During the past two years, business has dramatically cut back on information technology spending, and next year is shaping up to be nearly as bad. Considering the dismal state of the economy and the exorbitant IT spending that immediately preceded the slump, that should come as no surprise. No one expects IT spending to go through a boom like it did during the late 1990s.
But what may come as a surprise is this: Even when IT spending turns around, its prospects dont look so good. Spending over the long haul will likely grow at just half the rate it did during the last four decades. Consider the numbers: For 40 years, from 1960 to 2000, IT spending in the U.S. grew an average of 13 percent a year, peaking in 2000 when it jumped 23 percent, according to Morgan Stanley Dean Witter. Since then, its been downhill. In 2001, IT spending dropped 10 percent, according to the U.S. Department of Commerce. This year spending is expected to drop another 2 percent, and in 2003 spending will probably be flat, according to Goldman Sachs. (And dont be surprised to see 2003 spending go into the negative if the economy continues its current slide.)
Thats bad enough, but heres the rub: “In the long term, IT spending will grow about 6 percent to 7 percent a year,” says Laura Conigliaro, an analyst with Goldman Sachs—half the rate it grew from 1960 to 2000. And even that may be optimistic: “Demand will be up maybe 3 percent a year,” says Howard Anderson, senior managing director of YankeeTek Ventures. “The old days are over.”
We just might be witnessing the end of an era—the end of tech as a growth industry. That is a heretical view, particularly here in Silicon Valley, where hope springs eternal. The mantra of the faithful continues to be that as long as computer chips keep getting faster and cheaper, and programmers keep finding new uses for all that power, business will keep writing bigger and bigger checks. But that is all in the past.
If one had to pick a single reason why growth will dry up, it is this: The IT industry has simply gotten too big. Consider this statistic: In 2000, IT accounted for 47 percent of total capital equipment spending by U.S. business, according to the U.S. Department of Commerce. For every dollar businesses spent on trucks, planes, machinery, fixtures, furniture and other equipment, almost another dollar was spent on computer hardware and software.
Granted, 2000 was the height of the bubble, when business seemed to have an insatiable appetite for everything tech. Yet even before the excesses of 2000, spending on IT may have reached its natural level. From 1990 to 1999, IT accounted for an annual average of 43 percent of total capital equipment spending in the U.S.
As integral as computers are to running a modern corporation, this is still a material world. Goods must be manufactured, shipped and sold. Consumers still go out to restaurants for dinner and to stores for groceries. Workers continue to fly on planes and work in offices. Not every human activity can be encapsulated in a computer—as the dot-coms found out. And that wont change until we can say, “Beam me up, Scotty.”
But thats not the only barrier to growth. There is also something called the law of large numbers at work. The bigger an industry or company gets, the harder it is to achieve double-digit rates of growth. And IT has gotten big—real big. In 2000, U.S. business spent $447 billion on IT equipment. To achieve 7 percent growth would have meant convincing business to buy an additional $31 billion worth of hardware and software on top of the $447 billion—more than the combined revenues of EMC, Sun Microsystems and Oracle.
But size isnt the only reason IT growth is slowing. If business could continue to get demonstrable returns on its investment, it would keep spending. The problem is that too often there is little, if any, measurable return on new IT investments.
Take hardware upgrades, for example. In the current economic slowdown, hardware upgrades are being put off, as evidenced by the latest Goldman Sachs survey of CIOs that showed them as dead last in spending priorities. This situation is not expected to change much even after the economy recovers.
In the early years, new PCs offered substantial improvements in productivity. But is there really a cost justification for upgrading PCs every three years these days? Unless the user is a scientist pushing the limits of computation, the answer is probably no.
Its much the same in the back room. Businesses spent billions upgrading their mainframes and servers when they went through the Y2K conversion, and these machines continue to run just fine. Thanks to the efforts of IT vendors, computers are much more reliable today than ever before, and will likely become even more reliable in the future.
There is every reason to believe that the upgrade cycle for hardware will lengthen, not shorten, in the years ahead. If the average upgrade cycle for computers goes from three years to four years, or even five years, thats a lot fewer computers sold. The U.S. Department of Commerce, in its “Digital Economy 2002” report, says: “Some experts believe that businesses are finding that the IT equipment is not becoming obsolete as rapidly as a few years ago, and they are stretching out their replacement cycle from 3 to 3.5 or 4 years.”
And business spends more money on software than computers. Yet here too, there are reasons to anticipate a slowdown in spending growth. Name just about any function within a company—manufacturing, distribution, sales, accounting, administration—and chances are it has already been automated. Thats what business has spent the last decade doing, putting in enterprise software to automate the various business processes.
That isnt to say that there wont be incremental IT improvements in each of these areas—there will. But future software projects are not likely to be the massive, bet-the-company efforts that ERP vendors like SAP and Oracle made their billions on. New software projects like Web services will be implemented incrementally, not on the sort of wholesale scale as in the past, says Goldman Sachs analyst Rick Sherlund. “The CEO is not interested in big projects anymore,” he says.
The Next Big Thing
Of course, there is always the potential “Next Big Thing”—a technology that no one can predict and no one can ignore. Yet even if it comes along, CEOs will be much more cautious about how they invest their money after the billions of dollars that were wasted on superfluous Internet projects in the past few years.
The impact of such a long-term slowdown in IT spending will be widespread. We are already getting a glimpse of what it might be in the current economic slowdown. Venture capital money is already slowing dramatically, and will continue to erode if it becomes clear that IT spending will permanently slow. Less venture capital means fewer startups—and less innovation.
Slower growth also means greater competition among IT vendors. This is already evident in lower prices and slimmer profit margins among IT vendors. Lower prices are certainly good news for CIOs, but slimmer profit margins will mean less money available for research and development. “More companies will turn to the Dell model,” says Anderson. With R&D money cut back, technical breakthroughs will be harder to come by.
If IT becomes a slow-growth industry, it will also become more difficult for IT vendors and CIOs to attract talented people. Talent migrates to where the money and the action are, and the talent will quickly move elsewhere.
The end result? The IT industry could start looking a lot like the automobile industry, a large and important business, but not one known for growth or innovation. This was bound to happen sometime, but it is looking like the time may arrive a lot sooner than many people expected.
Eric Nee, a longtime observer of Silicon Valley, has served in a variety of editorial positions at Forbes, Fortune and Upside magazines. His next column will appear in January.