In your framework, barriers to entry are the only real source of competitive advantages for incumbent firms in a given industry. Can you explain the different types of incumbent competitive advantages?
Well, you’ve got demand and you’ve got supply. You can have a higher demand curve than a new competitor, which means that you have access to customers that the entrant can’t match. That’s customer captivity. The other side is, you can have a lower cost structure, and proprietary technology is the chief form of that — you’ve got a technology or you’ve got a cost structure that the entrant just can’t match. You would think that those would be the only two advantages, but in fact there’s another way that you can have lower costs, which is that you can have a scale that the entrant can’t match, and there may be either demand or cost advantages associated with scale. It turns out that those are the most powerful sources of competitive advantage.
Can you talk about why competitive advantages are easier to sustain in localized markets?
The obvious thing is that sustainable competitive advantages have to be competitive advantages that apply not just to existing customers, because they die off sooner or later. Teenage customers grow up and you’ve got new customers who are not attached to any firm. So customer captivity goes away sooner or later. Proprietary technologies go away too, because technologies die. And if they don’t die, they just become common property, and so nobody has an advantage after the technology diffuses. So proprietary technology by itself can’t be a truly sustainable competitive advantage.
What you need is something that’s an advantage in the market for new customers who are in the market for new technologies. That is what economies of scale do. When Intel goes after the next generation chip, because it’s got some degree of customer captivity — which is crucial to scale advantages — Intel can expect, if it’s successful, to get 10 times as many customers as AMD. That means Intel can spend 10 times as much on developing and marketing the new chip. That’s the advantage of scale. So who’s going to win that race every time? Intel.
Similarly, in software Microsoft is the company that’s most likely to add new features, so people are going to continue to learn how to use its programs. Microsoft’s costs, because it’s spreading them over 100 million customers and everybody else is spreading their costs over maybe 10 million, are massively lower than everybody else’s. So again, in the pursuit of new technology, Microsoft does better.
For Coca-Cola, it’s got an infrastructure with real economies of scale in any region in distribution and advertising, so when it incurs the fixed cost of advertising, it can expect to get more customers and conserve them at a lower cost. So it has an advantage in the pursuit of new customers.
Economies of scale are really the key to truly sustainable competitive advantages. It’s easier to be the dominant competitor in a market where you’ve got 60 percent of the market and there’s not enough left for everybody else. But if the market gets to be so big that you’ve got 10 percent and you’re the biggest guy in it, there’s plenty of room for everybody else, so you’re just not going to dominate big global markets where players have 7, 8, 9 percent of the market. There are going to be no elephants in those markets.
So where can you be an elephant? The answer is, in a local market in product space. The best examples of that are Intel and Microsoft. By contrast, Apple decided it was going to dominate personal computers, which is a huge market, so when Apple tried to go after the whole market, the market was essentially impossible to dominate. Microsoft started out with just operating systems, which is a much smaller market, and it was able to dominate that. Intel started out trying to do all chips, and then in the mid-’80s it said, “Look, we can’t compete in memory chips; let’s only concentrate on CPU chips.” So both companies concentrated on narrow niches, and then they basically expanded those niches.
As Microsoft moves away from its core strengths into other areas like search, games, etc., do you expect to see its profits decline?
This is where it’s useful to talk about Wal-Mart. Wal-Mart did exactly the same thing, not in product space but in geographic space. It started out dominating one region, and then it moved on in sort of an inkblot way to dominate adjacent regions where its advantages from its economies of scale in the core region continued to apply. To the extent that Microsoft concentrates on things like word processing or Excel that are intricately connected to its operating systems and are incremental additions there, it’s likely to do extremely well. When it decides to go off and do Xbox, which has nothing to do with its core area, it’s likely to get in trouble.
I think the answer to that question is that the inkblot strategy that Wal-Mart pursued initially, and no longer really has the discipline to pursue, is equally applicable in product space. Look at Cisco. It dominated intranets and then it decided to do telecommunications. Well, there were big, powerful companies already there. Similarly, when Wal-Mart tried to move into the West Coast, where Target is powerful, Wal-Mart was competing at a disadvantage.
