eMarkets: Fixing What Is Broken

Business to business marketplaces on the Internet, or eMarkets, are not having the kind of success or producing the results that were expected. Despite the fact that the business fundamentals for many eMarkets are sound, the eMarkets are having difficulty in attaining liquidity, where margins are workable and growth is self-sustaining.

The problem is due to a failure to deliver value across the entire chain of operations and transactions reaching from the manufacturer, through the distribution channel, and finally to the buyer and to the product's ultimate use. Instead, eMarkets have generally focused just on connecting buyers with sellers -- on "order matching" -- in an attempt to build transaction volume. This is a doomed strategy. The failure to remove substantial costs from the full transaction ends up providing too little value to any of the market participants, and the eMarket stagnates.

Relationships with buyers are particularly critical to eMarket success. This is because buyers have the most to gain from the initial operation of the eMarket if it can deliver its full potential value. But one unfortunate irony of eMarket development has been that eMarkets have had to spend most of their initial time and development on working with sellers just to get a catalog up so that the market can open its doors. The result is that most eMarkets have made only minimal investment in getting buyers fully on board and on meeting the full range of buyer needs. Weak, poorly managed relationships with buyers lead directly to low transaction volume and a sick eMarket.

The cure is to change the priorities during eMarket development. Rather than focusing first on building transaction volume and later on customization to meet individual buyer needs, the eMarket must focus on both goals in parallel. Systems to customize and manage relationships are just as important as systems to match orders when the eMarket opens its doors.

Companies will do business with an eMarket because the eMarket knows how to do business with them. Making this simple fact the core principle guiding eMarket development can get eMarkets moving forward again.

The Problem and the Question

Internet based B2B marketplaces -- I will call them eMarkets -- were supposed to be the slam dunk opportunities for new business growth. Unlike consumers, businesses are already online, have high speed connections, and have strong incentives to do things differently. A great deal of order processing still uses paper, phone, and fax. The old alternative, EDI, is too expensive and way too cumbersome for most business applications. Internet infrastructure provides a workable and inexpensive alternative that is here today and ready to roll out. There is real money to be saved and ready access to the tools that can save it. How could eMarkets go wrong?

But go wrong they did. Over this past year we have seen the failure of a number of new eMarkets and have followed others as they have had difficulty getting more than just a small foothold in the marketplaces that they hoped to revolutionize. Even the first movers in this category, the companies that were highest on the learning curve and that had their pick of opportunities, are running into trouble. Ventro, the granddaddy of exchanges and first to address the needs of the chemical industry, whose stock once traded for $231 a share, is now trading at just over $5 a share. VerticalNet, once trading at $148 a share, is bringing only $20 a share.

What's wrong? What can we learn from the last eighteen months of startups and industry initiatives that have not made their numbers or met expectations?

It Seemed So Easy

It was about a year ago that I was sitting in a conference room in North Carolina with a half dozen purchasing agents from a company that operated, maintained, and managed the business for several dozen large container ships that moved freight around the world. I was trying to get my hands around the dimensions of an eMarket opportunity that interested one of our clients, and so was trying to understand in detail the purchasing processes and systems that this ship management company used to order equipment and parts for repairs. Everything was done by phone and by fax. Most orders required that the purchasing agents obtain three bids. Since the bidding process sometimes involved questions and clarifications, just getting a single bid could involve multiple exchanges of faxes. One of the purchasing agents estimated that they spent more than two hours each day, as a group, simply sorting through the faxes and putting them in the right folders, getting each fax tied to its associated order. Thinking in terms of my client's interests, all of this was very exciting news. It seemed that my client should have no difficulty at all in bringing new efficiencies to this system, reducing costs and decreasing the time required to complete a purchase. It was like money lying on the floor waiting for someone to stoop down and pick it up.

In talking to another ship management company, I learned that the paper, phone, and fax system was so unwieldy that the company had great difficulty getting a clear idea of what it had purchased and what it had in inventory in ports and on ships. The purchasing system was so out of control that the company felt exposed to risk of losses from theft and from kickbacks and other graft in the purchasing process. Once again, the business motivation to replace the old system with something new was painfully obvious. Another slam dunk.

