Tom Davenport on Strategic Sourcing

WHEN I FIRST BEGAN TO RESEARCH e-commerce networks and markets with two colleagues (Jeff Brooks and Sue Cantrell) a year ago, I was not quite as bullish as they. I wasn’t as sure that e-markets would transform business-to-business relationships, or as optimistic that huge savings from purchasing efficiencies would be created. The diligent reader of this column (Joe is his name, I believe, and I’m very grateful to him) will remember that when I wrote about the topic in the April 1, 2000, issue of CIO (“Nets Upon Nets”), I was a little schizophrenic. That column was a dialogue between the skeptic and the visionary observer of e-markets, and although I said I thought the visionary won the debate, Joe might have observed that there were more skeptical than visionary paragraphs.

In the (much less important) world outside of my own mind, it appears that the skeptics are winning out. It wasn’t that long ago that I could read a press release each day on the formation of a new e-market; now I can read a daily dose of negativism in the IT and e-commerce press about these nascent companies.

Certainly there was potential for these e-markets to reshape B2B commerce. They offered access to more suppliers and customers, the potential exchange of virtually all types of information, and the ability to dynamically price goods and services through mechanisms such as auctions and yield management. The development of e-market standards across entire industries offered the possibility of new communities of commerce in which it would be just as easy to transact with another company as with your own. Because there were some real advantages to independent e-markets and because the world loves to think that a bunch of e-Davids could slay the B2B Goliaths, the whole concept took wing with reporters and equity analysts. It didn’t hurt that B2B e-markets came along at about the time that business-to-consumer businesses were flagging a bit.

But we were all being a little naive about how B2B e-markets would take off. They have not taken off?at least most of them aren’t thriving, and some are beginning to die. Our data, based on a sample of 120 e-markets, suggest that the median e-market is doing only 175 transactions a month, with a median value of $6,500 per transaction. That adds up to about $1.2 million a month passing through the market. Transaction fees are the primary means of getting paid in e-markets, and they are at best around 5 percent or so. That means that the average e-market is getting only about $60,000 a month in revenue, which will barely pay for the maintenance of the foosball table.

The big guys?that is, large companies that buy and sell within existing markets?have moved in, even though thus far there is little booty to plunder. They have developed their own e-market consortia working with other players in their own industries. It was only natural that large companies in, say, the oil business, would want to develop their own marketplaces rather than hand over big (well, they thought they would be big, and maybe they will be eventually) transaction fees. In March, April and May of 2000, consortia e-markets were formed in the automotive, aerospace, forest products, utilities, food, airline, rail, energy, chemicals, hospitality and computer industries. If you were an independent e-market, would that lineup intimidate you? I suspect it will scare many independent e-markets into selling themselves or their services to consortia.

If competition from e-market consortia weren’t enough, there are three other factors that have inhibited e-markets’ growth.

Integration is a huge issue in e-markets. It refers to the ability of one company to integrate its buying and selling systems with those of the e-market, or with other companies that are participating in the market. Certainly some progress has been made in the ability to connect interorganizationally; companies such as WebMethods have eased the way somewhat with integration tools. Eventually, companies will undoubtedly be able to issue purchase orders through their own enterprise systems, and have them seamlessly pass through an e-market to the enterprise system of another company. But don’t hold your breath while this is all being worked out; it will take many years.

The difficulty of implementing information and process standards. Sure, you can get a bunch of well-meaning people together from various companies in an industry, and they may all agree that a “9 millimeter stainless tube” will mean just that. Several industries?most notably the electronics industry through RosettaNet, an early e-market consortium?have already done so. But it is another matter altogether for everybody in an industry to actually implement common information and processes. That would mean reengineering an awful lot of businesses. It will be even harder for companies to agree on how to describe highly engineered or complex components. Even if these steps could be accomplished, there are undoubtedly dissidents within the organization who will say, “Why should we make all our processes and information the same as our competitors?so that we can compete only on price?” Sometimes we should listen to such dissidents.

The relationships companies have developed with their suppliers. Of course, they’re not really a problem from any other perspective; they are rather a great asset. Which is why companies don’t want to dissolve them in favor of an e-market in which they can save a few bucks by purchasing from suppliers they don’t know. My favorite example of this is Toyota, which was asked by the Big Three automakers to join Covisint, their e-market. Toyota managers asked themselves, “Should we dissolve the supplier relationships we’ve worked on for decades, which are a critical aspect of our quality, which is a critical aspect of why we’re a great car company?” They considered the issue for at least a second or two. Their response was, “OK, we’re in…for office supplies and a few nuts and bolts.” It would have made zero sense for Toyota, or any other organization with close supplier relationships for key components, to give up those relationships for a few dollars or yen saved through an e-market.

The end result of these factors is that strategic sourcing is alive and well. Companies continue to buy much of what they need and sell much of what they make to long-term trading partners with whom they have cultivated deep relationships. To be sure, many companies are developing private extranets to facilitate connections with existing partners. These private networks are similar to electronic data interchange connections, but they involve no value-added network intermediary and allow for a much broader range of information types to be transmitted. They’ll be particularly important in the next phase of interorganizational relationships, in which companies will truly collaborate, not just exchange information. The private networks do involve the Internet (albeit in a private form), but they don’t involve revolutionary change. Instead they are just another step in a long efficiency dance involving companies and the other companies with which they do business.