Skip to contentSkip to navigation

Yves Doz: The Thought Leader Interview

Multinational companies, says the INSEAD professor, must learn to discover, access, mobilize, and leverage knowledge from across the globe.

(originally published by Booz & Company)

Photograph by Jane Evelyn Atwood
The now-familiar criticism of multinational corporations is that they take too much from the communities in which they operate, in resources and cheap labor, and return too little. While acknowledging the legitimacy of much of the antiglobalization movement’s concerns, Yves Doz offers a different observation: Most multinationals fail to harvest the most precious resources — ideas and innovation — from the far-flung regions in which they operate, and they will suffer for it.

The Timken Professor of Global Technology and Innovation at INSEAD, the international business school with campuses in Singapore and Fontainebleau, France, Professor Doz posits a new type of global corporation attuned to the dynamics of the knowledge economy: the “metanational.” In his latest book, From Global to Metanational: How Companies Win in the Knowledge Economy (Harvard Business School Press, 2001), written with INSEAD colleagues José Santos and Peter Williamson, the metanational is described as “a company that builds a new kind of competitive advantage by discovering, accessing, mobilizing, and leveraging knowledge from many locations around the world.”

Becoming a global company once meant penetrating markets around the world. A traditional multinational develops a standard product for its home market, and then sells, or, in Professor Doz’s words, “projects” that standard around the world. But the demands of the knowledge economy are turning this strategy on its head, he says. Today, the challenge is to innovate by learning from the world. And because innovation drives growth, those companies that fail to learn will be left behind, he says.

Professor Doz and his colleagues define the metanational by three core capabilities: being the first to identify and capture new knowledge emerging all over the world; mobilizing this globally scattered knowledge to out-innovate competitors; and turning this innovation into value by producing, marketing, and delivering efficiently on a global scale. They say managers can build a metanational advantage for their own organizations by prospecting for and accessing untapped pockets of technology and emerging consumer trends from around the world, leveraging knowledge imprisoned in the multinational’s local subsidiaries, and mobilizing this fragmented knowledge to generate innovations, profits, and shareholder value.

Although he singles out no one company as a perfect metanational, Professor Doz identifies hotspots of learning capability in such innovators as the Nokia Corporation, in cell phones; Shiseido Company Ltd., in perfumes; and STMicroelectronics NV, in semiconductors. (See “STMicroelectronics: The Metaphysics of a Metanational Pioneer,” by Lawrence M. Fisher, s+b, Third Quarter 2002; Click here.) He notes that each of these companies benefited from “being born in the wrong place,’’ that is, outside the recognized centers of critical knowledge for their industries, and so had no choice but to learn from world markets.

Alliances are one tool of the metanational, and “beating the odds of failure in strategic alliances” has long been one of Professor Doz’s key themes. His Strategic Alliances executive seminar at INSEAD has had a waiting list and been standing-room-only for years, and an earlier book, Alliance Advantage: The Art of Creating Value through Partnering (Harvard Business School Press, 1998), written with Gary Hamel, was widely acclaimed as a strong practical treatise on the subject.

At INSEAD’s Fontainebleau campus, Professor Doz is dean of executive education. His executive seminars concentrate on strategic alliances, multinational management, and entrepreneurial innovation and renewal. Professor Doz has also led a cross-disciplinary research effort on technology and innovation and held the position of associate dean for research and development. Besides INSEAD, he has taught at Harvard and at Stanford in the U.S., and at Aoyama Gakuin University in Japan.

Over the years, Professor Doz has advised many major multinationals, including Philips, IBM, and Xerox, as well as such new multinationals as Intel, Nokia, Toyota, and Glaxo Wellcome, on global strategy and organization.

Strategy+business interviewed Professor Doz in Paris.

s+b: The proposal of a new corporate model, the metanational, suggests that the old multinational model has run out of gas or is in some other way obsolete. Why?

DOZ: I think it is increasingly true that if you take the more traditional models of multinationals, they are running slightly out of steam. They have become a bit self-defeating, and I think that there are several reasons for it.

