e-Marketplace Arms Merchants: Enabling the B2B Enablers.

Web-based business-to-business exchanges hold the promise of rewiring the world economy, but as usual, the promise is well ahead of the reality. Sorting the hype from the benefits, three B2B providers size up the implications of Net marketplaces for CEOs.

Last year, e-procurement was

Still in its infancy, B2B e-commerce is estimated by the Boston Consulting Group to account for $2.8 trillion in sales by 2003. GartnerGroup pegs the figure even higher at $7.2 trillion, and predicts 7,500 to 10,000 B2B marketplaces to emerge by 2002. Robertson Stephens estimates that even a 10 percent adoption rate by 2003 implies $1.8 trillion in B2B commerce volume. The global market could almost double that of the U.S.--exceeding $3 trillion worldwide. Somewhat more circumspect, Forrester Research forecasts that B2B transactions will grow to $2.7 trillion by 2004 from $406 million in 2000, with 53 percent of that passing through e-marketplaces compared to about 1 percent now.

Some 1,200 B2B Internet exchanges have blossomed since last year, with many claiming to have the killer app to bring buyers and sellers together, either in horizontal or vertical markets. The severity of last Spring's Nasdaq sell-off in tech stocks proved to be a corrective for a sector that was engulfed in its own mania. Some of these companies were little more than a T1 line and brochure-ware. Others had traction issues. Plastics-Net.com, with $1 million in revenues was ranked as the leading independent exchange in the plastics industry, but GE Plastics, a $7 billion market leader is not a part of it. Nor is Wal-Mart, the world's largest retailer to be found on the Worldwide Retail Exchange, which offers itself as the industry's largest B2B retail marketplace.

According to A.T. Kearney, less than 15 percent of Internet exchanges are really delivering value-added services or fulfilling end-to-end transactions. At issue is how to best attract buyers and sellers. Neutrality is prized, but will the middleman's cut survive as volumes increase? Will smaller suppliers get crushed in marketplaces set up by giant buyers such as big auto's Covisint? According to Deloitte Consulting's Steve Sprinkle, only a handful of these exchanges will get sufficient traction in each sector to thrive.

Initially, the promise of e-procurement--where companies had hoped to trim up to 20 percent of their total spending and drop transactions costs by two-thirds--is hard to resist. However, after studying both vendors and users, Forrester Research analyst Laurie Orlov found that implementation fell short of reality. Many simply automated their paper-based processes. "To be truly effective the process needs to be handsfree," she says. "Every supplier you need to buy from needs to be electronically accessible. And much depends on eliminating 'rogue' purchasing outside the system. These are very big cultural issues in some companies."

In most exchanges where buyers and sellers trade bids and offers until they make a deal, commodities or standardized products are the easiest to transact. GM's recent $147 million purchase of sealing packages from Opal through TradeXchange "is part of a growing trend of buyers pushing online procurement boundaries into direct materials used to produce finished goods," says Forrester analyst Stacie McCullough.

If e-marketplaces are to succeed, costs have to be removed from the entire supply chain, not simply pushed down on hapless vendors. Suppliers are naturally reluctant to join trading communities because they are concerned about pricing and information transparency.

Further, second and third tier suppliers typically have very old computer systems that are difficult to integrate. Product information is highly fragmented and often exists in forms that not even they can readily understand. Many underestimate the difficulty in managing these supplier "onramps" to an exchange. As these marketplaces evolve, 20 to 30 percent of safety stock can be eliminated from industry inventory right away, according to GE Global eXchange Services CEO Harvey Seegers. Given that $2 trillion is tied up in inventory, that amounts to $400-to-$600 billion coming out of the supply chain. Those are some sizeable savings--and the kind most CEOs want a piece of. Toward that end, CE asked a few of the leading exchange enablers to size-up both opportunities and pitfalls.

CommerceOne's Mark Hoffman Reaches for Global Scale

The Covisint deal was a big win for CommerceOne, a Pleasanton, CA-based software and e-commerce service provider. The company--which already had a deal with General Motors to handle all of the car giant's online procurement from janitorial services to car parts--got a boost when Ford, and later DaimlerChrysler and Renault/Nissan, combined forces to create a single marketplace. (Oracle, which had sold Ford on the idea of its own exchange, is part of the expanded arrangement.)

