About a year ago, the Harvard Business Review published an article titled "IT Doesn’t Matter." It ignited a vehement and often acrimonious debate over the value of information technology. Since then, Nicholas Carr, the author of the article, has expanded on his original thesis that while IT’s value will increase as it becomes more standardized and ubiquitous, "the ability of any one company to use IT in a distinctive way to gain competitive advantage will diminish until...it will make more sense to manage IT as a commodity input?something that is absolutely necessary but [that] isn’t going to set you apart from competitors."
Carr has faced off with detractors in print and onstage, and this month sees the publication of his new book, Does IT Matter? the title of which suggests Carr may have backed off his original position a bit. (He hasn’t.) Carr spoke recently with CIO Editor in Chief Abbie Lundberg to explore his conclusions and the assumptions underlying them, one of which is that all information technology is "infrastructure." (See "The Engine that Drives Success" on Page 36 for author Don Tapscott’s response to Carr.)
CIO: Most people distinguish between infrastructure technologies and the applications that ride on them. You seem not to make that distinction.
Nicholas Carr: Over time, what we call infrastructure has expanded to incorporate much more of the hardware, everything from PCs to mainframes, and much more of the software, from fairly esoteric utility software that’s been incorporated into operating systems through various application packages as well. I don’t think that process is over. More and more of the IT that companies use will, over time, become viewed as infrastructure. And what isn’t infrastructure continues to shrink and is meaningful to a smaller and smaller set of companies as a way to differentiate themselves.
So if you think of all of IT as a pie, what proportion of the pie falls into the infrastructure segment?
For most companies, you may as well think of it as 100 percent infrastructure and base your approach to IT management on that assumption. Ultimately, you’d want to just manage it as infrastructure, by which I mean a shared set of standardized hardware and software that is fairly homogenous across companies.
I agree with you that the greatest value from IT will come once the infrastructure is standardized, secure, reliable, interoperable, transparent. Where we differ is that I believe once we have that, there will be a lot of opportunity for differentiation and new types of business activities that are IT-enabled. Take Google’s search algorithms, which it has developed in a specialized, proprietary way on top of the common platform of the Internet to gain a unique advantage, which it has, in fact, been able to sustain.
A fairly small subset of companies can use distinctive secret algorithms, in effect, to create a fundamental competitive advantage. And if you’re a search engine, in which your entire existence basically is those algorithms, then sure. But how widely applicable is Google’s business model to other companies?
What about Amazon? It sells stuff. By leveraging the technologies it has developed on top of a common open infrastructure, the company has been able to achieve competitive advantage.
Whenever you have a new thing [like the Internet], you’re going to have new companies rise up and take advantage of that new thing. So you have companies like eBay and Amazon and Google. But what’s interesting to me about the Internet is how few companies you have like Amazon, eBay and Google.
The companies you cite in your book from 20, 30 years ago?American Airlines, American Hospital Supply, Reuters?are really the poster children of competitive advantage from that era. There weren’t very many of them either.
If you look back 20 years ago, even 10 years ago, companies were able to use creative IT systems themselves as competitive barriers. From American Airlines’ standpoint, it took an enormous amount of time to create Sabre, and it was horribly expensive. But because it was so difficult, it took a considerable period of time for competitors to replicate the system. And in that time in which American had a competitive advantage, based on its distinctive technology, it was able to get incredible economic and business value out of that. But I would argue that as IT has become more standardized?cheaper, more ubiquitous?it has become harder and harder to use the technology itself as a competitive barrier because it becomes easier for competitors to replicate your systems.
It’s important to distinguish between what is and what isn’t infrastructure. Amazon didn’t differentiate by having access to the Internet. It differentiated on how it set up its shopping carts, and how it tracked customer preferences, and all of those things it built into its business model. Those are the things that gave it its distinctiveness.
Certainly, Amazon has done a lot of great stuff with Internet technology.... Amazon is an example of a company that really did gain a first-mover advantage from the Internet, as eBay did as well.
So isn’t being able to leverage the right technology in the right way an important part of companies’ futures?
Sure, that is an important part, and that’s true of many business resources. The ability to build on that doesn’t in and of itself tell you that that resource, whether it’s IT or something else, is a strategic resource or not. It’s a competitive necessity, but it’s important not to confuse the technology, necessarily, with the source of advantage.
Are there other resources you could point to where the resource in and of itself does provide a competitive advantage?
It’s very rare. One of the reasons for writing my article and my book is to underscore the danger of trying to tie competitive advantage to a particular resource rather than to a complex system of activities and resources and such.
It’s been common wisdom for years that the value comes not from technology for technology’s sake, but what you do with the technology?how you use it to optimize or change your business model, how you use it to transform your processes. When you talk about the erosion of competitive advantage around information technology, are you talking strictly about hardware and software, or are you talking about those things?business processes, business models?that are enabled by IT?