The opportunities to use eMarkets to solve substantial business problems are, of course, not at all unique to ship management. We have found, for example, that the building materials market is one in which obtaining critical information about pricing is surprisingly expensive and time consuming. It currently takes up to five hours, once again using phone and fax, to verify pricing, supply, and shipping arrangements to complete an order. Market inefficiency in communicating pricing information creates constant concern among buyers that they may not be getting the best price. An eMarket should be able to bring much needed transparency to this market while, at the same time, making it much easier to exchange the rest of the information required to complete a transaction. Similarly, the business of ordering fresh fruit for use in juice processing and then of buying and selling the processed juice for bottling has been managed through phone and fax. In this complex market, which runs on both long term contract pricing and spot pricing, it is clear that using the Internet's ready made world-wide network can substantially reduce the cost of doing deals.

The examples proliferate. Phone calls followed by faxes are a poor way to place and track orders. Faxes do not integrate into other systems without rekeying the data. The point-to-point interactions between many buyers and many sellers are an obvious source of pricing and processing inefficiencies. It is clear that, over time, each of these markets will move to fully electronic transactions.

At the same time, eMarket development has proceeded more slowly and with more difficulty than many investors, entrepreneurs, buyers, and sellers had anticipated. At a recent conference focused on the use of e-business in the chemicals and plastics industries, there was certainly still a sense that the changes would come, eventually. As Eastman Chemical Company CEO Earnest Davenport said, "E-business is not the electronic version of the pet rock. It is not some fad that we can watch go by."[1] But it was also clear to conference attendees that deriving the anticipated benefits from eMarkets was going to take longer than anyone had anticipated as recently as earlier this year.

It is important to try to understand what is making eMarket development so difficult so that we can address the problems directly. eMarkets really can save money and make money; we just need to figure out how.

eMarkets: What and Why

Before we look at what has gone wrong in eMarkets, we need to first understand what they are and why, over the long run, they should eventually work, even if not in quite the same way and with the same ease as initially expected.[2] eMarkets take a number of different forms.

In all cases, an eMarket is a hub that connects multiple suppliers and multiple buyers. It is this "many-to-many" dimension of an eMarket that enables it to address important, difficult problems such as improving the efficiency in the exchange of pricing information for building materials or the creation of a more agile spot market for fruit juice. But, as we will see, it is also this many-to-many characteristic of eMarkets that makes them complex to setup and difficult to operate.

Another key characteristic of eMarkets is that they create "transparency" for both pricing and processes. Market inefficiencies, such as the difficulty in establishing the best price for building materials, arise when markets do not provide transparent views into the prices offered by different sellers. Buyers and sellers also benefit from transparent views of the purchasing process, so that they can identify problem accounts, difficulties in shipping and logistics operations, high volume purchases, and problems with payables and receivables. Replacing phone and fax transactions with electronic transactions on the Internet creates a step-by-step audit trail for each purchase that can be aggregated into reports, analyzed, and made available to anyone on the network who has the need and the right to access such information.

Making the business case for use of eMarkets always involves more than just the "order matching" component of the transaction. In other words, there must be a lot more to an eMarket than just connecting a buyer with a seller. The reason for this is that markets don't place a very high value on order matching as a market matures. If you doubt this, consider the fact that in 1998 the trading volume on the New York Stock Exchange reached $7.3 trillion and 169 billion shares, but revenues to the exchange were only $101 million.[3] If an eMarket is going to make money, it must look beyond matching buyer and seller to other value added services.

The case for an eMarket is most compelling when the market can improve transparency and efficiency at each step in the value chain from manufacturer to final use. It is easiest to see just what this means if we look at the kind of value chain analysis that most eMarkets do to demonstrate that they create enough new value to both be attractive to participants and to return a profit to the eMarket. What follows is taken from actual eMarket analyses. In order to respect NDA agreements I have created a composite picture that combines the analyses for several eMarkets. Although they are not drawn from any one eMarket, the numbers in the following table provide a realistic picture of potential eMarket performance.

Table 1. Savings available through use of an eMarket to address inefficiencies across the entire process of manufacturing, distributing, buying, and using a product.

