Two years ago, business-to-business (B2B) marketplaces barely existed, even as an idea. Today, there are plans to establish thousands of these on-line businesses. New deals—many involving some of the world’s biggest companies—are announced almost daily.

What is the attraction? For buyers, B2B marketplaces promise not only to deliver more competitive prices but also to rid the supply chain of a host of inefficiencies. Hopes are so high that money can be made before even a single transaction has been concluded: the mere announcement that General Motors was to form a marketplace powered by technology from Commerce One caused the latter’s share price to shoot up by 23 percent. GM benefited as well: recognizing that a successful marketplace needs liquidity, Commerce One gave the automaker warrants for up to 20 percent of itself, depending on how much liquidity it brought on-line.¹

Amid such optimism about B2B marketplaces, it is easy to forget that they are still in the very early stages of development (see sidebar "What is a B2B marketplace?"). The vast majority of Fortune 500 companies have yet to form or join a major marketplace, meaning that much value is still up for grabs, both for the buyers and for the marketplaces that succeed in attracting them.

Five factors for success

Indeed, no B2B marketplace is fully up and running so far. The technology standards needed to connect buyers and sellers, such as XML (an enhanced World Wide Web language designed to support commercial transactions), are still in development; meanwhile, many software vendors are advancing their own versions of these standards. Other technical hurdles remain, too. Consider the cost and time that must be devoted to building intracompany enterprise resource-planning (ERP) systems,² which have grown into a market worth some $20 billion a year.³ Then think how much more difficult and costly it will be to create a cross-enterprise system that integrates many different supply chains.

In addition, payment systems that ensure timely and secure settlement must be created; logistics systems must ensure that items ordered show up at the right places and times; and all this information must be provided in real time over shared networks. Most important, the marketplace must do whatever is needed to attract and retain enough buyers and sellers to thrive.

Given all this, it is quite likely that some marketplaces built at great expense will fail. Five factors will influence the outcome of the shakeout.

1. Early liquidity

As in all markets, the B2B marketplaces most likely to succeed are those with the greatest liquidity. The more buyers trade on a marketplace, the more suppliers will be tempted—or forced by strong buyers—to join them, which in many cases will lead to lower spreads. (Experience in many on-line financial markets shows that they survive only if they can maintain reasonably low spreads.) Moreover, successful marketplaces will endeavor to bring as much buying volume on-line as quickly as possible, in the same way that Amazon.com spent millions of dollars to acquire customers, reasoning that once it had them, it would be easy to sell more goods across more product categories.

2. The right owners

Those companies that bring liquidity to the marketplace are its logical founding partners and therefore have the best chance of capturing the value it creates in the form of lower prices for the goods and services purchased through it. Most industry-focused marketplaces, such as those announced in the automotive and petroleum industries, will be built around large pools of buyer spending. But the buyers are not always in control. A seller (such as Cargill, the producer of basic food ingredients) that has a wide range of buyers might also be in a position to take ownership of the marketplace. In functionally based marketplaces, where the buying and selling sides are likely to be more fragmented, the natural owner may well be a Web-based intermediary that steps in to roll up the volume on behalf of buyers and sellers.

As many Web-only intermediaries have learned after trying to build businesses around purchase-and-sales bulletin boards or content such as industry data, marketplaces established by companies that can’t readily generate liquidity are not likely to thrive. Large buyers—for example, companies in the chemicals industry—at first participated in these forums but have since come to recognize the value of their spending volume and are now building their own sites.

3. The right governance

If suppliers notice that buyers can’t come to agreement about issues such as common specifications, they will exploit such divisions

For marketplaces—especially multibuyer marketplaces—to work, good governance is needed to ensure that buyers agree on the terms of their involvement and commit themselves to supply liquidity. If suppliers notice that buyers in an on-line marketplace can’t agree about issues such as common specifications or how to limit the number of suppliers, they will soon exploit such divisions to drive through their own deals outside the marketplace, offering rewards to defectors. Conversely, to win business, suppliers will reward well-run marketplaces with better terms (see sidebar "A new sales channel").

Good governance is the way to avoid conflict between different buyers. It is likely to require the appointment of a team of managers, loyal to the marketplace, who are independent of the buyers but empowered by them to negotiate contracts with suppliers on their behalf. Such a management team will be critically important if a big buyer had a hand in setting up the marketplace, for rival companies will not participate in it unless they believe that it is neutral.

