TechWatch: Does IT Really Matter?

Gates may be the world’s most optimistic proponent of information technology (IT). But in a recent address to a group of university researchers in Redmond, WA, he outlined what he called “Harsh Realities of the Period 2001 — Now,” according to an article by Kim Peterson in the Seattle Times, July 29, 2003. His list of realities includes: the Internet “bust,” complexity of IT management, security crises, email spam and distractions, crashes and downtime, monolithic application silos, and cynicism about software and computer science.

Meanwhile, in May 2003, the Harvard Business Review featured an article by Nicolas Carr entitled “IT Doesn’t Matter.” His thesis is that the IT relationship to business is like electricity and the railroad. For a time, both electricity and the railroad were strategic, but now they are simply a part of the infrastructure of business, useful but no basis for advantage. IT has arrived at this same point, he argues, and companies should decrease their IT spending.

Microsoft’s response to the “harsh realities,” however, is to increase investment in R&D. In fact, for the coming year they are increasing R&D at Microsoft 8% to $6.9 billion and increasing long-term R&D by 15%. It would be easy to assume Microsoft has done this because it must do so to thrive as a high-tech company, and high tech customers should consider Carr’s argument more carefully. Having done that, it is my opinion that Carr is mostly wrong, because his case is built on some fundamental misunderstandings of IT. To justify my conclusion I need to introduce a bit more of his argument.

Carr’s Case for IT Commoditization

Carr summarizes his basic argument:

“As information technology’s power and ubiquity have grown, its strategic importance has diminished. The way you approach IT investment and management will need to change dramatically.”

Railroads grew rapidly from 1841 until 1876. Businesses that were early users gained competitive advantage from the way they used the railroad. The electrical utilities generating capacity grew rapidly from 1889 to 1920, and companies that used electricity strategically gained a competitive advantage. Similarly, he argues, IT has grown rapidly between 1990 and 2002 as measured by the number of host computers on the Internet.

“Spurred by massive investment, capacity soon skyrockets, leading to falling prices and quickly, commoditization.”

Commoditization in turn leads to a vanishing advantage.

“When a resource becomes essential to competition but inconsequential to strategy, the risks it creates become more important than the advantages it provides.”

The net of all this, he argues, is that IT management should move from offense to defense. They should spend less, follow rather than lead, and focus on vulnerabilities instead of opportunities. Supporting his case he states,

“Studies of corporate IT spending consistently show that greater expenditures rarely translate into superior financial results. In fact, the opposite is usually true.”

In Support of Carr’s Case

Carr has raised some good issues. In particular, his argument that there is no need for a more powerful PC to run standard applications such as word processing is valid. He is also correct in recognizing that more IT spending doesn’t lead to greater profitability. In fact, Paul Strassman demonstrated this case with survey data in his 1990 book The Business Value of Computing. IT has never been about how much money is spent, but how wisely that money has been spent.

Until the electricity generation capability was adequate, the standard interface to that system was developed (e.g. the shape of the plug), and the system was sufficiently reliable, companies had to be very careful in their use of electricity. Today, at least in the West, companies can assume the availability of such power and build their strategies independent of the electricity industry—the goal is cost control within these assumptions. Corres-pondingly, many of the components of the IT system have achieved adequate capability, standard interfaces, and sufficient reliability to make business decisions without regard for the IT industry except in looking for a low cost solution.

In the early days of the railroads, there were few standards and inadequate access to tracks, so companies planned the location of their plants based on railroad access. Today, railroad movement of freight continues to grow, but there is adequate access almost anywhere. Correspondingly, an adequate telecommunications network and low-cost components such as PCs mean not having to plan a business around computer access.

Thus far, I support Carr’s argument. But now to the flaws.

What Is IT?

The weaknesses of Carr’s arguments are many. Let’s start with his definition of IT. He jumps around in the article, never settling on a sharp focus for IT, but here are several.