GM estimates its aggregate spend at around $500 billion, with the whole market coming close to $1 trillion. CommerceOne charges, on average, one dollar per transaction regardless of size on its principal exchange mechanism, MarketSite. The company earns revenue from software licenses, upgrades, and services, which tend to be 20 percent of the initial price of the software. Tucker Anthony Cleary Gull analyst Richard Davis's revenue estimates for CommerceOne are $130 million in 2000, $270 million in 2001, and $435 million in 2002, with the Covisint deal boosting these numbers dramatically. As with its arch-rival Ariba, CommerceOne is expected to report diminishing losses until becoming profitable by 4Q next year.

A West Point engineering grad, CEO Mark Hoffman, prior to joining CommerceOne, co-founded Sybase, a relational database management systems startup, where he was also CEO in 1984. Charged with driving the overall direction of CommerceOne, his biggest challenge will be to get his company's technologies and those of Oracle to work together seamlessly. (There was talk of DaimlerChrysler planning to bring its own technology partner into the venture but this seems to have been tabled.)

The company sells a buying application called BuySite that allows e-procurement on the Internet and offers MarketSite as its trading platform, which is the foundation for interactive, real-time e-marketplaces. Last June, the company made a deal with SAP, the German software vendor, to integrate that company's supply chain management and product life cycle applications into CommerceOne's online exchange platforms. These advanced planning and optimization apps are expected to be rolled out for such customers as Boeing and GM by late third quarter of 2000.

Of its 197 customers, British Telecom stands out as a global Market-Site partner. Since the implementation of its exchange site, BT has reduced procurement administration costs by 90 percent and is seeing a 10 to 15 percent cost reduction in indirect goods purchased. BT has also reduced fulfillment from eight weeks to seven days and is seeing a $5 million ROI in the first 10 months. In addition to its deal with GM, CommerceOne is developing mega exchanges in aerospace with Boeing, Lockheed Martin, BAe Systems, and Raytheon. An energy exchange with Shell, BP Amoco, and 13 other energy companies is under way. A utilities exchange with 21 partners, including Bell South, is in progress.

What's driving Internet exchanges and e-marketplaces?

There are several forces. One is large companies coming together in their industries to create vertical marketplaces to streamline procurement. In other cases where there may be fewer dominant players, exchanges may be run by independents, but the principle of a many-to-many marketplace remains the same. Most companies have successfully automated their internal processes, so now they want to automate inter-company buy and sell transactions. For example, in the GM exchange, what they want us to do is to pick up a CAD-CAM file from GM, send it to suppliers who look at it, offer feedback, or make a bid or quote on whatever they can do.

There are also implications beyond procurement. Let's say you're an aerospace manufacturer, and you have a deep and complex supply chain of many tiers. This is an opportunity for communications. Behind your firewall you have planning and optimization software that enables you to apprise inventory and materials on hand. You may also have demand forecasting capability. After checking available stock, you realize that you need more materials, components, or assemblies. You send a demand signal to your tier one suppliers that indicates your need for X number of sub-assemblies by a given date. The tier one supplier receives the signal and realizes it's going to need more components to fulfill the order. The tier two supplier goes through a similar process to fulfill its order, so the signal goes all the way down the supply chain. Today, this process takes place between individual companies; tomorrow, it'll all happen on exchanges.

Many companies already use EDI, which is said to be more secure than the Internet and more reliable. So what's the big deal?

I don't agree with that at all. If you're worried about security, you can create dedicated bandwidth for private networks to your operations. You can put routers on both ends, which allow you to encrypt and decrypt. Yes, it'll cost more than just the Internet, but EDI is much more expensive. We don't have a single customer who has gone to the extent of using a virtual private network once they understand how the system works. Let me assure you that GM looks at it the same way because its exchange can't afford to shut down for any reason, either.

Which e-marketplaces are in the vanguard?

The aviation and aerospace exchange with Boeing, Lockheed Martin, Raytheon, and BAe is one. The chemical and energy industry exchanges are also pretty demanding. Early on, they have been able to drive out some of the inconsistencies in their spend. GM, for example, found that one of its suppliers had 24 different contracts with GM at different sites with different dollar values for exactly the same stuff. GM was able to use the exchange to drive one price for the item. Exchanges enforce the contract model, which allows different pricing models for different reasons. If you're GM, you "see" GM prices. If you're Chrysler, you only see Chrysler prices. So there's no overlap.