I’m talking beyond just the hardware and software. When you have this increasingly standardized new business infrastructure that we call IT, even at the process level, it starts to quickly erode certain traditional advantages that companies have. Best-practice automation diffuses very quickly throughout an industry. So you can see those advantages start to erode as everyone adapts to the new infrastructure.
So why are there still such dramatic differences between one company and another? Why is it that Kmart imploded, when it’s the same kind of business as Wal-Mart? Why is it that when you call L.L. Bean, you get a superior form of customer service than when you call Eddie Bauer?
It’s not all technology. Wal-Mart had a fundamentally different approach to the discount retailing business in terms of where it located its stores, the type of merchandise it carried.
Having said that, certainly Wal-Mart has been an excellent user of information technology and continues to be so. My point is not that technology could never supply a competitive advantage. My point is that its ability to differentiate one company from another has diminished over time as it’s become more and more standardized. Wal-Mart didn’t create its technology yesterday. When it really set itself apart from competitors?in particular on the supply chain side?was back many years ago, when that technology was new. But, over time, things get routinized. Processes become standardized, and the technologies underpinning them become standardized. That ability to use technology to set yourself apart shrinks.
The First-Mover Advantage
I just don’t see that it was easier to gain competitive advantage from IT 20 or 30 years ago.
It was hard back then. It was the very difficulty of doing it that meant that your advantages tended to last a long time. The important thing about a competitive advantage is not just having one, it’s how long you can hold on to it. And so the difficulty of innovating with IT in earlier times meant that if you did innovate successfully, you could hold on to that advantage for quite a long time. You can still innovate today. The question is, how quickly are your competitors going to be able to replicate those innovations? That competitive replication cycle has sped up and continues to speed up.
How does that play out with companies like Amazon and eBay? Their first-mover advantage has been sustained.
Right, and if you are able to use your IT advantage to leverage other advantages, then that’s great. But the point is, there are millions of companies in the world. They have to ask themselves, are Amazon and eBay models that we can follow? I think very few companies can follow the eBay or the Amazon model. But isn’t the point that companies shouldn’t be thinking about whom they should follow?they should be thinking about how they get advantage. What do they leverage? How do they use their resources?whether they’re human or technology resources?how do they use those in a unique and strategic way to make their companies prosper?
Sure, yes. It does tend to be how you use various resources in combination distinctively.
Let’s go back to Google, where you have the infrastructure that you want to be stable, reliable, low-cost, easy to manage. Then you have that distinctive asset that gives you your uniqueness.
Google’s most valuable lessons to other companies are in its management of basic IT. It’s a great user of commodity goods. It was very interesting that it came out and said, We’re not going to rush to buy the Titanium chip; the old stuff is good enough for us. The company obviously has an incredibly intense need for processing power and for IT sophistication, and yet it’s cobbling its systems together out of cheap, old components. That provides a great lesson to other companies in how to approach their own investments in IT. The fact that it’s a search engine built on algorithms is probably less meaningful to other companies.
Certainly that distinctive, unique thing is going to be different from company to company. As far as Google’s resourcefulness goes, those are the lessons of the recession. I don’t know a CIO who hasn’t been consolidating servers, using Linux wherever possible and renegotiating contracts with vendors. That’s been going on for three years now, and it’s the right thing to do, a smart thing to do. That’s the way to manage the infrastructure part of your IT.
For most companies, it’s all infrastructure at this point. You’re right that companies have to find those special things that can distinguish them from their competitors. But IT is only one place to look for those things. For most companies, I would say it’s not going to be in information technology anymore.
So is there no more place for innovation within corporate IT?
I think there will continue to be lots of innovation in corporate IT. But it will take place at the infrastructural level, it will be driven by vendors, and it will be shared.
If you look at IT innovation from an individual company’s standpoint, there are two questions you have to ask. One is, how much more are we going to have to spend to be an innovator than if we waited and took the copycat route? The second question, and this is extremely important, is how long are we going to be able to maintain that advantage before it’s replicated either directly by competitors or by vendors?
If companies look at those two things and find a way to reduce the extra costs that are usually involved in being an IT innovator, or to block competitors from replicating their systems, then innovation can still make sense. But at this point, those opportunities are rare. For a few companies, it’ll make sense to pursue them, but it makes more sense for most companies to be followers.
The CIO’s Burden
Most of the CIOs I talk with complain bitterly about having to put the bulk of their budget toward "keeping the lights on" in IT. Only a small portion is left for the things that are going to help the company be distinctive. Once companies are able to run their underlying IT cheaply, reliably, securely, isn’t there an opportunity to use some of the freed-up dollars to go from "let’s get this stuff to work" to "what can we really do with it?"
What can we really do with it is a very big question. Usually, those kinds of decisions are the essence of business management. It doesn’t seem to me that an IT staff would make those kind of decisions.
It doesn’t sound like you see much of a future for the CIO.
A lot of the most successful CIOs today are, in fact, ones who have implicitly or explicitly accepted the commoditization of IT and are focusing on how they can capitalize on that to drive down the costs and increase reliability.
Should the CIO still be on the executive committee?