He confuses IT with the Internet and with PC based technology. Then he says IT is “first of all a transport mechanism,” forgetting that IT is also involved in the generation of the content and the interpretation of it.

The IT system has many more components than Carr recognizes. Further, the IT system is much bigger than its individual components.

Be Careful with Analogies

Analogies, such as Carr’s comparison of IT with electricity and the railroads, can be very helpful when the limits are properly understood, but Carr didn’t recognize these limits. There are characteristics of both the electrical system and the railroad that make them fundamentally different from the IT system.

While we have some standards in computing, many of them de facto proprietary standards such as Windows, the IT system is a long way from the universal plug to 60-cycle power.

The railroad analogy provides a more interesting limitation. The speed and movement of goods by rail has changed only modestly over the last 100 years. But the introduction of airplanes has vastly changed the transportation system, both in terms of speed as well as eliminating the “land locked” constraint. Next day service, just in time manufacturing, reduction of inventory, and off-shore manufacturing are all linked to the transportation system, not the rail system.

The air system, for example, is an order of magnitude faster than the rail system, no longer land bound, and with a different cost structure than the rail system. This “order of magnitude change” happens every five years in the IT system because of Moore’s Law. It is like moving from a rail system to an air system to a space system and beyond twice each decade!

So with an IT system, we might conclude that certain functions are now standard, no longer strategic, and we should focus on reducing their cost. Then one day we wake up and realize we no longer need to do these functions at all. Changes in the overall IT system have made some of the old steps in the business completely irrelevant.

Following his analogy, Carr would urge the business to go to sleep using the railroad system at low cost, and completely miss the aerospace system as a business tool. This discontinuity is easier to see in transportation. It takes a whole new industry and infrastructure to move from railroads to airplanes. Not so in IT. The changes are more subtle, but equally profound in their impact.

Strategic Role of IT

We see this more readily looking back. It was only twenty years ago that PCs hit the business world. Like any new technology, they brought plusses and minuses, but few modern businesses could run without them. Then ten years ago the Internet was opened to business and the general public through the development of the web browser. It took almost a decade to sort out that this could be very powerful for business (on-line ordering, website communication, etc.) and also produce some very wasteful and absurd business models, e.g.

What might the IT system bring to business in the next ten years? Some candidate ideas from the developing IT system include true electronic collaboration, truly mobile working environments, much more natural human/IT systems interfaces and ubiquitous computing (where we forget about the computing itself and look at the use of vastly distributed special purpose computers that communicate with each other).

One of the difficulties in seeing the future use of technology is timing—a capability here today may not offer opportunity until many years later because of the things that must surround and support it. The VCR is a great example of this. Invented in 1952, it became a commercial success more than 25 years later for an application that was never envisioned.


The IT system has enough parallels with railroads and electricity to make Carr’s analysis seem right, and he is partially right. Companies should do as he suggests and scale back expenditures on today’s mature processes, and not add new technology to do the same old thing just because they can. But he is completely wrong when he suggests the IT system is no longer strategic to business.

His advice would have your company sitting on the sidelines while a competitor emerges from nowhere (as Dell, Wall Mart, and eBay did in the 1990s) to exploit the next generation of opportunity that comes from Moore’s Law. These business breakthroughs would not have been possible in the 1980s because of the state of technology. In ten years we will be able to look back at companies that made strategic use of emerging technology (or even creative use of existing technology) to create new business models, new business advantage, perhaps in your business area. Carr’s advice could put you out of business.

Can you have the best of both worlds? Absolutely. Look for ways to decrease spending on IT in the day-to-day areas. But look for the discontinuities of the future by understanding the potential strategic advantage that could come from existing and future technologies. The R&D investment in IT will make certain there is a lot to consider.


Al Erisman is executive editor of Ethix, which he co-founded in 1998.
He spent 32 years at The Boeing Company, the last 11 as director of technology.
He was selected as a senior technical fellow of The Boeing Company in 1990,
and received his Ph.D. in applied mathematics from Iowa State University.