Exchanges are also set up to do invoices and payments. Let's say you are a supplier, and you want to factor what you just received or sold. Systems can factor automatically. [You] simply indicate, "Yes, I'm going to sell to this company, here's a confirmation order, check here if you want to factor and boom-- you've factored the product and you've got the cash. This is good for small suppliers who are often cash poor.

What's the next wave in e-marketplaces?

When exchanges interact with other exchanges, you will see enormous changes. We have something we call the "global trading web," It takes different exchanges and ties them together. A buyer in, say, the North American auto exchange can communicate with a supplier plugged into the Deutsche Telekom exchange, with the P.O. in dollars and the supplier dealing in euros. The system will do all of the translations and transactions. We have established strategic relationships with British Telecom to host a MarketSite in the U.K., with Banamex to host one in Latin America, with NTT to host one in Japan, with Singapore Telecom to host one in Asia, with DT to do one in Germany and Swisscom to do one in Switzerland. Using the global trading portal, buyers from any exchange can access suppliers around the world, pay for it, and arrange shipping.

Ariba's Keith Krach

Ramps Up the Network

With almost 1,550 employees and an impressive 40 percent market share, Ariba has zoomed to a leading position in B2B e-commerce applications since its IPO one year ago. The three-and-a-half-year old Mountain View, CA startup, with annual revenues running at $320 million, offers end-to-end infrastructure for procurement and e-marketplace applications supported by the Ariba Network, which enables buyers and sellers to automate business transactions on the Internet. The company's Ariba Buyer (formerly ORMS) system enables companies to automate procurement within private exchanges. About $400 billion of total spend flows through the company's exchanges.

Current customers include Chevron, Cisco, FDX, HP, Merck, Nestle, Philips, Motorola, and Visa, among others. The company has signed up more than 20,000 suppliers to join Ariba Supplier Link (ASL) to provide electronic catalogs and conduct transactions. The company's "punch-out" technology is a key part of its procurement added value. Instead of displaying a static version of a supplier's catalog, Ariba's technology lets buyers link directly to a supplier's Web site via cXML to take advantage of the supplier's unique functionality while still remaining in the Ariba Network.

Last December, the company acquired Tradex, which specializes in platform software for horizontal and vertical marketplaces. More recently, it acquired SupplierMarket.com whose collaborative sourcing technology enables buyers and sellers to negotiate prices and logistics, something Ariba badly needed to fulfill its position as an end-to-end solution provider. Textron, which recently signed on to EDS's CoNext, a leveraged sourcing network powered by Ariba, aims to drive about $1.5 billion of indirect goods and services annually through the network. The manufacturing and electronics conglomerate hopes to reduce costs by $150 million initially and another $100 million over three years.

One of the firm's co-founders, Keith Krach, had been COO of Rasna Corp., a mechanical CAD software design firm, which he also helped to start. The 42-year-old Purdue- and Harvard-educated Ohio native (Rocky River near Cleveland) worked as a third shift production foreman on the Cadillac chassis line in his first job out of school. At the tender age of 26, he became VP and general manager of GMF Robotics, a joint venture with GM and Fanuc. Deciding that autos were not his future, he left GM at 30 and set Out for Silicon Valley, where he hooked up with a few scientists from IBM's research labs to form Rasna. When Parametric Technologies acquired Rasna, Krach became "entrepreneur-in-residence" at Benchmark Capital for six months before he and six others, then at Benchmark and Crosspoint Venture partners, were bitten by the B2B bug. Initially called "ProcureSoft" ("We changed that real quick"), Ariba, says Krach, is destined to be "the Cisco of Net exchanges." With Ariba shares having risen 1,882 percent since i ts July 1999 IPO rising (adjusted for stock splits), Wall Street, thus far, is inclined to agree.

With less than 15 percent of Internet exchanges actually delivering value added services on an end-to-end basis, is there more wheeling than dealing at present? When will business really embrace this?

It's just the beginning. Companies are moving from classic e-procurement to dynamic marketplaces. The press release wars are over. It takes awhile for multi-buyer and seller marketplaces to get liquid. When companies approach us asking how they should get started, I talk about three key things: the technological infrastructure, or the platform and the network; the overall business model--should it be a business unit or a newco, and if it's a newco, how do you divide the equity among partners; and the culture of the organization, because you have to throw out the Fortune 500 policies and procedures. One of the most critical decisions facing these exchanges is the choice of CEO. Many haven't picked one or they have multiple ones. Covisint hasn't even picked its CEO yet. It's not just building the technology; it's building a new organization.

