1

                              

Semiglobalization and Strategy

The globalization of markets is at hand. With that, the multinational commercial world nears its end, and so does the multinational corporation . . . The multinational corporation operates in a number of countries, and adjusts its products and processes in each, at high relative cost. The global corporation operates with resolute constancy . . . it sells the same things in the same way everywhere.

—Ted Levitt, “The Globalization of Markets,” 1983

A QUARTER OF A CENTURY after Ted Levitt’s bold pronouncements, excitement about the globalization of markets has given way to excitement about the globalization of production.¹ But what has remained constant is the vision of a globalization apocalypse, sweeping all before it. And this apocalyptic vision leads to a focus on strategies for a post-apocalyptic, integrated world—strategies that inevitably have a one-size-fits-all character. That is why Levitt’s definition of global strategy as a strategy for an integrated world still reigns.²

And, with apologies to my late colleague at the Harvard Business School, that definition is still as wrongheaded. In this book, I redefine global strategy to describe a broader set of strategic possibilities. I argue that differences between countries are larger than generally acknowledged. As a result, strategies that presume complete global integration tend to place far too much emphasis on international standardization and scalar expansion. While it is, of course, important to take advantage of similarities across borders, it is also critical to address differences. In the near and medium term, effective cross-border strategies will reckon with both, that is, with the reality that I call semiglobalization. The primary goal of this book is to stretch our thinking about cross-border strategies for a semiglobalized world.

This chapter begins by establishing that semiglobalization is, in fact, the real state of the world today—and tomorrow. It does so by taking some data on board, because, as the late Daniel Patrick Moynihan observed, we are all entitled to our own opinions, but not our own facts. The chapter then starts to address the implications for company strategy, using the example of one of the great border-crossing companies, Coca-Cola. Around the time that Levitt’s article appeared, Coke embarked on a global strategy of the sort that he recommended. The problems with that strategy took a while to surface, but by the millennium, Coke was adrift in a sea of troubles. Only recently has it started to regain its bearings. Other companies can either learn from Coke’s experience or rediscover the same lessons about semiglobalization the hard way, through trial and error.

Apocalypse Now?

According to the Library of Congress catalog, we are positively awash in books on globalization. More than five thousand such books were published between 2000 and 2004, compared with fewer than five hundred in the whole of the 1990s. In fact, between the mid-1990s and 2003, the rate of increase in globalization-related titles—more than doubling every eighteen months—surpassed the celebrated Moore’s Law!

Amid all this clutter, the books on globalization that have managed to attract significant attention have done so by painting visions of a “globalization apocalypse.” These volumes tend to exhibit what scholars cite as general characteristics of apocalyptic argumentation: emotional rather than cerebral appeals, reliance on prophecy, semiotic arousal (i.e., treating everything as a sign), an emphasis on creating “new” people, and, perhaps above all, a clamor for attention.³ The Flattening of the Earth is the globalization apocalypse that occupies center stage as of this writing.⁴ Thus, during a recent TV interview, the first question I was asked—quite earnestly—was why I still thought the world was round!⁵ But other visions of the globalization apocalypse have been propounded as well: the Death of Distance, the End of History, or Levitt’s own favorite, the Convergence of Tastes. Some writers in this vein view the apocalypse as a good thing—an escape from the ancient tribal rifts that have divided humans, or an opportunity to sell the same thing to everyone on earth. Others see it as a bad thing: a process that will lead to everyone eating the same fast food. But they all tend to assume (or predict) nearly complete internationalization.

This is where I disagree strenuously, but on the basis of data rather than opinion. Most types of economic activity that can be conducted either within or across borders are still quite localized by country.

Ask yourself, for example, how large total foreign direct investment (FDI) flows are in relation to gross global fixed capital formation. (To put it another way, of all the capital being invested around the world, how much is being done by companies outside their home countries?) Maybe you’ve heard the rhetoric about “investment knowing no boundaries,” and so on. The fact is, the ratio of FDI to overall fixed capital formation has been less than 10 percent for each of the last three years for which data are available (2003–2005). In other words, FDI accounts for less than a dime out of every capital dollar invested—or significantly less if one recognizes that much of FDI involves mergers and acquisitions, that is, investment that doesn’t actually generate incremental capital expenditures. And although merger waves can push the ratio of FDI to gross fixed capital formation to higher than 10 percent, the ratio has never quite reached 20 percent.⁶

