June 2004

Why I.T. Matters

The inventor of Ethernet refutes the claim that information technology has lost its strategic value.

By Robert M. Metcalfe

A year ago, Harvard Business Review published a now infamous article called "IT Doesn't Matter." Its author, the magazine's then executive editor Nicholas G. Carr, argued that information technology no longer gives businesses a competitive edge. Carr called information technology managers impatient, wasteful, passive, and lured by the chorus of hype about the so-called strategic value of IT.

Harvard Business Review has 243,000 extremely influential readers. So if it publishes an article saying that information technology doesn't matter, then an awful lot of important business leaders are going to believe it. And if they do, they'll run their companies-and our economy-into a ditch.

Many commentators have debunked Carr's article since it appeared last year. So many in fact that I feel like Elizabeth Taylor's ninth husband: I know what to do, but how to make it interesting? But Carr's article just won't stay debunked. And now he has expanded his thesis into a new book called Does IT Matter?, which the Harvard Business School Press published in April. The question-style title hints at some backpedaling, but Carr's point is basically unchanged-and it needs debunking yet again.

Since I do not subscribe to the ink-on-dead-trees version of the magazine, I bought my copy of Carr's May 2003 paper through Amazon.com. It was delivered over the Internet in minutes as a PDF file for $7.00. Carr's new book is also listed on Amazon.com, a triumph of IT-enabled corporate strategy. We see that IT apparently matters to Harvard.

Carr himself has a website, nicholasgcarr.com. IT apparently matters to Carr.

Let's face it: IT matters to everyone.

Two Trillion Reasons that I.T. Matters

I asked how much IT matters of Frank Gens, senior vice president for the information technology market research giant IDC. (Full disclosure: IDC is owned by IDG, on whose board I serve.) IDC reports that the global investment in information technology (including telecommunications) totaled $1.9 trillion in 2003 and, despite Carr, will climb to $2.0 trillion in 2004.

According to a 2003 IDC survey, non-IT business executives spend 20 percent of their time thinking about IT. Are they wasting their time? Again despite Carr, almost 60 percent say that the strategic importance of IT is increasing; only 2 percent say the importance is decreasing. Carr may claim these Harvard-MBA-type executives are foolish or misguided, but 55 percent feel that their companies should use information technology more aggressively; 43 percent feel their usage is just right; and only 2 percent feel that they should be less aggressive.

In Carr's world, information technology managers are apparently fools, or even frauds, to the tune of $2 trillion per year. Presumably, these managers slavishly upgrade to whatever new thing vendors want to sell. But in the real world, millions of people already work hard to spend their IT budgets wisely. The computer-trade press has been covering this complicated process for almost 40 years.

In warding off his debunkers, Carr has offered some clarifications of his argument. He doesn't really mean that information technology doesn't matter; rather, he says, his point is that because IT has been commoditized, like electricity, it confers upon its business users no competitive advantage. He also clarifies that he does not mean that information itself doesn't matter, nor does he mean that the people using the technology don't matter. What really doesn't matter, he says, is the no-longer-proprietary technology infrastructure for storing, processing, and transmitting information. So we can only hope that most of Harvard Business Review's captains of industry read beyond the article titles before dropping the magazine on their coffee tables.

Carr concludes that since information technology no longer provides a competitive advantage to businesses, they should stop spending wildly on advanced information technology products and services. He admonishes managers to stop being suckers for the latest cool products from Cisco, Intel, Microsoft, Oracle, et al. IT managers should stop squandering corporate assets and begin acting in the best interests of their shareholders. They should become boring minimizers of IT cost and risk.

As evidence, Carr points out that my 30-year-old baby, Ethernet, has been standardized and commoditized. It's true that last year more than 184 million new Ethernet ports were shipped, at a value of $12.5 billion, and that anyone can buy them. Most of those ports are the current mainstream version of Ethernet, which carries data over wires on local-area networks at 10 or 100 megabits per second.

But now that the post-Internet-bubble nuclear winter is almost over, Ethernet is speeding up, to beyond 1,000 megabits (one gigabit) per second. Ethernet is going into wide-area networks. It's going wireless. It's going into embedded systems-the eight billion microprocessors shipped every year that don't go into PCs.

New Ethernet standards are being created, new commoditization races are being started, and Ethernet, if ever it wasn't, is once again a tool of corporate strategy. In the article and now again in his book, Carr wrongly equates today's information technologies with electricity, and then he wrongly characterizes electricity as static. In short, Carr, deep into a post-bubble depression, wrongly declares the end of history.

The history of electricity is not over, however. Controlling electrical power grids is still famously problematic, and that's to say nothing of the exciting developments in technologies such as wind, solar, fission, fusion, hydrogen, and batteries, all of which present strategic opportunities. And information technology is bigger and more recent than electricity. Both are still rapidly evolving; both are very much alive as important elements of corporate strategy.

Much of the research on information technology usage that Carr cites is of dubious validity. Take, for example, the studies that, as Carr puts it, "consistently show" that expenditure on IT as a fraction of company revenue is inversely correlated with financial performance. One study that Carr cites states that the 25 companies with the highest economic returns spent on average just .8 percent of revenues on IT, while the typical company spent 3.7 percent. But this hardly proves Carr's conclusion. Rather, it indicates that companies investing wisely in IT increase revenues much faster than those that invest unwisely, too little, or not at all. Companies that invest poorly in IT don't increase revenues as quickly, so their IT expenditures are higher as a fraction of revenue. Companies that invest unwisely in IT go out of business and are not counted in the studies. IT still matters.