If Microsoft tries to do something like that, it’s going to get in trouble. But if you actually look at the history of search engines, there’s been very little stability in terms of the dominant competitor. The earliest search engine was Netscape, and then it was Yahoo! and Ask Jeeves and then it was AltaVista for a while and then it became Google. Now, maybe people will all use Google, but as that market gets to be very big and the cost of search engines gets to be quite small, Google will not be able to dominate the market. People tend not to be tied to an individual search engine, so there isn’t that much customer captivity. The cost of producing a search engine is quite small, so a company can be viable with 2 percent of the market. And since all the revenue is from advertising, and advertising can be priced per impression, there’s going to be a lot of natural competition in specialized search engines. So Google is going to be much harder to protect because the fixed costs are small relative to the size of the market.
In markets with no barriers to entry, you tell firms to forget about strategy and focus on internal efficiency. If you look at the roster of Fortune 500 companies, can a company make it to that level based on operating efficiencies alone?
No. Some of them do very well, but I think it’s extremely rare, because everybody else just has to copy them. That was the problem with the Japanese, of course, that they were extremely efficient. They were entirely focused in industries like global automobiles and machine tools with no competitive advantages, and guess what happened?
Let’s talk about Caterpillar or Deere. Tractors can be made anyplace. But the thing about tractors is they break down, and the servicing is an inherent part of the product. So if you look at Deere or Caterpillar, which look like massive manufacturing companies, I’m sure they make almost all their money from their dealership and support networks. And there you’ve got local economies of scale, because you have to have an infrastructure in place, and if you have 70 percent of the tractor market in New Hampshire, then you’re going to be offering much better service. So there are lots of products where there are no global economies of scale — and therefore there are very few companies like Microsoft or Coca-Cola that have economies of scale everyplace — but almost every company has some area in which it is dominant. And I think Caterpillar and Deere are good examples of that, because the dealership networks give them a competitive advantage.
In automobiles in the United States, nobody has any advantage in manufacturing. But if an auto company had a regional franchise, where it had more dealerships and support and better knowledge of credit ratings in the area — because financial institutions have local information — it could have a competitive advantage. Unfortunately, the auto companies are now in a situation where there are dealerships everyplace and nobody has a competitive advantage. But you can imagine that if they’d been more sensible, they could have sustained those competitive advantages.
So I think in a way it’s the wrong question. You don’t want to think, does this company have a competitive advantage or not? The question is, where does it have competitive advantages, and where does it not have competitive advantages? And that’s why we say that strategy is local, because you’ve got to work within the company.
You mentioned the importance of services. What impact will the continued growth of the service sector have on corporate strategy?
I think it’s going to make local strategy much more important. Look at something like HMOs, where Aetna tried to be everyplace, and they’re much less profitable than a very focused company like Oxford, which dominated the New York metropolitan area. I think that services are inherently provided locally, and within those markets you inherently have more opportunity to develop economies of scale, broadly construed.
I’ll give you a good example. Airlines, for some reason, have never been successful at cooperating. But if you look at origin-destination pairs, there are very often just two airlines serving any particular number of origin-destination pairs. And there are economies of scale, because you want to call Delta first if you’re going to Atlanta, and so on. If they could only have the discipline to learn to get out of each other’s way and each focus on their own little niche and cooperate within that niche, they would do much better.
So I think the answer is that when you deal with these essentially geographically localized services, the importance of strategy is actually going to go up. That’s really counterintuitive, because people have always thought of strategy as something that the big global companies pursue, and it’s not at all, because those are actually the markets, like automobiles, where nobody is going to dominate anything, because there are no barriers to entry and they can’t really be erected.
In terms of business strategy, what has been the biggest disaster of the past five years?
It’s the biggest disaster of the past 30 years, and it was AT&T’s determination to get into information processing. The real crimes are when a company tries to get into a market that it’s never going to get into. It assigns marketing people impossible tasks and it invests in facilities that are never going to earn a return and it takes it best managers and, rather than putting them on uses where they’re really going to make money, it assigns them to these impossible tasks.
AT&T had an enormously profitable core business, the long-distance business. There were big economies of scale that it could have used to dominate local regions. It could have done a lot of stuff. But what it couldn’t do is fight at a competitive disadvantage in data processing, where everybody else like IBM had the captive customers and the industry expertise and the software expertise. And then AT&T tried to go up against the local telecom companies, trying to do cable, and so forth. The real crime has been AT&T. The company has disappeared. It went from the preeminent American enterprise to a subsidiary of one of its former units. That’s an extraordinary strategic performance.
Bruce Greenwald is the Robert Heilbrunn Professor of Finance and Asset Management and director of the Heilbrunn Center for Graham & Dodd Investing at Columbia Business School.
Read the Research
"Competition Demystified: A Radically Simplified Approach to Business Strategy"
"All Strategy Is Local"