	Source of Savings or Added Value	%
Manufacturer
	Inventory Reduction (due to reduced lead time and better information on what is selling)	1.0
	Reduced order processing costs	0.5 - 1.0
	Reduced returns (100% acceptance of orders)	0.2 - 0.5
	Reduced cost of sales	1.0
	Increased sales due to greater market penetration (assume 15%)	2.0 - 4.0
	Total Value to Manufacturer (% of wholesale price)	4.7 - 7.5
Distributor
	Improved Inventory management/faster turns	1.5
	Reduced order processing costs	1.0 - 1.5
	Reduced cost of sales	1.0 - 2.5
	Increased sales (assume 10% - 15% growth)	0.8 - 2.0
	Total Value to Distributor (% of wholesale price)	4.3 - 7.5
Buyer
	Reduced product costs due to pricing transparency	4.0 - 6.0
	Inventory reduction and waste elimination	1.0 - 1.5
	Reduced purchasing and order processing costs	1.5 - 2.0
	Reduced logistics/shipping costs from elimination of emergency orders	0.8 - 1.3
	Improved productivity due to increased equipment uptime	0.5 - 1.0
	Reduced costs due to lifecycle maintenance/replacement	0.2 - 0.5
	Total Value to Buyer (% of wholesale price)	8.0 - 12.3
Total Savings and New Value from eMarket (% of wholesale price)		17.0 - 27.3

There are three features of this analysis that are important to understanding how eMarkets work and why they have been difficult to put into place.

Liquidity

We could have just as well called it "critical mass" or "the tipping point," but "liquidity" has emerged as the term of art to describe an eMarket's transition from just being an interesting alternative purchasing source to achieving standing as the main play, as the ineluctable economic fact of life in a particular industry. Liquidity is the holy grail for eMarkets.

In practice, liquidity means different things at different stages of market development. Investors and eMarket makers graciously lower the bar for early markets but then jack it back up as the market matures. If liquidity is the holy grail, it is all too often only glimpsed at the rear of a train that is headed away over the horizon.

For an eMarket that is just starting out, achieving liquidity is understood to mean getting signed purchasing agreements with enough key buyers so that the eMarket can anticipate a significant potential transaction volume, usually in the range of a billion dollars or more in the early years. eMarkets often try to put such agreements in place even before the market is up and running. If we are willing to mix a couple of metaphors, we can think of this as "vapor liquidity" -- it suggests potential and intent, but money isn't really flowing yet. Vapor liquidity is what today's leading markets can claim, along with a much smaller actual transaction volume. Achieving vapor liquidity is an important first step toward something bigger, since marketplaces tend to become natural monopolies. But is also just vapor. Nobody is yet making money or saving money.

Once an eMarket starts operating, growth toward liquidity proceeds in two dimensions, as shown in figure 1. Growth along the horizontal, X-axis, moving to the right, is growth in the volume of transactions being handled in the eMarket. It is useful to distinguish between transaction volume growth within a single company and growth in the number of companies served. Both kinds of growth contribute to greater overall transaction volume, but they are qualitatively different in terms of the "stickiness" or staying power of the eMarket within the overall market. Only the first kind of volume – lots of transactions within a company – creates strong connection and stickiness between the customer and the eMarket. Obviously, the goal is to multiply this strong transaction volume within companies by many market participants. The question is, "How do you get there?" The safest route is to build strong relationships with each buyer as you grow, rather than to spread the market broadly but thinly at the outset.

How does one build such strong relationships? This question is connected to the vertical, Y-Axis dimension in the graph. As a market moves from the bottom to the top of the graph it succeeds in offering increasing value to the eMarket participants. Framing this in terms of the sources of savings and added value listed in Table 1, moving upward in the graph means that the eMarket is successfully addressing more and more of the sources of savings and value listed in the table.

Achieving liquidity requires moving both to the right and toward the top of the graph: an eMarket most both build transaction volume and deliver a high proportion of the potential savings and value in order to be self sustaining over the long run.

The horizontal, volume building dimension of liquidity is the one that has received the most attention from eMarkets to date. The roadmap for moving from the lower left corner to full liquidity in the upper right corner usually proceeds first to the right, and only later heads upward. There are a number of reasons for this.

Unfortunately for eMarkets, although building volume feels good and is tremendously important, the decision to move first to the right and then upward, focusing first on volume and only later on increased value, is a dangerous one. The problem is that nobody outside of the eMarket itself is very happy with the outcome as the eMarket focuses on transaction volume. Suppliers hate the arrangement because, from their standpoint, the only thing that the eMarket is doing is squeezing margins by eliminating the occasional high margin sales that used to be possible when pricing was less transparent. As one supplier said in responding to an interview that was done in preparation of the Morgan Stanley Dean Witter B2B Internet Report,

