Debate Transcript: Does IT Matter?

The following is the transcript of a debate between Nicholas G. Carr, author of "Does IT Matter? Information Technology and the Corrosion of Competitive Advantage" and Bob Metcalfe, Ethernet inventor, 3Com founder and venture capitalist at the Premier 100 Conference for IT Leaders, March 8, 2004.

Carr opened the debate, responding to Metcalfe's pre-debate presentation.

CARR: Bob's VC investments must be doing very poorly, because he obviously has a lot of anger. I'm glad to give him the opportunity to vent here. His dramatic misreading of my article ["IT Doesn't Matter," published by the Harvard Business Review. (see related stories)] was quite impressive, and I can thus put him at the head of the pack of the failed debunkers that he listed at the start of his presentation.
I think the interesting thing, when you think back over everything that Bob has just said, is that he never really addressed the point of whether IT - information technology - can give a company a defensible advantage. Whether, in fact, we're at the stage of the development of the IT infrastructure that most if not virtually all IT innovations will be rapidly replicated by competitors - which is, after all, my point.
He raised an issue that others have raised, which is, "Oh, well, IT never mattered, then." But I think that's not true, and I think by saying that, you kind of deliberately skew the debate. Because in fact if you look back at the history of IT, you see that information technology - the systems themselves - in the early stages of the development of IT, actually provided extremely strong competitive advantages.
You had, for instance, American Airlines with its Saber system, which kept the competition at bay for many years. And even when the competition was able to leapfrog American's system, as United did, American had had that advantage for so long, and had leveraged that advantage so strongly, that United's system was unable to dislodge it. And why did it have that advantage? Because it was hard to do. It took American Airlines 10 years to develop that system, and loads of money. At that point in the development of the IT infrastructure, it just took a long time for competitors to come behind and replicate the system. Today, because of all the advances of IT - because for many users in fact the capabilities are out beyond their needs -- that replication cycle goes much, much faster.
Another example is that when we had closed networks, you could set up a closed network with your suppliers on one end and your customers on the other, and you could lock out your competitors. And we saw companies like American Hospital Supply do that. It was the first of the medical suppliers to go out and put its ordering terminals in hospital buying offices. Essentially, because the buyers didn't want to go out and get terminals from other suppliers, it locked out all the competitors. Similarly Reuters did that with Reuters Monitor - set up a private network to send financial information out to financial traders, and locked out competitors.
But what's happened there as well? When we moved to open networks and the Internet in particular, not only did closed networks no longer have advantages, but they turned into disadvantages. And so through this process - and Bob may contend that history doesn't matter, and that's fine, but I actually think it does - through this process which we've seen over and over again in business, we see the advantage potential of the technology become neutralized as it spreads, as it advances, as it becomes standardized, as it becomes homogenized - and this is the point where Bob and many of those who have tried to rebut the article gloss over. They default back to, "Oh, Ethernet, we're going beyond 1 gigabit." Fine. But that's available to everyone - that's the point. So yes, I'm not arguing that IT is unimportant, I'm not arguing that innovation is over; in fact I think we'll see continued very strong innovation. But the point is it's innovation in the infrastructure. It's shared by all companies; it provides competitive advantage to none. If you go with Bob's advice, I think what you'll end up doing is simply wasting a lot of your company's money.

METCALFE: In that context then, what would be a list of the things that are a source of competitive advantage?

CARR: Superior products, superior services, scale, proprietary manufacturing processes, brand. . .

METCALFE: But don't all of those things rely on information technology, except possibly for brand? And maybe even brand also. . .

CARR: As I said, information technology is a part of almost every modern business process. But that doesn't tell us whether IT itself can provide competitive advantage. Think of flour. I can give a ton of flour to two different bakers; one baker is going to create a better loaf of bread. Does that mean flour is a strategic resource? No - it's essential, but that doesn't mean the baker should go out and spend more money on flour. It's a commodity input and it should be managed like a commodity input.

METCALFE: How about new technology? That is, information technology (singular) - it's not really one technology, it's many technologies. So Intel's Pentium, Oracle SQL databases, Cisco routers [aside] - what about things like robotics and GPS and RFID and the spectrum above 50 GHz and embedded networking, etc.?