One is that there has been growing knowledge dispersion. There are more and more places around the world where interesting things are happening market-wise, where interesting innovations, technologies, and knowledge are being created; and relatively few multinational companies have recognized this growing knowledge dispersion. Most tend to rely too much on their home base as the fount of knowledge and the source of innovations, and to ignore the potential of tapping the world for pockets of specialized knowledge, and of bringing that knowledge together to fuel innovations. It’s a major missed opportunity.

A second reason is that the people who are criticizing globalization, the people who demonstrate in the streets of Genoa or Seattle, have a point, which is that the idea of globalization was basically specialization in trade based on comparative advantage, where everyone would excel in what he or she does best. But instead, increasingly, global companies are really scanning the world for cheaper and cheaper labor sources on the one side, and on the other looking for new markets wherever they develop, and serving them with a relatively undifferentiated approach from the one they use at home. So in that sense, the people who complain about exclusion have a relatively strong point. This centralized global form of multinational expansion is just not going to be able to work forever. The multinationals are going to run into increasing social tensions and increasing political strife.

s+b: You used the term “self-defeating.” Self-defeating because they ignore this dispersed knowledge?

DOZ: No, self-defeating for two reasons. Partly because, albeit unwillingly, they are to some extent agents of dispersion. If you go back to the classical economists who looked at international trade, the whole idea of globalization was more or less reciprocal, where various locations in a circle of exchanges would provide the others with labor and markets on a relatively stable basis. But what you find in lots of industries are companies that are locked into global races and cannot really do much except basically follow a very migratory, and imitative, search for cheap labor. They move from Malaysia to Mauritius to Madagascar.

On the other hand, there are countries and governments that for the most part haven’t really been very strategic in their dealings with these companies. So they themselves are migrating from industry to industry, as each in turn seeks cheap labor. You don’t see these governments upgrading the local educational institutions, or upgrading the local capabilities and infrastructure, so they can move from shirts to consumer electronics, to specialty electronics, to product development, to science, to whatever. So there is this dangerous situation in which high-value innovative activities remain in the home country and the lower-value activities, mostly manufacturing, migrate from one part of the world to another.

That system might still work if the governments involved were competent in terms of raising the knowledge level while the companies lower the cost level. Then the race to the bottom among companies would be matched by a race to the top among countries or locations.

s+b: Is there a company that serves today as an exemplar of the metanational strategy, or do you have to look to multiple companies for multiple elements of it as you do in From Global to Metanational?

DOZ: If you are asking, is there already an accomplished metanational that we can point to as an exemplary company, I don’t think so, except perhaps Nestlé. STMicroelectronics comes close. I think there are metanational behaviors, activities, and processes present in a number of companies, but what we did in our work was not to report on the best practice in Company X or Company Y. Our approach is really much more to develop a logical argument, and to say, “If you want to learn from the world, then how can you go about it?”

Our argument is that there are three core steps: a step of sensing, which is identifying and accessing existing knowledge; a step we call mobilizing, which is integrating scattered capabilities and emerging market opportunities in new products and services; and then we stress a third step, which is optimizing operations to maximize the return on these new offerings.

Then we use examples from the practices of various companies as a way to illustrate the argument, and also as a way to make it a more tangible process — to say, “There’s something really going on here,” not just a set of ideas in some academics’ brains. Some American companies, like Procter & Gamble, are moving in this direction. A lot of companies are building up metanational capabilities.

We are not saying that metanational capabilities are going to fully substitute for what companies already do. They are going to complement what companies already do. It is not an issue where you can say, “Tomorrow morning the flow of products and components and so on is going to be quite different from what it was yesterday.” The day-to-day operations, the worldwide operation or integration processes, all these things will remain the same. But the way you fuel and renew innovation is going to depend increasingly on your ability to accept knowledge that lies in different places, not just in your home base.

So our goal is not to suggest an entirely different type of company. Our goal is to find companies that do many of the usual things, yet nonetheless fuel and renew and learn in a more interesting and richer way, around new knowledge creation and innovative integration, than most companies do.