Buyer's aren't completely happy either, since they are realizing only a portion of the projected benefits that were the reason for signing on with the eMarket in the first place. In no instance in Fastwater's experience with clients and in our research into eMarkets have we found buyers who say that finding cheaper prices, alone, is a sufficient reason to begin participation in an eMarket. The strongest selling points for buyers always revolve around improved process efficiency and better access to the information and reports required to run the business. When an eMarket focuses only on building up the number of transactions without handling more of the complete transaction, they are not delivering the value that the buyer expects to receive. The problem then compounds itself. Disappointed buyers use the system less. Particularly in markets where buyers and sellers know each other well, it is easy to save even more money on high volume, routine transactions by cutting the eMarket out of the deal. The eMarket's role recedes to one of being the resource of last resort. Buyers turn to the eMarket only for hard to find equipment, unusual requirements, or spot purchasing.

What this means is that unless an eMarket moves upward as well as to the right on the grid in figure 1, even in the early stages of growth, it can quickly run into trouble. The path to liquidity needs to follow something more like a diagonal running from lower left to upper right rather than a jog to the right followed by a sharp left turn. Most eMarkets in operation today are not building enough upward momentum. Changing that requires taking a fresh look at what it means to bring a customer "on board" the eMarket.

On-Boarding: Breaking Through

"On-boarding" is another eMarket term of art. It refers to the process of engaging the participants in an eMarket so that the eMarket achieves liquidity. On-boarding, like liquidity, happens in stages, and the stages match the progressively more complex definitions of liquidity. In the earliest, vapor liquidity stages of a market, on-boarding consists of securing engagement at the executive level of buying and selling organizations in order to receive high level commitments to participate in the eMarket. This kind of on-boarding might potentially involve the exchange of equity or warrants in return for conditional commitments to participate. Because the business case for eMarkets is good, at least on paper, numerous eMarkets have been able to succeed at this level of on-boarding.

On-boarding becomes increasingly demanding as the eMarket begins operation and its quest for liquidity.

On-Boarding at the Outset of eMarket Growth

In the early stages of an eMarket's operation on-boarding tends to focus on suppliers and on getting their products listed on the eMarket site. The reason for this is simple: if you open a store with empty shelves, nobody can do any business. The initial software investments and development focus for typical eMarkets reflect this focus on getting the store up and running. eMarkets start out by buying transaction engines, catalog presentation and search tools, and software and services to make supplier catalog information available through the eMarket database.

This last item, making supplier catalogs available through the eMarket, can be an enormously complex and expensive undertaking that involves solving different conversion and translation problems for each supplier. The problems that eMarkets typically encounter include:

The complexity of these problems and the expense involved in solving them, again and again for each supplier, ensures that catalog loading and access is almost always the primary on-boarding focus during the first months of operating an eMarket. The eMarket needs buyers, too, of course. But the demands of catalog loading typically result in giving only minimal attention to on-boarding for buyers, which often consists of little more than a registration process. In some cases this registration might involve establishing financial qualifications for buyers so that there is assurance that they can and will pay for what they buy. In some specialized markets, such as those dealing with the purchase of chemicals, there may be additional registration requirements in order to establish that the buyers are using the products for recognized, permissible business purposes rather than, say, building bombs or other instruments of terror. But none of this really addresses the concerns that brought buyers into the market in the first place.

That is a mistake. Success requires working with both dimensions of liquidity. That means that on-boarding must extend beyond catalog creation and buyer registration to deal with the value propositions that attracted the buyers to the eMarket.

The failure to fully engage buyers is particularly damaging because, in a sense, a young eMarket needs buyers more than it needs suppliers. Suppliers are much less likely than buyers to see value in an eMarket. A quick look at Table 1 explains why. Add to that the fact that suppliers would be just as happy -- and often happier -- in a market without pricing transparency. It is ironic, then, that the early on-boarding focus in most eMarkets is necessarily centered on connecting the systems used by suppliers. Catalog building is clearly essential, but it is not sufficient.

Now we are getting close to the heart of the problem that eMarkets are encountering. Most early eMarket efforts have gotten to the point where there is a website to do business and catalog information from the suppliers. And now the market has hit a dead spot and has lost momentum. Having focused attention on connecting to suppliers at the cost of not dealing fully with the on-boarding needs of buyers, an eMarket is vulnerable because it risks losing the attention and support of the core constituency that it must depend on to begin growing again. The impulse -- a desperate one -- is to focus on building order volume -- to focus on the low hanging fruit. Unfortunately this impulse is fundamentally wrong and self defeating. At this point the eMarket must turn its attention to meeting the real needs of buyers, needs that go beyond order matching. To do this, the eMarket must focus on a much broader implementation of on-boarding. Few markets are doing that. It's hard. Let's look at what is involved.