CARR: I think we're at the stage in the development of the infrastructure that new innovations become commoditized very quickly. Look at RFID. What is Wal-Mart doing by going out and telling its 100 largest suppliers, which is in effect the entire consumer products industry of America, "You have to adopt RFID and send us pallets and product boxes with tags within a year?" Wal-Mart is commoditizing RFID as quickly as possible - it's neutralizing RFID as a potential competitive weapon that a competitor could use. And it's offloading the cost of doing that onto its suppliers. And why is Wal-Mart doing that? For a very simple reason - because it's the cost leader; because it's the dominant force in the retail supply chain, it knows that any productivity gains that come out of the commoditization of RFID, a disproportionate amount will fall to it. So I think in one new area of IT after another, you see these forces that just push them to become part of the commodity infrastructure very quickly. They may create greater productivity gains, but those gains ultimately go back to the consumer; they're not going to one company's bottom line.

METCALFE: Before going to RFID, how did Wal-Mart achieve this preeminence? Didn't they do that by the strategic use of technology?

CARR: I think they've been a great user of technology, and particularly in the earlier stages of the development of the retail supply line. I think there are other reasons - for instance, store location, merchandising strategy, selling brand-name products at a lower price, that also contributed to it. But I certainly agree that they've been a very good user of IT, and have used it in a way that buttressed the advantage that they've established.

METCALFE: Now in the field of RFID, there have been announcements by Wal-Mart and the Department of Defense and others. But if you go inside that industry, there's a huge disagreement as to timing and what the proper standards are. Some companies are racing to adopt early; others are dragging their feet. Aren't we seeing the playing out of a strategy game involving information technology?

CARR: No. I think we're seeing what we saw with barcoding, which is you take a new technology that improves productivity greatly; you turn it into a standard; it very rapidly becomes a standard. It boosts productivity throughout the retail supply chain, but quickly becomes invisible at a competitive level. And I think the same thing's going to happen very quickly with RFID, particularly as vendors -- as vendors do -- quickly replicate best practices into their products.

METCALFE: So Sears used to be the biggest retailer; now it's not anymore. Was that because they chose bar codes or didn't choose bar codes or chose supply chain management or didn't? What was the problem with Sears?

CARR: Crappy clothes.

METCALFE: Isn't that a Kenmore suit you're wearing?

QUESTION FROM THE FLOOR: Some user companies such as American Express are actually getting patents for their IT-driven business processes and systems. Isn't that proprietary in your model?

CARR: Yeah, and I heard [CIO] Glen [Salow] from American Express talk about that [in the opening keynote (see story)]. Patents continue to be a great source of competitive advantage. So if you can get a patent on some particular use of IT, yes, I would say that that is a use of IT that can translate into a competitive advantage. How broadly companies will be able to do that is the question.

METCALFE: Incidentally, in your terminology, I have a problem with your making "proprietary" and "infrastructural" as opposites. The opposite of "proprietary" is usually "open." The opposite of "infrastructure" is usually, like, "applications." You confound them in such a way. Why do you do that?

CARR: To say the opposite of "infrastructure" is "application" is looking at the term "infrastructure" from a very inside-IT way. The way I'm using it is that infrastructure - and it doesn't have to be IT - is something that's broadly shared. Something that's proprietary is held and controlled by an individual company.

QUESTION FROM THE FLOOR: It seems like your argument is that innovation will happen, but it won't do you any good.

CARR: As I said, I think innovation continues. Companies need to innovate to stay ahead of their competition. But IT isn't the only place - there are many different areas to innovate. Innovation always carries cost and carries risks. My point is that we've reached the point where IT, for the vast majority of companies, isn't the best focus for innovation. You'll just end up taking on too much risk and spending too much. IT capabilities will continue to progress, and maintaining competitive parity will continue to be essential to survival. But if you try to jump ahead and turn it into a competitive advantage, I think you're going to spend a lot of money; you're going to find competitors are able to keep pace with any gains you get.