There are three reasons why companies typically try to achieve value through marketplaces. One is to cut costs, which is where much of the focus is today. The second is to increase revenue, and the third is to retain customers. Take BankAmerica. They utilize our platform for e-procurement, reducing indirect purchasing costs anywhere from 5 percent to 20 percent. They're also developing a marketplace for their two million small- and medium-size customers, which will create revenue opportunities and a delivery mechanism for new services. They've increased the switching cost for those customers which help retain customers.

Do companies feel vulnerable when asked to share sensitive inventory or price information in a vertical exchange?

Each organization is a little different. What's proprietary and what you're willing to share for the good of the consortium varies. Petrocosm, an oil and gas marketplace founded with Ariba by Chevron, Texaco, and KPMG, shares a lot of information in terms of buying equipment. Mindtrac, in Singapore, is a rubber exchange that acts as a tire marketplace for wholesalers and manufacturers around the world. It shares a lot of information and has a tremendous amount of liquidity.

Are you at all concerned that the Department of Justice is looking into e-marketplaces as possible venues of monopoly power or cartel-like price-fixing?

When I testified before Congress recently, I emphasized that the consumer should benefit from these exchanges because prices will come down. I also recommended that third-party management of these exchanges is vital to achieve neutrality. As far as monopoly power is concerned, much depends on the corporate governance and deciding what's truly appropriate. Everything depends upon trust. If it's not a 60/60 win-win for buyers and sellers, it won't survive very long anyway.

Is EDI dead?

I don't know, If it's not dead, it's in the process of being replaced by open XML networks. About one-third of our customers hook up to our network via EDI, because you want to handle every single customer regardless of their sophistication. For some, it's high level XML, for others it's EDI, and for others it's a simple Internet connection. I don't see EDI growing by leaps and bounds.

GE Global eXchange Services' Harvey Seegers

Collaborates to Compete

"E-business was made for GE, and the 'E' in GE has a whole new meaning," wrote Jack Welch in the company's current annual report. If there were any further doubts about GE's commitment, last March he announced that GE would make a sizeable investment in two new Internet service companies that were formed from the split of GE Information Services: GE Global eXchange Services (GEXS) and GE InterBusiness Operations.

The latter was formed to create the "Internet factory of the future" by migrating GE's installed base to open systems and offering Six Sigma service. With 500 employees based in Gaithersburg, MD, GEXS intends to focus on Internet data exchange enterprise application software, procurement services, and e-marketplace exchanges, leveraging one of the largest electronic trading communities with 100,000 partners in the world.

Welch's choice of unit leader is interesting. A former infantry commander in the South Pacific and aide to the Commandant of the Marine Corps, Harvey Seegers exhibits a crisp take-that-hill confidence in staking GE's flag in B2B territory. GE may not have the same reach as Ariba or CommerceOne, says AMR Research analyst Scott Latham, but it has the potential to challenge their leadership.

Claiming to be technology agnostic, Seegers affirms that GEXS will use whatever software a customer wants, including those of market leaders Ariba and CommerceOne. But Seegers recently formed an alliance with CommerceOne to offset the formidable linkup of Ariba with IBM and i2 Technologies. The alliance does not preclude GEXS from working with others, but it underscores the furious pace to get total end-to-end transaction solutions to market. Industry observers say true B2B won't come until trading is connected with disparate back office functions in a complete "hands-free" environment, which is B2B's holy grail.

GEXS helped Coca-Cola establish a private exchange with its bottlers and suppliers, as well as British food retailer Tesco, which set up Trading Information Exchange (TIE) to share daily point-of-sale data and procurement transactions. In the Q3 2000, GEXS will launch a healthcare private exchange currently backed by Johnson & Johnson, GE Medical Systems, Baxter International, Medtronic, and Abbott Labs. It will allow online ordering, customer-directed distribution and value-added services, such as clinical information on a broad range of products. Ambitious in scope, the exchange will also be supported by the Ariba, IBM, and i2 partnership.

Without disclosing exact figures, Seegers says his unit runs between $250 million and $500 million, with projected annual growth of 30 to 40 percent. He plans to top $1 billion in revenue in five years.