On-Boarding to Full Liquidity

eMarkets that hope to move up the Y-axis in figure 1, providing more of the full benefits of the eMarket to participants, must expand on-boarding to include a variety of new capabilities. Here are a few of the most important ones:

Customization of business rules, workflows, and processes for each buyer

Forcing buyers to make their own adaptations to a one size fits all purchasing system runs counter to the reason that buyers signed on to the eMarket in the first place. Such an approach increases the buyer's costs, rather than decreasing them. In business purchasing environments different buyers will have different needs, requirements, and internal practices. eMarkets must build a central system that can be quickly and inexpensively tailored to meet individual buyer needs.

This is one of those problems that is especially vexing for eMarkets, with their many-to-many relationships. If you were simply setting up a system for use by on supplier in selling to many buyers, you would need to provide a single customized relationship with each buyer. That would be hard enough. But in an eMarket, each buyer will need to support different styles and forms of relationship with a variety of different suppliers. The different approaches within just one buying organization will vary depending on the nature of the purchase (e.g., perishable goods as opposed to staple goods, trusted suppliers and new suppliers, large products with complex shipping requirements as opposed to ones that can be express mailed, and so on). All of this variation is then multiplied again, for each registered buyer. The rules engines and customer profile information for eMarkets must be designed to handle these complex, many-to-many relations -- not a simple task.

Bringing existing long-term buying arrangements into the eMarket

Much of the spending in many industries is under long term contracts. Other markets that we have examined closely are moving in that direction. This can pose a serious problem for eMarkets in that it can keep them from reaching liquidity anytime soon; they are shut out of the business before they even knock on the door.

But we have also found that both buyers and suppliers often have difficulty in managing these supply contracts. The same supplier may have very different arrangements with different buyers for the same goods, and of course buyers have different arrangements with different suppliers. As in the rest of their purchasing operation, both buyers and suppliers often rely on manual systems to ensure that goods are being exchanged for the contracted price. We found numerous suppliers and buyers who were quite sure that the manual systems were not really working well, and that either too much or too little was being charged. The arrangements were too complex and varied to be handled just with spreadsheets and price books.

Solving this management problem is a very important opportunity -- and on-boarding objective -- for eMarkets. There is probably no other single mechanism that can so quickly move an eMarket into the center of a buyer's business. But, once again, providing this service means buying or developing a system that can deal effectively and accurately with the multitude of different contract relations between multiple buyers and multiple sellers. It is another many-to-many problem, with a vengeance.

Integration with existing enterprise systems

This is never easy, but is always a requirement if an eMarket expects to tie in to the rest of the accounting and maintenance management operations in a buyer's business. It is, of course, made more difficult by the fact that different buyers have different environments. It is also a more complicated problem in eMarkets, which often serve very specialized vertical markets, because vertical markets tend to use at least a few unique and distinctly non-standard software packages.

For example, in assisting our client in building an eMarket for equipment purchasing for ship management companies, we discovered that a large number of ship managers and operators used a maintenance management system called "AMOS" that keeps track of the makeup and maintenance of the systems on a ship. Buyers were adamant that any eMarket system must tie into their AMOS software. Unfortunately AMOS is a proprietary system with proprietary interfaces. Unlike integration with PeopleSoft, Oracle, SAP, and the other systems used by the buyers in this market, there is no easy way to turn AMOS integration over to consultants. It is an expensive additional development task, but also a critically important one if this eMarket ever hopes to bring buyers fully on board.

Reporting

The ability to get a better understanding of overall operations by taking advantage of the aggregate view of purchasing that an eMarket provides has been a top priority for every buyer in every market that we have worked in or studied. In talking with buyers and suppliers about what they want in the reporting systems that provide this aggregate view, we have found that different market participants have a wide variety of different requirements and are interested in very different things. On-boarding necessarily involves being able to provide each buyer and seller with the reports that it needs in the form that it prefers. This means that, once again, the eMarket must solve a substantial profiling and customer information management problem in order to get market participants on board.

User training and support

In many cases eMarkets are replacing manual systems. That means that the purchasing agents, engineers, and others involved in the purchasing process are often not routine, seasoned users of computers, networks, and the Internet. In each of the eMarkets we have studied and participated in we have found that training and supporting these people is a critical requirement for obtaining the volume of use that the eMarket desires and anticipates.