METCALFE: On innovation, I think innovation isn't over, and that there are new technologies coming. I've listed some; I could list many others. Companies will differ in their adoption patterns of these new technologies, and therefore some will become Wal-Mart and others will become Sears. RFID is a great example. Companies are now deciding one by one whether they're going to adopt RFID - how soon, how late. And in the course of making those decisions, they are sealing their fate. They are making strategy decisions based on emerging technology. If they make them wrong, they will suffer.

CARR: I think right now the better strategy is to wait. It's true that IT advances so quickly; it's also true that costs drop extremely quickly. To be an innovator means higher costs, higher risk. You have to ask yourself, "How long is that IT-based advantage going to last?" Because it's got to last long enough to pay back what you're sacrificing with higher costs and higher risk. And for most companies now, that equation will come out with the answer: Don't get out ahead. It's just not worth it strategically.

METCALFE: I'm struggling to get around to actually rebutting your argument. I think we're right at core there. It has to do with how long you can sustain a competitive advantage. How long, which you argue is getting much shorter now. So therefore whatever tradeoffs we were doing in the past, they're different now, because the time [you retain the competitive advantage] is shorter relative to the risk and cost of being out on the cutting edge. Is that it?

CARR: That's one of the essential points, yes. With any innovation, in order for it to be strategic, you have to be able to hold onto those benefits at least for long enough to pay back your initial investment. I think that time clearly has shrunk.

METCALFE: My counterargument would be that it has shrunk for the things that you think are information technology. But there are other technologies emerging for which the shrinkage has not occurred yet. For example, GPS is now generating latitudes and longitudes. Businesses are now forming to exploit that wealth of information about where things and people are. The payback period on that is likely to be larger than, say, buying faster Ethernet.

CARR: I'm certainly not arguing that we're at the end of technology progress. I think technology comes in waves, and there will be another wave. If you ask me about GPS itself, I don't see that many barriers to the commoditization of that technology. So I'm not sure you're going to be able to create a distinctive application and then defend it against replication. I don't see where those competitive barriers are. But I completely agree with you that we're going to see additional waves of other technologies, some of which may well build on the IT infrastructure.

Nothing is more of a commodity than money or capital. So would Nick claim that capital is not a differentiator?

CARR: No. But it's a cute analogy.

METCALFE: Money does matter.

CARR: How you spend money matters.

METCALFE: How you employ IT matters.

CARR: You can use any business resource well, or profitably or unprofitably. I don't see IT as being different from that. But still you're left with, is it a commodity resource? Do we manage it and invest in it as something that can give us a competitive advantage? Even at the usage level, what we're seeing through things like business process outsourcing is that increasingly, even the uses of IT can become commoditized, and I think that will be a trend that continues as well, and even picks up. But there's no question that given the cost and ubiquity of IT, a company can certainly put itself at a disadvantage by being a poor, sloppy, wasteful user.

QUESTION FROM THE FLOOR: A question for Bob - what would you say is the biggest IT innovation on the horizon right now?

METCALFE: There are so many to choose from. Embedded processing. There are 8 billion microprocessors shipped every year now, which is more than one for every person on earth. About 1% of them are networked. So networking those embedded processors will be a source of competitive advantage for companies that do it well.

CARR: I certainly agree on the consumer side, the embedding of information technology into consumer products is going to be an enormous area for innovation. And as I've said, in general I think we are seeing the real locus of IT innovation moving from the corporate side to the consumer side.

METCALFE: The frequencies above 50 GHz have recently become unlicensed by the FCC. And the chips necessary to access those much higher frequencies, which is a huge bandwidth - more than all of television and radio many times over - those frequencies are becoming available, and they'll be employed in lots of proprietary and creative ways to bring competitive advantage to companies over time.

CARR: Vendors are awfully good at qualifying, saying, "Of course it's not the IT, it's the way we use it." Well gee, you sold a lot of IT for quite good margins, and now you're coming back and saying, "Oh, now all that stuff we sold you doesn't matter - it's only the latest business intelligence applications that we're peddling today." So I'm very suspicious of this very swishy use of definitions by the IT industry.

More From the Premier 100 Conference:
Full coverage overview
Latest conference headlines
Complete preconference survey results (registration required)
Conference blog

From CIO: 8 Free Online Courses to Grow Your Tech Skills
You Might Like
Join the discussion
Be the first to comment on this article. Our Commenting Policies