E-strategy brief

While Welch waited
May 17th 2001
From The Economist print edition

This week we launch a series of case studies of how big established companies are developing their e-business strategies. We begin with a look at how the biggest of them all, General Electric, almost missed the boat

Get article background

THE razzmatazz that once surrounded business-to-business (B2B) e-commerce has vanished, leaving a trail of dead dotcoms. That does not mean an end to the changes that new information and communications technologies are bringing to companies. But from now on, the transformation will occur mainly in established companies, which will continue their quiet search for new electronic ways to carry out familiar tasks.

The coming of the Internet, with its associated technologies, is changing both how companies deal with their suppliers and customers, and how they organise work. Some of these shifts will be profound; others are, for now, mere extensions of the proprietary networks that many big firms were already using. The case studies that we will run over the next few weeks will look at the progress of this revolution in America, Europe, Japan and emerging economies.

One of the companies farthest along the track is General Electric. This old-economy icon now buys and sells more through its private online marketplaces—an estimated $20 billion this year alone—than is traded in all the independent B2B marketplaces together. Management books celebrate the e-business epiphany of its boss, Jack Welch, as proof that even elephants can dance.

Yet, for most of the late 1990s, GE seemed to have blown a spectacular e-business opportunity. When the Internet wave hit, it was one of the world’s biggest industrial buyers, with one of its biggest electronic buying networks. In 1997 its GE Information Services (GEIS) division was practically the only B2B game in town, albeit based on pre-Internet technologies.

But GEIS faded mysteriously from sight just as such B2B companies as Ariba and Commerce One began to produce fashionable software for electronic commerce. Although GEIS executives knew that they stood on the verge of something big, they were unable to do much about it until it was almost too late.

The reason was a bad case of conglomeritis. GE, the world’s largest company, offers vast opportunities to use technology inventively. But its obsession with earnings growth means that it suffers from an inherent conservatism: unprofitable new ventures must win extraordinary executive and board support to go ahead.

In such a culture, it was hard to win support for a new B2B business. Keeping up with the dotcom start-ups would mean spending as profligately as they had, but without the freedom that came, in those golden days, from investors’ tolerance of big losses. “We wanted to do some of what Ariba and Commerce One do, but we couldn’t do it at the cost of hundreds of millions of dollars of losses,” says Harvey Seegers, then GEIS’s boss and now in charge of the main part of GE’s e-commerce activities. GEIS bosses found it hard to get a green light from headquarters. While they waited, their “first-mover” advantage withered.

The main obstacle was the prospect of cannibalising a profitable proprietary data business: GEIS itself. From the early 1980s until the mid-1990s, e-business referred not to the Internet but to electronic data interchange (EDI), a way of exchanging formatted purchase orders and other documents electronically. EDI transactions tend to be carried over private, secure networks or leased lines, usually run by third-party service providers (see chart).

GEIS was the biggest of these. Its data centre in Rockville, Maryland boasted a NASA-like command centre run by former military technicians still sporting crew cuts and smartly addressing supervisors as “sir”. It served 100,000 companies that collectively process more than a billion transactions a year.

This offered GE a vision of the future: more and more companies doing business with each other electronically as they automated their internal processes. The natural extension of the “enterprise resource planning” movement led by SAP, a German software company, was connecting one enterprise to another. This laid the groundwork for widespread, automated machine-to-machine transactions between companies.

Enter the Internet

But at the same time the Internet was exploding. GEIS at first saw its potential mostly as a cheap on-ramp to EDI networks for small and medium-sized firms that did not want a full EDI connection. Smaller suppliers often had to be dragged reluctantly on to larger firms’ EDI networks: for the minnows, the costs often outweighed the savings. Perhaps the Internet could help there, replacing expensive private networks with cheaper public ones. So GEIS offered a service that would accept EDI forms submitted via a website and transmit them on its own proprietary network to their destination, one of GEIS’s larger clients.

But that was as far as it went. For all the Internet’s consumer appeal, GEIS reasoned, the business world is far more conservative. For more than two decades, EDI had been the only commonly agreed standard with which to conduct business-to-business transactions. It would not be overturned easily.

Yet EDI has many flaws. In essence it is nothing more than a machine-readable e-mail message with strict rules about what to say. It is not interactive, and it operates, like all e-mail, in an asynchronous fashion, with long delays between call and response.

Thus, Wal-Mart’s ordering system might use EDI to send a purchase order to a supplier in the morning; once the supplier accepted the order it might direct its computers to send a confirmation back. Similar formatted messages might travel back and forth as the order is completed, shipped and received. EDI saves companies money, mostly by taking people and errors out of the equation, but it does not fundamentally change the way business is done or who it is done with.

That changed when the Internet came along. B2B exchanges did more than simply take existing relationships and transactions and turn them into digital form: they offered the potential for new relationships and new sorts of transactions, from auctions to direct sales without middlemen and brokers. Between 1995 and 2000, more than 700 such exchanges were founded. Some were independent marketplaces and some were run by big buyers or sellers.