In too many cases we have seen eMarkets handle user training and support in much the same way that a software company would. In other words, they set up a help desk and assume that, after a customer struggles for awhile, he or she will call the help desk. Although it works for software, where the buyer is committed to use of the new system, it is a bad model for eMarkets. It is much more likely, when the buyer picks up the phone, he or she will simply call the supplier, place the order, and forget about the eMarket and its help desk. The buyers have a job to do, and have alternate ways to do it. If the eMarket is hard to use or confusing, buyers will run around it.

What this means is that on-boarding must extend to building a relationship with each individual purchasing agent or other authorized user. It is essential that the eMarket is able to maintain a profile for each user so that, when he or she does need help, the response can be appropriate for that user's level of computer sophistication, experience with the system, and for the kind of work that he or she needs to do.

On-Boarding in a Nutshell

Business-to-business selling is a relationship business. If my job is ordering lumber, I expect my suppliers to know how I want the load stacked and whether I want it "tarped" (wrapped with a tarpaulin) for delivery. I don't want to have to explain what I need, and I don't want to have to key it in for every order. I do business with a supplier because the supplier knows how to do business with me. If I find that someone else is selling dimensional lumber for less, I don't automatically jump to the new supplier, but I call up my regular supplier -- the one who knows my business -- and ask whether he or she can match the other supplier's price.

I would expect no less of an eMarket. The eMarket, too, needs to understand my business so that I can do deals quickly and receive orders that are always right. From the eMarket's standpoint, this means being able to keep detailed information about each buyer, categorized by products, suppliers, and other requirements. Building this kind of relationship tracking system is necessary just to get in the game. If the eMarket can move beyond that and also do a better job of managing long term contracts, integrating with other systems, and providing useful reports, then the eMarket can become the purchasing resource of choice. At that point, and only then, can the eMarket attain liquidity.

If connecting with the unique needs of each buyer is a requirement for eMarkets if they hope to take their place among the rich relationships that characterize B2B purchasing, it is also an opportunity. Rich customization allows an eMarket to develop a powerful kind of "stickiness" that extends far beyond the strategies available to consumer businesses. By working with buyers to ensure that the eMarket fits tightly with the buyer's other systems, and by storing, collecting, and reporting on information that the buyer needs, the eMarket can create a stickiness that motivates buyers to drive the majority of their purchasing through the eMarket. Note that in following such a course the eMarket addresses its need to build transaction volume by first solving the problem of delivering full potential value -- just the opposite of the pattern typically followed by eMarkets to date. Also note that "customization" does not imply loss or dilution of the eMarket's brand identity. The strategy of choice is to deliver highly customized service while at the same time retaining a consistent brand identity. The movement toward liquidity then accelerates as the buyer associates the satisfaction of its unique needs with the eMarket's brand.[5]

Unfortunately, in most cases eMarkets seem to have lost track of the sticky, relationship-centered nature of B2B buying and selling. The savings and added value from addressing the entire purchase process is usually in the eMarket's plan, but too often the eMarket has spent the last six months focusing only on order matching with the hope that it can build some volume. This can't work. There is no good reason for the buyer to play, and if the buyer won't play the seller most certainly won't.

The overemphasis on order matching is due, no doubt, to the very understandable desire to build revenues and the mistaken believe that there is a simple, direct path to doing that. But another reason that eMarkets might be failing to address the broader dimensions of on boarding, beyond simple registration, is that the building the complete on-boarding infrastructure is difficult. It requires different software products than the transaction engines, catalog search tools, and catalog loaders that have been the focus of eMarket makers to date. The infrastructure investment made to date is important, to be sure, but a successful eMarket also needs software that can track and act on the complex, many-to-many relationships that are at the heart of making a market work.

Relationship Management Redefined

In our interviews with eMarkets it is clear that, for many markets, management thinks of relationship management as "help desk." This is most often a consequence of the market's narrow focus on order matching: do the deals and, if someone needs help, give them hand. Put bluntly, the eMarket wants to do deals and if doing a deal requires talking to someone, then, sure, let's have a relationship. The deal comes first, and maybe the relationship follows. We have seen that this cannot work. The sequence needs to be turned on its head: build the relationship and the deals will follow.