FDI isn’t an isolated, unrepresentative example. Figure 1-1 summarizes data on internationalization along ten dimensions. As you can see, the levels of internationalization along these dimensions all cluster much closer to 10 percent—which also happens to be the average across the ten categories—than to 100 percent.⁷ The biggest exception in absolute terms—the trade-to-GDP ratio shown at the bottom of the figure—probably recedes most of the way back toward 20 percent if you adjust for double-counting.⁸ So if I had to guess the internationalization level of some activity about which I had no particular information, I would guess it to be much closer to 10 percent than to 100 percent! I call this the “10 percent presumption.”

The 10 percent presumption notwithstanding, I prefer to talk of semiglobalization rather than “deciglobalization.” One reason is that 10 percent is not meant to be any kind of global constant: my best guess is that the next few decades will witness increased internationalization of many of the categories in figure 1-1, and a (slow) upward drift in their average. Second, if internationalization levels are setting new records in many respects, international activity probably warrants a share of attention that exceeds its current share of total economic activity—it is increasingly important and its surge is taking it into uncharted territory. Third, business interest in internationalization may also exceed general internationalization levels because businesses have some distinct advantages—as well as disadvantages—compared with other channels for cross-border coordination. Thus, the largest companies are significantly more internationalized than the 10 percent level: the one hundred largest nonfinancial corporations, for example, have, on average, one-half their sales, assets, and employment overseas.⁹ And many smaller companies aspire to increase their internationalization levels.

So the point of the data presented in figure 1-1—and other data on cross-border market integration that I discuss at much greater length and more systematically in my published academic research—is not that we should neglect cross-border issues, but that we should see them from a semiglobalized perspective.¹⁰ From this perspective, the most astonishing aspect of various announcements of the globalization apocalypse is the extent of exaggeration involved.

Apocalypse in the Near Term?

There is an obvious rejoinder available to apocalyptics: the assertion that even if the world isn’t quite flat today, it will be tomorrow.¹¹

To deal with such an assertion, we have to look at trends rather than at levels of integration at one point in time. The results are interesting. Along a few dimensions, integration reached its all-time high many years ago. For example, rough calculations suggest that the fraction of the world’s population accounted for by long-term international immigrants was slightly higher in 1900—the high-water mark of an earlier era of migration—than in 2005.¹²

Along other dimensions, new records are being set. But this has happened only relatively recently, and only after long periods of stagnation and reversal. For example, FDI stocks as a ratio of GDP peaked before World War I and didn’t return to that level until the 1990s. In fact, some economists have argued that the most remarkable development over the last few centuries was the declining level of internationalization between the two world wars, of which FDI is a particularly striking illustration.¹³

And finally, there are dimensions along which pre–World War I levels of integration were matched and surpassed relatively soon after World War II. International trade in relation to GDP is the leading example: it surpassed pre–World War I records during the 1960s, reached the 20 percent level for the first time in 1979, and, over the next twenty-five years, increased to 27 percent. Extrapolating this rate of increase would imply a trade-to-GDP ratio of less than 35 percent by 2030. Unprecedented yes, but hardly apocalyptic.¹⁴

It is useful to supplement such extrapolations with some consideration of the forces behind the trends. Consider the two drivers of cross-border integration that have been emphasized the most by apocalyptics:¹⁵

• Technological improvements, particularly in communications technologies

• Policy changes that have engaged more countries in the world economy

The question we have to ask is this: Do these two important forces really push us toward an integrated world in the near term?

Improving Communications Technologies

Technological improvements seem to be the most frequently cited drivers of the alleged globalization apocalypse.¹⁶ Given their rate of improvement over the last century, transportation and, particularly, communications technologies have attracted the most attention. For example, the cost of a three-minute telephone call from New York to London fell from $350 in 1930 to about 40 cents in 1999 and is now approaching zero for voice-over-Internet telephony. And the Internet itself is just one of many newer forms of connectivity—enabled by digitalization and the convergence of communications and computing—that have progressed several times faster than plain old telephone service. This pace of improvement has inspired many apocalyptic announcements, including this one, from one of the better books of this sort, Frances Cairncross’s Death of Distance:

New ideas will spread faster, leaping borders. Poor countries will have immediate access to information that was once restricted to the industrial world and traveled only slowly, if at all, beyond it. Entire electorates will learn things that once only a few bureaucrats knew. Small companies will offer services that previously only giants could provide. In all these ways, the communications revolution is profoundly democratic and liberating, leveling the imbalance between large and small, rich and poor.¹⁷

There is a kernel of plausibility to some of Cairncross’s ideas. Technologies and standards do enable connectivity and collaboration at a distance, and that is important. It is also likely that, as Cairncross asserts, the separation of where certain services can be performed from where they are delivered will matter a great deal.