Most of those have since disappeared. Yet B2B continues to grow today, although companies now tend to work with their own suppliers in private marketplaces rather than in independent exchanges. Indeed, many of the virtues of GEIS’s model—and especially its focus on improving existing business relationships, rather than reinventing them—are now, ironically, back in vogue.

Fomenting revolution

A year and a half ago, though, that was not the prevailing wisdom. As B2B dotcom start-ups were going public at multi-billion-dollar valuations, GEIS was starting to have doubts about its continuing focus on traditional EDI. Fears of the Y2K computer bug had distracted it for years, as it spent heavily to avoid trouble. Its customers—mostly larger, established industrial firms—were equally distracted.

Mr Seegers had joined GEIS in 1996. Within the company, a mainframe culture dominated, with an emphasis on white-shirted professionalism and reliability; beyond it, the growth of the Internet was in full spate. Mr Seegers, a former marine, was trained to handle conflict; yet this was a case of two worlds that preferred not to acknowledge each other’s existence at all.

GEIS was a small division by GE standards, not even a line item in its annual reports—although its concentration on EDI meant that it certainly satisfied Mr Welch’s decree that all divisions must be number one or two in their market. Mr Seegers now began to agitate for a larger role for GEIS. It took Mr Welch’s e-business awakening to get it.

That moment has, like most big shifts in GE, taken on an unduly heroic sheen. It was, after all, in early 1999, a time when it was impossible to ignore Internet hype; Mr Welch was hardly a brave pioneer. But it is not easy to turn around the world’s biggest firm, even for the most commanding of bosses.

In fact, there had been something of an e-business movement already brewing in GE, from the bottom up and in just a few isolated spots. As early as 1994, GE Plastics, one of the conglomerate’s divisions, had a website with some basic information. By 1997, its Polymerland distribution operations had become the first GE unit to allow customers to place online orders. GEIS was also operating a partly web-based service called the Trading Partner Network, which simplified and automated the bid and procurement process for other divisions, led by GE Lighting.

But until 1999, when Mr Welch, after a Christmas watching his family shop online, famously told GE that e-business would be every division’s “priority one, two, three and four”, progress was slow. Buying from suppliers online was one thing, but selling to customers online risked putting GE’s sales force out of business. Big Internet investments risked hurting each division’s bottom line, against which bosses are mainly measured. And GEIS, the one division that could have pushed the hardest, was fussing over Y2K.

By the time Y2K had passed, Mr Welch had devised “destroyyourbusiness.com”, encouraging each division to reinvent itself before somebody else did. Mr Seegers proposed splitting up GEIS: rather than trying to alter the mainframe culture that ran the EDI networks, he suggested spinning off a more entrepreneurial group to pursue Internet marketplaces. An hour and a half later, Mr Welch said, “Go for it.”

So, in March 2000, GEIS became GE Global eXchange Services (GXS), a software and marketplace builder, and GE Systems Services, an EDI network provider. Tellingly, Mr Seegers chose to run GXS, the smaller of the two by assets, but the one with fewer legacy obligations that would hamper innovation.

The culture clash

Today, all of GE’s big divisions run their own web marketplaces, both for internal and external use. Three initiatives—“e-buy”; “e-sell”; and “e-make”—are digitising the main functions of running the conglomerate, saving money, reaching customers faster and using Internet technology to extend the focus on quality that was the driving-force of Mr Welch’s previous passion, six-sigma manufacturing.

The greatest hurdle has been not technology but culture. Sales staff, worried that they might be destroying their jobs, had to be offered bonuses for helping customers to use GE websites to order. Managers had to watch carefully for reprobate employees using “parallel paths” (the telephone, for instance, or a walk to a store) to order supplies, say, or arrange travel. Some offices even closed their mail rooms for all but one day a week (and that only for the incorrigible legal department) to stop employees from using regular post. Others locked their printer rooms except for occasional days when bosses would station themselves at the door and demand from those who came through an explanation for their sad inability to shake old paper habits.

For now, most of this is about cost savings and improved quality. But over time GE hopes its e-business projects will allow it to come up with new products, such as services that turn one-time purchases into recurring revenue streams, and sell to new customers that its existing salesforce cannot reach, such as small businesses. It also hopes to expand its supplier base by letting more firms bid online, through auctions and other marketplaces.

GXS has restored GE’s reputation in (depleted) B2B circles. But although the company is now free to invest in its own research and development at a rate closer to that of its independent competitors, it still cannot buy other firms, as they can, with shares. GE may now be flexible about internal e-business spending, but earnings-diluting acquisitions are still hard for it to swallow. This constrains its ability to profit from the industry’s consolidation: it must stick mostly to its strategy of organic growth and internal development. GE breaks many moulds, but there are limits to what even the most entrepreneurial arm of a conglomerate can do.

That does not undermine the scale of GXS’s achievement; rather, it suggests how hard it is for big companies to keep up with nimble newcomers. Asked if he can think of any other successful high-tech companies that have grown up within a diversified industrial conglomerate, Mr Seegers ponders. “No,” he finally concedes with a smile. “We’ll be the first.”