which B2B can facilitate a wide range of business interactions. They say that the savings involved – some delivering one-time gains and some facilitating continuing improvement – should powerfully fuel the diffusion of B2B in the coming years. Current estimates of the savings from automotive B2B is estimated by Goldman Sachs to be more than $2,000 per vehicle, although the authors view this as too high because it overestimates how many purchases can be carried out via B2B auctions.

The authors cite five potential sources of performance improvement:

  • Automating the procurement process: The reduction in paperwork may bring costs per purchase order from $75-$150 down to $10-$30.
  • Interoperability: Easy communication between many users has been facilitated by the Internet. XML, a computer language still being refined, promises to make e-market communication even more standardized.
  • Auctions: Online auctions have a tremendous potential to reduce prices, as large numbers of suppliers compete for contracts.
  • Collaborative planning: Major savings can be achieved if plants in the supply chain can quickly view each others’ inventory levels and production schedule and plan accordingly.
  • Collaborative design: Designers in the same firm or different firms would be able to work in parallel. Various forms of proprietary software already exist, but the difficulties of linking them, and data security issues, still present obstacles.
  • Helper and MacDuffie show how these benefits tie in with the auto industry’s existing behaviors toward suppliers. Since the 1930s the U.S. auto industry has been largely characterized – with some exceptions – by “exit” relationships, and the Japanese industry by “voice” relationships.

In the exit mode, automakers resolve problems with a supplier by getting a new supplier. With the voice mode, automakers work with longtime existing suppliers to resolve problems. The authors assert that B2B can facilitate either approach. “If you want to be free to choose any supplier you want, and switch suppliers at a moment’s notice, then B2B helps you do that, through auctions. If you want to develop close relations with suppliers and exchange proprietary data, B2B makes that more efficient too,” says Helper.

Given what many would characterize as “cutthroat” auction markets, will the Internet weaken the close ties which some automakers, Toyota for example, have for decades nurtured with their suppliers? The authors predict that B2B exchanges will not push Japanese automakers toward exit mode; rather, it will help them strengthen their existing voice approach.

Certain B2B advantages will accrue to either exit or voice modes of exchange. On the inventory side, all links in the supply chain can have quick access to information about inventory levels and delivery schedules.

Helper talks about potential design-side advantages as well. “B2B can help facilitate the voice mode on the design side, but there is a need for trust. People worry that their intellectual property is going to appear in someone else’s hands. So if you have had an exit relationship with a supplier it will be harder for you to use these collaborative tools. On other hand, if you have already built a collaborative relationship, then the benefits of being able to share designs on the web and have suppliers around the world mark up each others’ designs are quite high.”

The move by a majority of the world’s largest automakers to establish their industry-wide B2B exchange – Covisint, founded in February 2000 – provides the authors with a specific B2B model to consider. They wonder whether Covisint can overcome technical and organizational challenges while also helping align the conflicting strategies across exit and voice modes and between automakers and their suppliers.

So far, they find Covisint’s progress to date impressive, considering the disappointing performance of most B2B exchanges today. In the first six months of 2001, Covisint managed transactions worth more than $33 billion – 13% of the $240 billion that Ford, GM and DaimlerChrysler buy annually. The main equity partners report that they are already seeing results from their investment; in July, Ford announced that it had saved $70 million in reduced paperwork and lower supplier prices, much more than the cost of its investment in Covisint.

Yet the consortium must work to encompass the different procurement approaches of the industry, the authors believe. “Unless Covisint can support the exit and voice models of procurement equally well, it may come to be dominated by certain automakers and shunned by others,” their paper suggests. The researchers’ interviews with consultants to Covisint suggest that Toyota is unlikely to join in any substantial way, for example, and VW and BMW have kept away from the start.

Helper and MacDuffie highlight the advantages of the B2B consortia approach not just for the automotive industry but for others as well. These include a guaranteed source of transactions; financial liquidity, and an ability to work towards standards that facilitate interoperability. Other prominent consortia B2Bs they name include:

  • Worldwide Retail Exchange, which includes more than 50 large multinational retailers such as Best Buy, K-Mart, and Safeway
  • Transora, a consumer products marketplace with over 50 companies including Kraft, Procter & Gamble, General Mills in the U.S. and Unilever, Cadbury Schweppes and Nestle overseas
  • Metalspectrum, a raw materials marketplace backed by over 20 of the leading metals companies including Alcoa, Kaiser, and Reynolds
  • Pantellos, a consortium of 21 energy and utility companies.

The authors hold that industry consortia exchanges are increasingly providing genuine cost savings through automation and standardization. They are also offering collaborative planning and design, although this latter option is more problematic. While the consortia are working hard to develop the software and to persuade trading partners that they can handle the intellectual property issues, some companies are gravitating toward the idea of a private B2B exchange between themselves and their supply chain and design partners.

Helper and MacDuffie offer the example of the telecommunications firm Ericsson, which recently installed supply chain software from Chicago-based PipeChain Inc. to aid collaborative planning. Even some of the founding members of Covisint are following suit: GM and the U.S. arm of DaimlerChrysler are each developing their own private B2B systems (GM SupplyPower and FastCar). Helper and MacDuffie wonder if the B2B consortia exchanges will ultimately be undercut by individual firms restless to make progress through private exchanges. “There will be no shortage of software vendors rushing to support private network activities of this kind,” their research notes.

While private networks overcome the governance problems of consortia, they may well lead to the problems of non-standard software that the open architecture of the Internet was supposed to prevent. For example, suppliers to the auto industry currently have to invest in three separate computer-aided design systems, because Ford, GM and DaimlerChrysler have each chosen different software. So far, Covisint has not done much to change this situation.

The role of public exchanges, ultimately, may be to help buyers and sellers find each other. Subsequent transactions might take place on private networks, where logistics arrangements could be optimized and proprietary information about designs could be safely exchanged. A recent Boston Consulting Group report on B2B described the goal of this strategy as “exploiting public exchanges to create private sources of advantage.”

To some extent, the software companies that write the code used by B2B exchanges are among the authors of B2B’s future. “It may well be the code writers at Oracle and Commerce One that have the most influence over how much Covisint affects the mode of exchange of these different companies,” says MacDuffie. The B2B software used might better facilitate exit capabilities such as auctions, for example, or voice capabilities such as collaborative design. And because Covisint is so big, any software that comes to predominate may end up laying the groundwork for other e-marketplaces as well.

Given the potential power of the software, the authors believe that B2B exchange founders – whether of consortia or private exchanges – would do well to make sure their software providers understand the workings of their particular industry. “I spent time in California talking to code writers,” says Helper, “and they have a very theoretical notion of how manufacturing actually takes place. They were surprised to find out that not all products can be easily described like a shirt in an L. L. Bean catalog. In the auto, aircraft and many other industries, it is very difficult to specify everything you want in a product or supplier. Designs and demand conditions change frequently, so what you really want is a firm that will respond quickly to conditions that are difficult to identify in advance.”

MacDuffie agrees, noting that one factor that brought down the pure startup B2B exchanges was an insufficient knowledge of the industry they were trying to serve. “Many people have been saying the next stage of the Internet will require the combination of deep industry knowledge with IT and Internet capabilities. Neither by itself is likely to be successful.”