Nevertheless, it’s a gross exaggeration to jump from these kernels to proclaiming the “death of distance” based on improving communications technologies. Reconsider the Internet itself. The internationalization of Internet traffic is impossible to measure precisely, particularly because of problems sizing up domestic flows. But the best estimates I have been able to locate indicate an internationalization level a bit below 20 percent, that is, within a factor of two of 10 percent.¹⁸ In terms of changes rather than levels, the international share and, particularly, the intercontinental share of total traffic is supposed to be decreasing rather than increasing, for reasons ranging from surging peer-to-peer traffic to the development of alternatives to the United States, which until recently was the hub for virtually all international switching.

Business-focused examples permit more definite statements based on better data. Look at information technology (IT) services, which are often cited as an illustration of technologically enabled globalization. A total of 2 percent or 11 percent of such work—depending on whether one looks at the total potential market or only the immediately addressable part of it—is currently offshored.¹⁹ Or for an even more net-centric example that helps explain barriers at borders as well as exemplifying their effects, consider Google.

The company boasts of supporting more than one hundred languages and, partly as a result, has recently been rated the top global Web site. But Google’s reach in Russia—cofounder Sergey Brin’s country of origin—was only 28 percent in 2006, versus 64 percent for the market leader in search services, Yandex, and 53 percent for Rambler, two local competitors that account for 91 percent of the Russian market for ads linked to Web searches.²⁰ Google’s problems reflect, in part, linguistic complexities: Russian nouns have three genders and up to six cases, verbs are very irregular, and the meaning of words can depend on their ending or the context. In addition, local competitors have adapted better to the local context by, for example, developing payment mechanisms through traditional banks to compensate for the dearth of credit cards and online payment infrastructure. And though Google has doubled its reach since 2003, this has required setting up a physical presence in Russia and hiring engineers there, underlining the continued importance of physical location.

Google’s highly publicized travails with Chinese censors illustrate a different set of reasons that borders continue to matter: governments have become more adept at creating closed national networks and enforcing local laws (aided, in part, by Internet geographic identification technologies that continue to improve). Nor is it just totalitarian governments that flex their muscles in such ways. Many experts view the success of the French government’s 2000 effort to restrict sales of Nazi memorabilia by Yahoo! as the key legal precedent in this regard. And the intervention that has probably had the biggest economic impact is the U.S. government’s 2006 ban on online gambling.

The implications of all these barriers at the borders for the Internet are discussed at length in a book with the telling subtitle Illusions of a Borderless World, which argues that: “What we once called a global network is becoming a collection of nation-state networks.”²¹ Chapter 2 looks more generally at barriers to cross-border economic activity and collects and classifies them in terms of the CAGE (cultural-administrative-geographic-economic) distance framework for thinking about the differences between countries.

Policy Openings

A second notable driver of cross-border integration is a constellation of policy changes that led many countries—particularly China, India, and the former Soviet Union—to come in from the cold and participate more extensively in the international economy. Economists Jeffrey Sachs and Andrew Warner provide one of the better-researched (although still apocalyptic) descriptions of these policy changes and their implications:

The years between 1970 and 1995, and especially the last decade, have witnessed the most remarkable institutional harmonization and economic integration among nations in world history. While economic integration was increasing throughout the 1970s and 1980s, the extent of integration has come sharply into focus only since the collapse of communism in 1989. In 1995, one dominant global economic system is emerging.²²

Yes, such policy openings are important. But to paint them as a sea change is inaccurate, at best. Remember that integration is still relatively limited. Meanwhile, the policies that we changeable humans enact are surprisingly reversible. Thus, Francis Fukuyama’s End of History, in which liberal democracy and technologically driven capitalism were supposed to have triumphed over other ideologies, seems quite quaint today.²³ Especially in the wake of September 11, 2001, Samuel Huntington’s Clash of Civilizations looks a bit more prescient.²⁴