What does GE bring to the e-marketplace party that isn't already available from other arms merchants?

A lot of the vertical markets do not yet have the ability to completely automate a transaction. There's a fair amount of what I call "swivel chair commerce" going on, where someone will look at something on a Web site, make a decision to order it, and then turn to the telephone to place the order where it gets keyed into another system. There is a good deal of useful information being presented in these marketplaces, but the real mission-critical stuff isn't there yet--things like the availability to order, logistics tracking, invoice status, and other things that bring economies of scale to e-commerce. Our business model is to provide companies with an integrated suite of point-to-point solutions.

Is GE eating the GEXS dogfood? Are you doing this for GE'S operating units?

Clearly our mission is to become a leading top business e-commerce provider to companies around the world. To the extent we can assist GE with its e-commerce implementation efforts, we do and we're involved in a number of projects inside GE. But the responsibility for transforming GE lies, along with the chairman, with operating heads in each division. We compete for GE's business just like anyone else. There's no free lunch.

So what does GEXS offer that, say, Ariba or CommerceOne doesn't?

We provide the ability to complete transactions between computer systems of buyers and sellers that would not normally talk to one another. Our help centers around the world provide seven-day, 24-hour user support in local languages. We have a host of applications outside the CommerceOne or Ariba models, such as content management services. We have a 50 percent ownership in TPN Register, which is a supplier to Ariba. So an Ariba customer can come online, and the Ariba system links into TPN Register, which hosts all product data for suppliers on a given service. The customer can compare blue widgets across 10 different suppliers, get current pricing information, and make a meaningful choice as to which widget he wants at what price.

Is the Covisint exchange involving GM, Ford, Daimler-Chrysler, and Nissan going to be the e-marketplace model for others or is it unique to big auto?

It's too early to tell. My view is that that exchange got announced a little faster than the players fully understood how it was going to go forward. There are huge opportunities to improve the purchasing efficiency of indirect spending, because the direct spend is in pretty good shape, largely through EDI. The big question mark is the government's inquiry into that exchange. How the DoJ comes out on this may determine whether it becomes an emerging model. The key here is that industry players have come together to form a marketplace as opposed to independent third parties getting between suppliers and their customers. This latter model doesn't have a long runway. Businesses that allow third parties to get in between them and their customers have a problem.

The other issue is one antitrust officials call monopsony power, where buyer power is aggregated to such an extent that it creates a concentration of power that otherwise would not exist. This social policy theory holds that if buyers unnaturally come together to force down supplier prices, you will have an unnatural consolidation of supply which results, long-term, in increasing prices to buyers that are then passed on to customers.

What's the model for the future?

I see companies developing private exchanges where they have special relationships with suppliers and can interface with them in various ways to offer special benefits. Customers may come into a company's Web site and access that company's suppliers on a preferred customer basis. The greater the preference, the greater the access. It's not a free-for-all auction. Companies with either very strong buyer or supplier power will establish this private exchange and then extend invitations to customers and suppliers, not competitors.

When most of the mission-critical functions are incorporated onto these exchanges, is it bye-bye EDI?

Everybody's got his own risk curve. Today, our EDI business is growing 20 percent to 30 percent annually. So when the popular press paints EDI as an old technology that's dying, nothing could be further from the truth. We're the No. 1 EDI services provider in the world and it's because those customers that need to make certain these transactions go through or shut down production if they don't, aren't ready to trust the Internet. Will all direct materials eventually go to the Internet? The answer is yes, but not as rapidly as some people would have you believe. There needs to be a little more bandwidth, security, and reliability. Every time one of these monster viruses is released on the Internet, people's laptops get attacked and they get the jitters.

What do CEOs misunderstand most about e-commerce or e-procurement?

They misunderstand how important it is to achieve back-end integration between computer systems and the integration of these back-end systems with the back-end systems of their trading partners. There is a general belief that all one has to do is get customers and suppliers on the Internet exchanging visual information. Nothing could be further from the truth. Real economies of scale come only when these computer systems are fully integrated so that the amount of required human intervention for a transaction to be fulfilled is minimal.

What are the initial useful steps for CEOs to take to achieve this?

They should take a hard look at the architecture of their physical supply chain processes. What are the many sequential steps between a company and its suppliers? Where is all the inventory tied up and why? Only then can CEOs implement e-commerce solutions to streamline the process. The last thing you want to do is to pave over the old cow path.