The mind-set we see now is almost turning the old slogan of “think global, act local” on its head. We are saying you need to think local and act global. In other words, you need to constantly ask yourself, “What can I draw from a particular local environment that is unique and different, that is going to best make use of its capabilities and competencies, that is going to best leverage this uniqueness on a worldwide basis?” So the approach is very much to think about the local uniqueness, think about the local rootedness, and think about how you can then take these things to a global level.

The role of management in metanational companies is to identify and connect the best from the various locations in which they are involved. Thinking local means saying, “We are going to leverage our local capabilities globally, in a much more effective way.” That’s not the same thing as leveraging your global capabilities locally. It is a change of mind-set, a change of attitude, and a change of perspective in how you look at the world. But it is not a fundamental change in the day-to-day operations of sales forces, manufacturing plants, or even traditional R&D labs.

It comes down to this: If you want to compete successfully, there is an unprecedented opportunity to pick up under-exploited, poorly understood innovative knowledge in all kinds of places. But you have to be looking for it, and you have to capture it, and then you have to make sure that it’s used.

Making Knowledge Useful

s+b: You describe the metanational as an organization designed to prospect for, sense, and leverage dispersed knowledge. Many companies spent the ’90s buying knowledge systems, and adopting processes intended to make them into “learning companies.” Why has the return on those investments been so limited?

DOZ: Return is a complex term, and although it is something we have researched, I’m not sure we can measure the return on knowledge systems investment or be very conclusive about that. But I do think that most so-called knowledge management systems act like the Yellow Pages. They have been good at essentially two things: locating sources of knowledge internally, and tagging and cataloging existing knowledge in the company. But they haven’t been designed that well for prospecting and accessing knowledge outside the company. Most are fairly inward-looking, which is good for some consulting companies that have a lot of in-house knowledge, but isn’t necessarily as good for many other companies that don’t.

Also, once you have identified the locations of various types of knowledge, or the sources of knowledge in the company, how do you actually connect them effectively? That goes well beyond the point most knowledge systems have reached. They can handle only explicit, articulated knowledge. The ability to move from finding out where the knowledge is, to making that knowledge useful for your own purposes and in your own location, is a big jump.

In the book we describe four layers of knowledge complexity. Essentially, knowledge systems have been good at managing what we call “simple knowledge.” By “simple” we mean knowledge that can be made explicit, that can be codified in a reasonably universal way, and not an ambiguous way. But we need to become able to share another knowledge level: “complex knowledge,” which is tacit, context-dependent, or both. That becomes a lot more difficult to achieve. Most companies haven’t really learned how to manage and integrate complex knowledge. Complex knowledge is hard to learn and hard to understand unless you learn it in context, in the place where the knowledge exists. Yet learning complex knowledge from a distance is perhaps the most daunting challenge multinationals face today.

s+b: Learn by doing, in other words.

DOZ: Yes, if you learn by doing in the right place. In the book we use the example of Shiseido, which is not only a technology company, but now also a fragrance company, created here in Paris from scratch, with several new luxury brands. Shiseido, which is Japanese, plugged itself into France to gain knowledge about how to develop and market perfumes. Shiseido went to great lengths — by acquiring luxury beauty parlors to give it access to elite customers, by creating a French joint venture, and by collaborating with Issey Miyake and Jean-Paul Gaultier, two top designers. It also opened its own high-end beauty salon in Paris. The company learned by close observation and interaction with customers and collaborators in these various operations. This is, I believe, the key to acquiring complex knowledge.

The point is that easily codified knowledge, the kind that knowledge systems manage pretty well, is probably the least deeply interesting knowledge, because it is not likely to provide very sustainable competitive differentiation. Learning by doing indeed leads to original operating knowledge, of a very valuable, hard-to-imitate type.

s+b: Do you see a role for someone like a chief knowledge officer in a metanational?

DOZ: Yes, depending on how you define the role. I think part of the problem of so-called CKOs is the title. I can see very clearly a role for a CIO, someone who manages information systems, information tools and libraries, of various types. You are basically talking about someone who worries about intelligent idea infrastructure.