But even if you stay on the economic plane, as Sachs and Warner mostly do, you quickly see counterevidence to the supposed irreversibility of policy openings. The so-called Washington consensus around market-friendly, open policies ran up against the Asian currency crisis and has since frayed substantially—in the swing toward “neopopulism” across much of Latin America, for example—to the point where we are starting to see working papers with titles such as “Is the Washington Consensus Dead?” In terms of outcomes, the number of countries—in Latin America, coastal Africa, and the former Soviet Union—that have dropped out of the “convergence club” (defined in terms of narrowing productivity and structural gaps vis-à-vis the advanced industrialized countries) is at least as impressive as the number of countries that have joined the club.²⁵ At a multilateral level, the suspension of the Doha round of trade talks in the summer of 2006—prompting The Economist to run a cover titled “The Future of Globalization” and depicting a beached wreck—is not a good omen.²⁶ In addition, the recent wave of cross-border mergers and acquisitions seems to be encountering more protectionism in a broader range of countries than did the previous wave in the late 1990s.

Of course, given that sentiments in this regard have shifted more than once in recent decades, they may yet shift again in the future. Such possible inflection points are discussed further in chapter 8. The point here is not only that it is possible to turn the clock back on globalization-friendly policies, but also that we have a relatively recent example of that actually happening, during the interwar period. In particular, we have to consider the possibility that really deep international economic integration may not mix well with national sovereignty.²⁷

So while the technological drivers of increased cross-border integration may be irreversible, we can’t say the same about policy drivers. Policy drivers are, therefore, an even shakier basis for apocalyptic visions of complete cross-border integration—not to mention strategy making predicated on such visions!

It is interesting to speculate about why people’s beliefs in globalization overshoot the reality of semiglobalization to the extent that they do. Jean de la Fontaine’s aphorism “Everyone believes very easily what they fear or desire” subsumes at least some of the explanations: the paranoia of those who fear world domination by multinationals, the smug superiority of elites who have been variously characterized as Davos men and cosmocrats, the terminal insecurity of those trying to be “with it,” the naive utopianism of internationalists, and so forth. But spending more time on this issue is a little bit like H. L. Mencken’s characterization of going to the zoo: engaging, but not especially productive. So it’s time to turn to the implications for companies and their global strategies, which we will pursue by looking at the rather amazing story of Coke.

The Case of Coke

Even companies with substantial global experience, presence, and success can fall prey to visions of the apocalypse—and thereby place themselves in great peril. A particularly cautionary case is that of Coca-Cola, which has a broader global presence than just about any other company in the world, owns what is reckoned to be the world’s most valuable brand, and is much more profitable overseas than at home. Until the late 1990s, Coke was also regarded as an exemplar of global management. But since then, it has experienced a fall from grace, from which it is still recovering. Consider Coke under successive CEOs.

Background

Founded in 1886, Coke made its first move outside the United States in 1902, when it entered Cuba—the same year that archrival Pepsi-Cola started up business. By 1929, five years before Pepsi set up its first foreign venture (in Canada), Coke was being sold in seventy-six countries around the world. Its international presence was greatly bolstered by World War II, when keeping U.S. troops overseas supplied with the soft drink became government policy. Coke, exempted from wartime sugar rationing, built sixty-three bottling plants around the world. Its global push continued after the war, under Robert Woodruff, who ran the company from the early 1920s to the beginning of the 1980s and was an avowed flag-planter: “In every country in the world, [Coca-]Cola dominates. We feel that we have to plant our flag everywhere, even before the Christians arrive. Cola’s destiny is to inherit the earth.”²⁸

But despite this triumphalism (which one wit referred to as “Coca-Colonization”), Coke’s strategy continued to be “multilocal” over this period. Its local operations were more or less independently managed. Their primary objective was to support a network of over one thousand bottlers, which employed more than fifty times as many people as did the mother ship and actually undertook most of the activities performed by the Coke system.

