Recovery in Technology Spending Is Still Elusive

The New York Times ^ | June 23, 2002 | RUSS MITCHELL

[sarcasm]

very quarter since the technology boom ended, hopeful forecasters have predicted that a recovery in business technology spending is just a couple of quarters away. "Eventually, they're going to be right," joked Kevin M. Landis, who runs Firsthand Funds, which focuses on technology stocks.

It is not likely to happen anytime soon. While other areas are showing tentative signs of strength, technology remains in a deep funk.

The reasons go beyond the weak economy. Perhaps for the first time since the invention of the computer, technology is outpacing customers' needs, industry executives say. The equipment of today, they add, can generally handle most of what software users throw at it, meaning that companies will not replace equipment as often — a trend that they say will outlast the current economic sluggishness. The business software sector, meanwhile, has grown intensely competitive, and vendors are under severe price pressure.

Technology executives have pounded a steady beat of dour prognostication in recent weeks. First, I.B.M.'s chief executive, Samuel J. Palmisano, said, "It's clear the industry is not bouncing back this year," adding that next year did not look much better. Then the chief of Hewlett-Packard, Carleton S. Fiorina, said the "spending environment remains tough around the world" for technology.

On June 11, Kenneth A. Goldman, the chief financial officer of Siebel Systems, said that trying to sell software is "frankly more challenging" now than it was last year. And last year was dismal. Siebel's shares tumbled 14 percent the next day. The company makes software for customer relationship management, a software area that analysts expect to perk up early in any technology comeback.

Also this month, Intel warned that microprocessor sales were weakening, sending chip stocks reeling. And on Tuesday, Advanced Micro Devices said its sales are expected to be well below already low estimates.

Chief executives who have clamped down on technology spending to protect scarce profits lack vision, some technology boosters say. "Technology is one of the main drivers of productivity," said Harris N. Miller, president of the Information Technology Association of America in Arlington, Va. "The challenge is to remind C.E.O.'s and C.F.O.'s of that, so they're more willing to spend money."

But many top executives say they do not need to be reminded. As Nigel W. Morris, the president and chief operating officer at Capital One Financial, puts it: "We're willing to spend a dollar to make more than a dollar back." What is different now is that top executives see no need to keep increasing technology spending at anywhere near the pace of the last several years.

After splurging on multimillion-dollar, multiyear software projects from Oracle, SAP, Siebel and PeopleSoft and others, businesses are buying software in smaller pieces — when they buy at all. They also report a move toward lower prices and better contract terms, which they expect to continue as software companies compete more aggressively.

If technology faces a long, grinding road to recovery, it is not because large corporations doubt technology's power. The impressive productivity growth of the last decade continued through the recession — up 4.2 percent over the last year — and at an annual rate of 8.4 percent in the first quarter of 2002, the highest in 19 years. Figuring into those statistics, of course, are wide layoffs that force remaining workers to pick up the slack. But few economists doubt that technology has played a big role.

"Technology is the fulcrum of our business," said Mr. Morris of Capital One. Considered by industry analysts as perhaps the shrewdest user of information technology in the credit card business, Capital One slices and dices data to offer high-margin products. Its customer base has increased to 45 million worldwide today from 5 million in 1995, while profits have grown 20 to 40 percent annually since 1995.

 
XECUTIVES at Waste Management, the long-troubled garbage hauler, focused on new systems to help revive its fortunes. "Our stock went from 13 to 28 bucks over the last 30 months, and we couldn't have done that without technology," says Thomas L. Smith, senior vice president for information technology.

At Royal Bank of Canada, Martin J. Lippert, vice chairman and chief information officer, acknowledged that his company became caught up in the "silliness" of the late 1990's, expecting, for example, that customers wanted to check bank statements on wireless devices.

Still, Mr. Lippert said, the bank continues to "take a lot of the technologies that grew up in the dot-com boom and put them to practical uses." The company has moved aggressively into online banking and now aims to eliminate $1 billion in costs over several years, in large part through technology. A Web-based system to let customers reorder checks online saves $19 million a year, Mr. Lippert said. Another system allows a customer to make address changes remotely in a single step, saving Royal $38 million in annual mailing costs.

All these executives say they will continue to modernize aggressively. But because they will be spending less on each piece of hardware and software, and because they will extend the lifetime of equipment they already have, technology companies may well find a return to their glory days elusive.

It is almost as if Moore's Law has turned around to bite the industry that created it. Moore's Law says processing power doubles about every 18 months, or, conversely, that the cost of a single unit of computing power is cut in half.