The management team will also have to make critical decisions about such issues as technology, settlement (which payment services to use), buying policies (for instance, rewarding compliance and punishing defections), and the resolution of disputes among buyers or between them and suppliers. These decisions can’t be made by committees consisting of buyers, since individual interests will bog down the process and therefore constrain the flow of liquidity to the marketplace.

4. Openness

To make the marketplace as efficient as possible, it must attract as much buying and selling as possible, which means that it must operate under open standards. Most marketplace architects, such as Ariba and Commerce One, are committed to open technical standards but still have to agree on them. When such standards will emerge is still unclear, but in the meantime start-ups such as WebMethods are developing the equivalent of translation software to ensure open communication between buyers and sellers as well as between marketplaces.

5. A full range of services

Many marketplaces are being formed with a view to making them efficient from a pricing perspective. But simply pushing down prices won’t suffice in the longer term. Companies that already use purchasing as a source of competitive advantage want to cut their total cost of ownership (TCO)⁴—that is, not just the price they pay for their supplies but also other supply-related costs, such as excess inventory, spoilage, and rush orders. Many marketplace architects are therefore thinking about how to provide other services that are related to the supply chain, such as fulfillment logistics, the management of customer relations, and tracking the performance of suppliers.

How buyers should proceed

Despite the difficulties of establishing B2B marketplaces and the fact that they will inevitably involve a learning process for all concerned, big buyers can’t afford to ignore the trend (see sidebar "First on-line"). Doing nothing isn’t an option: to reap the benefits of improved purchasing, competitors are moving quickly to integrate B2B marketplaces into their electronic-

commerce strategies. Financial markets are handsomely rewarding buyers that are building such marketplaces. And suppliers are planning to counterattack by forming their own Web-based purchasing systems for their customers to use. Big buyers should thus embark on an all-out campaign to create or join a marketplace and bring their liquidity on-line.

To build successful marketplaces, companies should pay especially close attention in two areas: managing change and finding the right partners.

Managing change

Since liquidity is the key, companies will have to get their off-line spending on-line as quickly as possible. This is no mean feat: it took Amazon three years to bring just under 2 percent of book purchases on-line, and analysts expect it will take an additional three years for 15 percent to be brought on-line. And in contrast to the business-to-consumer (B2C) world, where purchasers generally act for no one but themselves, many B2B buyers are not principals but agents, buying on behalf of others. Yet another layer of managers and companies must therefore be persuaded that on-line purchasing is the way to go.

To get everyone on board the project, strong support from the chief executive officer is needed, and this means setting up a steering committee comprising the CEO and key business unit leaders as well as a program office that is directed by a senior manager. The CEO must play a visible role in announcing and supporting the entire effort, working with unit managers to emphasize the benefits to the whole company of funneling purchasing transactions through the marketplace.

Of course, the program leader must determine the order in which different kinds of purchases move on-line, taking into account the possible savings to be made in each of them and the difficulties involved in the effort. An electric utility, for example, might very well decide to buy nonstrategic items such as wire and cable through its marketplace well before it bought more important goods and services there, since changing suppliers or specifications on, say, mission-critical power station equipment could mean real trouble. High-potential categories in the early going frequently include nonproduction (or white-collar) "indirects," such as personal-computer hardware and software or travel and entertainment, as well as factory floor (or blue-collar) MRO (materials, repair, and operations) goods, since buyers tend to be somewhat less concerned about their specifications and the companies that supply them.

Technologically unsophisticated procurement employees may feel threatened by the possibility that software could replace them

Resistance to change within a company can be a huge obstacle to bringing its purchases on-line. For every company that is well on its way to achieving world-class purchasing and supply management, many others are only in the early stages of attempts to improve their procurement practices. Lots of people charged with procurement are not technologically sophisticated and have little or no experience in buying on-line. Some may feel threatened by the possibility that they could be replaced by software. They need to be told about the benefits of on-line purchasing—more time for contact with strategic suppliers, for example, and less paper pushing—and about its impact on long-standing supplier relationships. CEO support will be critical in motivating the laggards.

Finding the right partners

In all likelihood, a large company wishing to take its purchases on-line will choose as its partner a technology firm that provides the market architecture. The many B2B marketplace architects differ in important ways. Ariba, for example, has a good deal of experience developing procurement software, having signed up the largest number of customers to buy their own supplies over the Web. Commerce One points to its early entry into the B2B market space. Oracle argues that its B2B marketplace customers benefit from its experience designing ERP systems.