For many eMarkets this will mean resetting priorities and focus. It will also usually mean adding to the eMarket's software infrastructure so that it is really possible to store and manage all of this information about buyer and supplier preferences and relationships. If I expect my loads to be tarped, the only measure of success is seeing the tarp around the load as it pulls into the yard. If the truck pulls in without a tarp, knowing that the eMarket tried to manage the information but lost it or confused it doesn't make the situation better. My next order would run around the eMarket and go to the supplier who knows what I need, even if it means sending a fax. Having systems that allow an eMarket to get this right is critical to the eMarket's survival.

The key elements of the relationship management software that is required go beyond what most CRM software offers today. Part of the reason for this is that the eMarket must be able to make unusually broad, proactive use of the relationship information. Another factor is the complex, many-to-many nature of eMarket relationships. The high level requirements include:

The central role of relationships in B2B purchasing requires eMarkets to fundamentally rethink and extend the notion of what a relationship management system does.

Summary: Making eMarkets Work

The idea behind eMarkets is right: current paper, fax, and phone based ordering systems are expensive and inefficient, and will almost certainly be replaced by Internet based systems. eMarkets can provide transparency into pricing and processing that can squeeze costs out of the purchasing process and, just as important, help markets expand. eMarkets are an important way to make purchasing more productive and efficient and will eventually become a routine, important part of the business landscape.

But the problem today is that most eMarkets are not growing as planned and are not getting the level of use that investors, market builders, and even buyers and sellers anticipated. The question addressed in this analysis was why this should be so. Our focus has been on understanding what has gone wrong and on discovering, if possible, a way to fix the problem.

The key insight that opens a solution to this problem is the realization that achieving liquidity -- the point at which an eMarket can stop fighting for existence because it has become a central part of the market landscape -- depends both on attaining significant transaction volume and on extracting the full potential value from the eMarket's existence. This last objective, extracting all of the potential value, implies that the eMarket must move beyond mere order matching. The market must succeed in addressing problems across the entire value chain.

The current eMarket stagnation grows from the eMarket's initial requirement to have goods on the shelves before opening the doors. What this has meant, in practice, is that most eMarkets have spend substantial sums just creating the catalog and search capabilities required to start the eMarket. In particular, eMarkets have had to spend substantial sums to convert and load supplier catalogs.

Spending so much money on connecting to suppliers was clearly necessary, but it has also been unfortunate in important ways, since it is the buyers, not the suppliers, who are the eMarket's strongest potential allies and supporters. But most eMarkets have spent very little time on the infrastructure to support close relationships with buyers. They are "infrastructured out" at this point, and would just like to do some order matching and build some transaction volume in order to realize some revenues. Unfortunately, without the mechanisms in place to build and use buyer relationships, it is unlikely that the desired transaction volume is attainable. Buyers didn't sign up to do order matching, and that fact that order matching is all that most eMarkets have to offer only means that buyers will continue to do purchasing using alternate routes that run around the eMarket.

Before eMarkets can begin growing again, they need to address the problem of building the connections to buyers that will allow buyers to get the same level of service from the eMarket that they currently get by dealing directly with suppliers with whom they have been doing business for a long time, hundreds of years in some cases. That means recording all of the buyer preference and requirement information so that the eMarket's automated system can do as good a job as the human systems that are currently in place. Once the eMarket can do that, it can play the game. It can then move forward and start displacing current ways of doing business by focusing on the added value derived from the eMarket's ability to reduce processing costs, to integrate more completely with other automated systems, and to produce valuable reports.

Reestablishing the forward momentum for eMarkets is a solvable problem. The fundamental value is there. What is required is a refocusing of eMarket attention and investment from building transaction volume to building more complete relationships with buyers. If the relationships are there, and are encoded and realized in every automated transaction between the buyer, the eMarket, and the seller, the transaction volume will follow.

Notes

1. Christina Cheddar, "Pressure Builds For Laggards As More Chem Companies Go Online," Dow Jones Online News, Monday, October 09, 2000.

2. If you are interested in a much more in-depth look at B2B markets, I recommend reading Morgan Stanley Dean Witter's report titled The B2B Internet Report: Collaborative Commerce. by Charles Phillips and Mary Meeker, published in April of 2000. As of this writing it is available on the MSDW website at http://www.msdw.com/techresearch/index.html.

5. For a much more complete discussion of B2B personalization and customization see Web Engagement: Connecting to Customers in e-Business (Addison Wesley, 2000).