Equipment capabilities are advancing faster than demand for extra power — a switch from the days when a new computer seemed outmoded the minute it was bought. "The PC's coming out today have 2-gigahertz processors, 512K memory, 60-gigabyte hard drives," Mr. Lippert said. "There just aren't a lot of applications" big enough to drive that.

Several chief information officers said they were not planning to buy many more PC's, and that when they did, they would stretch the usual three-year replacement period to four or years or more — unless a big, unanticipated application came along to chew up processing power again. "We used to refresh frequently," said Kenneth W. Lacy, the chief information officer at United Parcel Service. "We're not doing it as much."

Mr. Lacy is hardly abandoning technology investment. U.P.S. is putting the finishing touches on a highly automated distribution center in Louisville, Ky., that will pump through 50,000 more packages an hour than the center it replaces. Like other information officers, Mr. Lacy is finding more powerful equipment available at cheaper prices.

This is true for servers, for example, the specialized computers that store data and applications and "serve" them to PC's and other machines, including home computers via the Web. Users are finding ways to turn one big server into several small ones, saving on both hardware and maintenance. An I.B.M. mainframe can, in effect, be "partitioned" into multiple servers that run the Linux operating system. Big Sun Microsystems machines can be carved up the same way. And the market for "blade" servers — basically, small servers built on computer cards that can all fit on a single rack — is growing fast. They can be linked and can provide the power of a big server for a fraction of the price.

Hand-held devices are replacing bigger computers, too. One information officer says that as soon as devices like the BlackBerry and the Palm can handle e-mail attachments reliably, he plans to start providing them as replacements for more expensive laptops. That is good news for makers of hand-held devices. But each generation of smaller, faster, more efficient machines means smaller profit margins as well. That is good for the consumer, but tougher to make money on.

The software industry has long had rich profit margins. Either a software company makes it or it does not, and when it does the returns are huge, especially if it dominates the market, as Microsoft does with Windows, or Oracle had until recently with its databases. Information officers expect software margins to narrow as competition increases and prices fall, particularly the software made by companies like Oracle, SAP, PeopleSoft and Siebel.
 
HESE companies make what is called enterprise software, which is at the heart of important business operations, from finance to human resources to customer service. They were big winners in the 1990's, when demand and prices were high.

But now, demand has collapsed and terms are changing. Prices are coming down. "The power has shifted," Mr. Morris of Capital One said.

The pricing pressure for software is expected to last beyond the tough times, just as it is for hardware. Each enterprise software maker originally specialized in something different — Oracle in databases, PeopleSoft in human resources, Siebel in customer-relationship management, SAP in enterprise-resource planning, which covers a variety of back-office operations.

Now SAP and PeopleSoft are trying to get into Siebel's business, while Oracle tries to nose in on all three. Microsoft, meanwhile, is coming up from the low end, selling databases and other back-office software to smaller companies.

Some buyers appear to regard the turn of events with glee. "It's come full circle on them now," Robert L. Mayo, the chief information officer at Subaru of America, a part of Fuji Heavy Industries, said of the new price pressure on vendors. For example, Mr. Mayo said, "in the late 90's you couldn't get a fixed-price contract to save your life." Instead, software vendors charged by time and materials, and the costs ballooned, he said. Today, vendors readily agree to fixed-price contracts, he said, and are more willing to sell a big project in smaller pieces, rather than lock up a customer with a major contract over several years.

The entire enterprise software industry came under fire for enormous multiyear contracts that went out of control. One of the biggest was Nestlé USA's contract with SAP — a $200 million software project begun in 1997 and only now being completed.

Such "big bang" contracts proved enormously lucrative for software vendors. But their days may be numbered. "I talk to other C.E.O.'s, and I'm surprised at the emotion around this issue," says William A. Zollars, the chief executive of the Yellow Corporation, a trucking and logistics company. "Companies have gotten a little disillusioned with big-bang solutions."

Yellow created a subsidiary in 1995 to manage technology; it now has 325 employees. Mr. Zollars continues to invest, spending $80 million annually, but on "bite size" projects. "We never do anything we can't complete in under a year," he said.

Likewise, Mr. Morris of Capital One enforces what he calls the 10-10 rule: a project must take "no more than 10 months, no more than $10 million."

So technology investment continues — but more efficient, more focused, with equipment and software that falls in price. By continuing to increase productivity, technology will keep helping the economy. How much the technology vendors will benefit is not yet known.