Before choosing a company to design the architecture of the marketplace, large companies should explore at least eight issues.

1. Experience. How many live installations has the marketplace architect worked on? What is the transaction flow through them? Experience counts in winning the race to liquidity.

2. Integration. How will the marketplace software be integrated into the company’s or companies’ existing systems, and how long will it take to integrate them? It is important to avoid a long path to integration, which will slow down the push to achieve liquidity on-line as well as delay the benefits of on-line buying. To find out how hard it is to integrate the new systems into the old ones, talk to clients of the architecture provider.

3. Services. What additional services, such as supplier identification and qualification, payment, credit, or logistics, will be offered in the marketplace? When will they be available? As noted earlier, the best marketplaces will offer more than just efficient pricing.

4. Ownership. Many marketplace agreements include equity stakes not only for buyers but also for technology providers. What will the marketplace architect provide in return for this stake? Make sure that incentives are tied to building the marketplace as quickly and efficiently as possible.

5. Technology. How open will the marketplace be? Open standards are needed to foster a rapid growth in liquidity. What guarantees can be offered to ensure that buyers will be able to communicate and transact with another marketplace if they wish? How robust is the underlying technology: can it handle millions of transactions a year securely? What happens if the marketplace crashes and production lines are starved of inputs?

6. Suppliers. How can the marketplace give suppliers quick, easy, and cheap access to its offerings? In all probability, the larger the number of suppliers competing for business, the higher the liquidity of the marketplace. And on the other side of the equation, the assumption that buyers will show up on their own is dangerous: in most cases, they too will need help getting on-line. ANX (the Automotive Network Exchange)—an early attempt by Chrysler (before its merger with Daimler-Benz), Ford, and GM to form a marketplace—failed partly because suppliers couldn’t log on and start selling their goods and services cheaply and easily.

A buyer would be ill-advised to demand that its suppliers go on-line; they must be persuaded that doing so will help them sell more goods and services to their current customers, gain access to potential new ones, or cut logistical costs in areas such as order processing and returns. To contain the costs borne by suppliers, buyers should give them suitable technology tools, such as ’’hosted’’ software (from an application service provider) that runs over the Web and requires neither deep integration into their existing systems nor costly in-house maintenance.

7. Cost. How much will all this cost? Since no fully functioning marketplace has yet been built, what benchmarks and guarantees could ensure that the expense of the system doesn’t spiral out of control? This issue is especially significant in the context of thorny issues such as integration deep into the supply chain—an integration involving many often incompatible systems. (Remember those costly ERP installations!)

Also consider the various methods of payment. Some marketplace participants are essentially paying for the technology in the form of equity; others have the money to pay up-front license fees; still others prefer the pay-as-you-go approach, in the form of transaction fees. The bottom line is that marketplaces won’t create value for buyers if too much money is spent on technology.

8. Buyers. Having the right buyer-partners on board is equally important. Many buyers underestimate just how tough it will be to partner with other parties and jointly cede control over matters such as the negotiation of supplier contracts to the marketplace-management team. Savvy buyers will create incentives (such as equity stakes in the new marketplace) to tempt buyer-partners to bring their spending on-line and subsequently impose penalties for opting out of agreed-upon contracts. They will also try to bring buyers from a range of industries into the marketplace. Since auto companies and engine makers, for example, source from similar supply markets, companies in those two industries might make ideal marketplace partners that could not only pool their spending but also have the advantage of not being direct competitors.

Finally, the chemistry between buyer-partners is critical. In any industry, some companies have a history of collaboration with competitors, and others are bare-knuckle adversaries. Instead of aiming for high numbers, it makes sense to assemble a group of buyers that can work together smoothly to bring their liquidity on-line.

B2B marketplaces will fundamentally change how buyers and sellers interact. The remaining uncertainty revolves around how quickly such marketplaces will develop. If the problems of standards, security, costs, and incentives are not resolved, buyers could end up wasting millions of dollars building doomed marketplaces, without ever figuring out how to direct spending to them.

The likelier outcome is that within five years, many large and successful marketplaces will have formed around single buyers and a few around a number of them. Some 50 percent of the purchasing requirements of big companies will probably be satisfied on-line. Much, though not all, on-line spending will likely involve white-collar and industrial MRO goods and services and nonproduction indirects.

A buyer’s nirvana—with up to 80 percent of total spending brought on-line and real-time inventory checks, quality data, and supplier performance tabulations—is still only a distant hope. Nevertheless, it is a goal worth aspiring to.