But I can also see a role for someone who could be more externally oriented — learning, scouting — and much less oriented toward structures, who works less inside the organization. This would be a bit more of a roving ambassador, or an explorer. This is really a discovery function, and I would have liked the so-called CKOs to move in that direction, as lead explorers. But what I’ve seen is closer to managing internal processes and internal infrastructures, and making sure that the e-mail works right, or at best worrying about sharing best practices, which I think is really a missed opportunity. That’s the world as we know it already. The priority should be to explore the world, and find out what is going on that we don’t know yet.

s+b: That’s a much tougher job description to write, a much tougher person to recruit. What would that person’s CV look like?

DOZ: I know, it’s a difficult assignment. You need someone who is partly a kind of salesperson, someone who is curious, willing to rub shoulders with customers, with partners, competitors, and so on, and therefore has this sense of being able to open doors. But the important knowledge doesn’t come to you very easily, so at the same time this person must be smart and strategic enough to not just go in with the view of what kind of gizmo he or she can sell you next, but what can be learned from you. How can you make me aware of things I am not yet sensitive to, but which I should be familiar with? Typically, people who have managed distant labs and companies have this characteristic of curiosity toward the new and unknown, and are used to operating in contexts they can’t control.

Companies need to be very sensitive to the fact that these people need to be treated as explorers, that on their own terms, they are almost the equivalent of adventurers. From that standpoint they need to be managed, financed, and rewarded according to a logic that measures something very different from higher sales or greater internal efficiency.

Learning from Many Markets

s+b: You use the term “projection” for what multinationals do, implying that they push their products and processes into global markets, whereas metanationals pull ideas from dispersed sources. How does the metanational translate that knowledge into a coherent go-to-market strategy?

DOZ: First of all, there is one goal that is perhaps obvious, which is that a lot of what we call learning and sensing is about capturing market knowledge — understanding different markets and different customers. It’s very much concerned with perceiving emerging trends in consumer behavior. A lot of it may end up in product development and distribution strategy and so on, but in any case the knowledge has a market dividend. So the more market dividends you capture to begin with, the less you run the risk of ending up with the proverbial horse designed by committee, the risk of producing something that is not going to sell anywhere.

The other important point is that we need some form of market discipline fairly early in the process. For instance, if you take this approach to new projects, then it’s easier if you find the customer before you determine that the idea is a real project. Essentially, the idea for industrial product or service suppliers is to team up with customers or with one large customer. That way you don’t try to sustain something and push something that doesn’t really have a market proposition attached to it. You need early market tests. Then the challenge becomes making those market tests, and the market knowledge gained in them, applicable elsewhere.

The danger of the projection model is that domestic companies, which are sensitive only to the domestic market, fail to recognize two things: They fail to recognize the need for differences in different markets around the world; and worse, they fail to recognize the fact that at some point in time the domestic market may stop being a lead market. And they become very hardwired to very detailed knowledge that they get from their domestic market.

Look at a company like Kao, another Japanese beauty-product company, which has tremendous capability in Japan in terms of market sensitivity, through its distribution channels and all kinds of intermediaries. It has tremendous infrastructure for market sensing, for understanding market trends, like nerves extending down to Japanese villages, and it’s fantastic. But you can become a hostage of your home market and the richness of the understanding you’ve developed there, because unless you have something comparable elsewhere, you are bound to keep overrelying on your domestic market’s feedback to develop new products and services. And you may fail to realize that it is no longer the center of the world.

That’s what happened to Motorola, which went through that very problem of having been the absolute leader in radio communications, mobile communications, and cellular telephones, and then failed to see early enough that the game was shifting from technology leadership to fashion accessory design, marketing, and distribution. Nokia managed to understand this much faster and much sooner, partly because they weren’t at the center of the universe and they knew they had to learn from markets far away from Finland.

s+b: In Alliance Advantage, your book with Gary Hamel, you wrote about how companies can add value through partnering. Many companies have created a global network of alliances as a way to gain capabilities or knowledge without the more disruptive and costly steps of a merger or acquisition. Does this strategy need to be modified for the metanational?

