Information Age: News, analysis & insight for IT & business leaders

When Carr's subsequent book is published shortly (for a review of the book, see Information Age, May 2004), it will bear the title: 'Does IT matter?' Does this small change signal that Carr, perhaps buckling under the weight of criticism, has been persuaded to soften his stance?

There was little sign of this in his keynote presentation at the Effective IT Summit in London in March.

In a reasoned and academic presentation, Carr demonstrated that, in the time since his article caused such a furore, he has had time to fill out and develop his arguments. Carr began on a disarming note: "Does IT Matter? Of course it does," he said.

IT, he noted, is ubiquitous, is integral to modern business processes, is the dominant capital expense for most companies and is a pre-requisite to business survival. He even agreed that, "in the right circumstances", IT can significantly boost industrial productivity.

"But," he argued, "none of this means that IT is a competitive resource, that IT provides a competitive edge that lasts long enough to repay the investment."

In Carr's view, there are two types of technology: 'proprietary', which means it is patented, protected and exclusive; and 'infrastructural', which is technology that is shared broadly by all companies in an industry or region, such as rail, telephones or electricity.

This is where Carr becomes controversial: he places IT in the infrastructural category. Just as electricity or water no longer provide competitive advantage, so it is increasingly with IT.

In its early stages, he says, an infrastructural technology acts like a proprietary technology, because once one company invests in it, it is hard for other companies to replicate the technology. That gives the richest companies a perceptible competitive edge.

But that advantage diminishes over time as the technology becomes cheaper and ubiquitous. And in the final stages, companies tend to fall behind the curve, says Carr; "They are still looking to IT to provide a competitive advantage, even though it is not capable of doing so." This, he says, leads to overspending and wasteful investments.

Carr points out that the move from proprietary to infrastructure does not happen evenly. Financial services, he notes, is ahead of healthcare, and the developed world is ahead of the developing world.

But his thesis is that across industry generally, "we are much further along the process than most people believe".

This is the core of the argument that appeared in Carr's infamous article, and the one that, as he put it, "struck a nerve" with CEOs of IT suppliers as well as with many IT directors and CIOs. The explicit strategy that results from this analysis is to spend less and to lower one's expectations from IT.

But Carr has developed his analysis with two further, related, points. First, a lot of people said that IT is fundamentally different to say, the railways - IT is malleable and flexible, they argued, and has endless innovation potential.

But Carr argues that if it is different, it is different in a way that is actually exaggerating the move from proprietary to infrastructural. First, intense competition in hardware is leading to 'overshooting'.

This is a concept borrowed from Harvard Professor Clayton Christensen, and describes how suppliers keen to please their most demanding customers tend to push the technology beyond the needs of most potential customers. In this way, powerful technology becomes quickly and cheaply available to all.

A second factor is what Carr calls "the extreme economics of scale" of software. The open source movement, offshore development and 'factory' production of software are all examples of how software is becoming easier and cheaper to develop and re-use.

Carr points out that in the early stages of IT, various companies were able to get long-lasting advantages from their technology investments.

He cites the Sabre reservation system, the American Hospital Supply ASAP ordering systems, and Reuters' Monitor, the information and trading system.

Now, he says, these advantages have rapidly eroded as accessibility, affordability and standardisation have increased. And while there are always exceptions, few organisations will gain a competitive edge through IT and most of those that do will probably be making a mistake.

All of this requires a new set of imperatives for IT management, says Carr.

These include spending less, following rather than leading, innovating only when the risks are low, and focusing more on vulnerabilities (risk management) than opportunities.

Carr cites several companies in the US that are doing this: Verizon has cut its IT spending from 6% to 4% of total spending through a number of initiatives, including, for example, a moratorium on server spending until vendors reduced prices; General Motors has cut $800 million from its IT spending through similar initiatives, including the use of offshore outsourcing; General Electric has cut its IT spending from 2.8% to 2.5%, partly by using open source software.

Carr finished his presentation with a quote from Max Hopper, a former CIO of American Airlines: "IT will come to be thought of more like electricity or the telephone network, rather than as a decisive source of organisational advantage. People like me will have succeeded when we have worked ourselves out of our jobs. Only then will our organisations be capable of embracing the true promise of information technology."

That was in 1991.

Do you agree with Nicholas Carr? Write to letters@information-age.com to let us know.