DOZ: No, actually I think learning alliances can be a key part of a metanational strategy. Companies need to say, “How do I access knowledge that I haven’t mastered yet, that I don’t really understand yet, that can be learned in the practical work of an alliance?” Alliances are very powerful metanational learning tools, provided we forge alliances very much with a learning mind-set and a learning agenda, and turn that agenda into some pretty down-to-earth, concrete learning steps.

The French–Italian chip company STMicroelectronics is an example of a company with all types of alliances with customers, which it uses very actively for learning. In another domain, there is a Portuguese company, one of the largest Portuguese groups, called Sonae, which does things like supermarkets and shopping mall development. That’s not very high-tech global stuff, but they have a very clearly articulated strategy of using alliances. It’s a way for them, first, to learn new businesses, and second, to acquire new knowledge from different national communities, in order to expand their existing businesses internationally.

s+b: Since last September 11, senior executives are thinking about security at a strategic level. Since a network of alliance partners is only as secure as the weakest partner, whether it’s vulnerable to physical attack or cyber attack, do we need a new level of due diligence in forming alliances?

DOZ: I think that is actually a good point. It depends on the field of activity of the alliance. You can argue that if it’s focused on running the airlines or integrated logistics, then yes, by all means. If it’s running advanced product development on a worldwide basis, one should probably say yes, although you can argue that there you could essentially save, store, reprocess, and create backups within the system. That way, if a lab is blown up, then you still have a real problem, but if nothing more than some of your data is lost, you can reconstitute it and you are not going to lose too much time. It depends quite a lot on the nature of the activity. Anything that is real-time, worldwide, service oriented, data intensive, information intensive, or is a security-related or national defense–related type of business, by all means factor that into due diligence.

s+b: Even before the September 11 attacks, some multinationals were finding that local companies, whether in South America, South Africa, or Asia, were far tougher competitors than they had anticipated, and that many of their presumed advantages did not pan out in these markets. Is this the limitation of the projection model?

DOZ: Yes, and there is one point that you picked up on earlier, which is that if your market intelligence and your market understanding is overly dependent on the home base, then it becomes very difficult to be sensitive to market differences. I would argue that it is not just a matter of tough local competitors, but that there may be very contextual knowledge involved, rich knowledge that can be acquired only by the experience of investing in a company.

SAB, South African Breweries, has been expanding in all kinds of lesser-developed markets quite successfully because it has acquired local expertise by dealing with underdeveloped infrastructures in South Africa and adjacent markets. Most Western beer suppliers, except perhaps Heineken, with its rich experience in the Congo, have had a tough time competing against SAB in developing country markets. That is a kind of expertise that most breweries, except perhaps companies like Heineken, don’t really have. There is a lot of latent expertise, passive knowledge, that is embedded there.

We’re touching here on one of the limitations of the projection model. Companies that are too deep in a projection mode just leave aside, and ignore, too much available knowledge from peripheral locations. Some of that knowledge involves ideas for new products, some of that knowledge refers to new kinds of competencies, some of it is awareness and understanding of process features.

These are all arguments that say, “Don’t overrely on the home base, and don’t ignore all the pockets of knowledge that are available around the world, just because they may be a little more difficult to access than what is next door to you.” The weakness of the projection model is not just facing tough local competitors — it is leaving pockets of knowledge under-exploited, when in fact you could access them on a relatively inexpensive basis and probably extract a lot of value from them. To the metanational companies, the world is a canvas dotted with pockets of expertise that are waiting to be brought together.

Future of Clusters

s+b: As someone who grew up, literally, in Silicon Valley, I find some of your assertions about the sources of innovation difficult to accept. Will not such clusters, with their rich academic ties, venture capital, and the virtuous cycle of concentrated and highly mobile talent, remain vital sources of innovation for decades to come?

DOZ: We are certainly not making the point that we are going to see the demise of Silicon Valley or the demise of geographical capability clusters. The point we are making is we are going to see the multiplication of such clusters with perhaps more original, more differentiated capabilities. In other words, more narrow, but more intense capabilities, as you see now with software in India, as you see around biotechnology in Cambridge in the U.K. In a sense you don’t have the same richness and the same diversity as you get in Silicon Valley, but you get actually quite a lot of expertise, and probably more depth and greater focus. So our argument is not to say that clusters are about to disappear.

Our argument is really twofold: Clusters are about to multiply, and they will develop and emerge, or in some cases be created by government policies or encouraged by government policies, in more diverse locations around the world.

Point number two is that a company that learns how to connect the knowledge found within those various clusters in a smarter, faster, and more efficient way than other companies is going to gain a tremendous advantage, way beyond what would be available to companies that rely only on a single cluster.

So our message to Silicon Valley companies is not, “You guys should move out of there.” Not at all. The message is really much more, “We know that Silicon Valley is extremely strong, but don’t be too provincial. Look at the rest of the world as well, and think about what you can learn and bring home from the rest of the world.”

s+b: What of the classic research centers, like Xerox PARC and Bell Labs? Does the failure to thrive of Xerox and Lucent suggest that such centers are relics?

DOZ: My gut feeling is, “I hope not.” The real issues are twofold. One is that some of these centers, like PARC and Bell Labs, have become kind of gated communities, isolated islands, looking pretty much inward, and not always having the richness of external contacts they would benefit from. On the other hand, one of the things we certainly do not advocate, and that has been done by some companies, is to set up relatively large research establishments overseas. They then find that the local companies are already connected to the local communities and have first pick of the best local students, the best local researchers, and so on. That’s what happened to a number of Western companies that opened research centers or R&D centers in Japan, only to be disappointed.

And then the second problem, if you look at the history of Xerox and Lucent, is that unless you also work very actively on the absorptive capacity of the rest of the organization — its ability to put to good use what is developed in the labs — then someone else might benefit instead. Which is exactly what happened with both of these centers. Big companies have been fairly reluctant, and in many cases not so much reluctant as unable, to absorb the new knowledge created in their research centers, which at the end of the day leads entrepreneurial companies like Apple in the 1980s to basically get the benefits. But I don’t think one should draw the conclusion that this is the problem of the labs. To my mind that is much more the problem of the rest of the organization. And whether you look at Xerox, Lucent, IBM, and so on, I think it would be very unfortunate if at the end of the day that knowledge got dispersed or got lost.

s+b: If Nokia or ST’s successes suggest that the next metanational superstar could as easily emerge from Iceland as Palo Alto, why have European startups, like Business Objects in software or SangStat in biotech, included a Silicon Valley base and a Nasdaq listing as integral parts of their strategy?

DOZ: I think we need to separate the two sides of the question. Is the Nasdaq listing going to remain essential to companies? The answer is yes, for all kinds of reasons that have to do with venture capital funding, standards of reporting, the way you look at stock valuations, and so forth. So I would expect a lot of companies started by Europeans to essentially be transformed into American companies, in order to have a Nasdaq listing and in order to have U.S. appeal.

The second point is a different one. If you are in a catch-up mode, like many companies have been, starting with big companies in Korea and small European companies like Business Objects, and you basically say, “I need to learn from the leaders, and I know where the leaders are,” then you want to go and set up more than just a listening post. You want to set up a substantial operation, and I think you will continue to see a lot of that taking place.

The more intriguing question is, if you are no longer in the catch-up mode, or if you start with a bright, innovative idea that really puts you ahead of others, then where do you go? And I think that’s the challenge increasingly for a lot of companies. Then you have to start thinking about convergence of technologies, convergence of skills, migration of skilled populations — and it is a much more complex exercise. It’s an exercise that is never going to turn out 100 percent right, so at best you can try to identify which are the best territories. California is likely to remain one of these locations, but it will be one location among others.

Reprint No. 02408

Lawrence M. Fisher,, covered technology for the New York Times for 15 years and has written for dozens of other publications. Mr. Fisher, who is based in San Francisco, is a recipient of the Hearst Award for investigative journalism.
Get s+b's award-winning newsletter delivered to your inbox. Sign up No, thanks
Illustration of flying birds delivering information
Get the newsletter

Sign up now to get our top insights on business strategy and management trends, delivered straight to your inbox twice a week.