Roberto Goizueta: Exploiting Similarities

Roberto Goizueta, who took over as Coke’s CEO in 1981, continued Woodruff’s push into international markets but also set out to transform how they were managed. Over the course of his tenure, Coke came to define the aggressively globalized corporation: its strategy reflected Goizueta’s sense that the only fundamental difference between markets in the United States and other countries was the lower average levels of market penetration overseas. As he put it in one speech, “At this point in time in the United States, people consume more soft drinks than any other liquid, including ordinary tap water. If we take full advantage of our opportunities, some day, not too many years into our second century, we will see the same wave catching on in market after market.”²⁹

This core belief in similarities across countries underpinned a global strategy that placed ever more emphasis on international growth, scale economies, statelessness, ubiquity, and centralization with standardization:

• Growth fever: Although U.S. volume growth slowed in the mid-1980s, Goizueta clung to historic targets and placed ever more emphasis on the non-U.S. operations as a way of meeting them. Given a belief in similarities across countries, the rest of the world seemed to be a blue ocean of growth opportunities: in Goizueta’s last year as CEO, for example, Coke sold thirty gallons of soft drinks per capita in the United States (5 percent of the world’s population), versus an average of three and one-half gallons per capita in the rest of the world. Room to grow!

• Economies of scale: Goizueta was also convinced of endless scale economies that would increasingly concentrate market share in Coke’s hands. As he explained to Coke’s bottlers in a speech shortly before his death, “We already have the most popular brand in the world. In fact, we have four of the top five soft drink brands . . . As I see it, that is a giant head start. I cannot think of one business that is in a better position to succeed than ours . . . in a time when trade barriers are tumbling.”³⁰ Again, this element of Goizueta’s strategy went hand in hand with a belief in similarities across countries.

• Statelessness: In 1996, Goizueta declared that “the labels international and domestic, which adequately described our business structure in the past, no longer apply. Today, our company, which just happens to be headquartered in the United States, is truly a global company.”³¹ And he acted on this assertion by officially embedding the U.S. organization in what used to be the international organization—although in practice, the U.S. operation continued to be an entity unto itself. The point is that such a move would make perfect conceptual sense, given a faith in cross-border similarities, because maintaining separate U.S. and international organizations would then be duplicative at best and probably dysfunctional (in the sense of creating unnecessary silos).

• Ubiquity: Goizueta inherited an enterprise that already operated in 160 countries; by the time he departed, that number had nearly reached 200. Some of this expansion—such as into East Europe as the Berlin Wall came down—made clear sense. But other market penetration efforts seem to have been justified on the basis of faith rather than market analysis. Thus, after the Soviet Union left Afghanistan—but as turmoil there continued—Coke successfully raced Pepsi to be the first to bring its soft drinks back to the Afghani market, in 1991.³²

• Centralization and standardization: In pursuit of the objectives described above, Goizueta engaged in an unprecedented amount of centralization and standardization. Divisions were consolidated, and regional groups headquartered in Atlanta. Consumer research, creative services, TV commercials, and most promotions were put under the supervision of Edge Creative, Coke’s internal ad agency, with the idea of standardizing these marketing activities—and the effect of further increasing the head count at headquarters. At the same time, the company designated so-called anchor bottlers, which would often operate in more than one country, and in which Coke would take equity stakes ranging from 20 to 49 percent—leading the company to become more involved internationally in decisions that it had previously delegated to (more) independent bottlers.

The emphasis on centralization and standardization obviously created a bias toward a one-size-fits-all strategy. But few were inclined to argue with it at the time. Coke had been rated the most admired U.S. corporation by Fortune in 1995 and 1996, and would be again in 1997. More objectively, its market value increased from $4 billion to $140 billion over Goizueta’s sixteen-year tenure. Still, these impressive accomplishments reflected Coke’s fundamental strengths and Goizueta’s assiduity in exploiting them (as well as some creative accounting toward the end of this period in buying and selling bottlers), rather than the fundamental soundness of the one-size-fits-all approach. In light of his successors’ travails, that approach turned out to have been grossly overrated.

Douglas Ivester: Staying the Course

When Goizueta died unexpectedly in 1997, he was succeeded by his chief financial officer, Douglas Ivester. The CFO had masterminded the strategy of buying bottlers and reselling them to Coke affiliates, with the gains booked as operating income; this practice helped mask pressures on concentrate profitability. Ivester shared Goizueta’s vision of limitless international growth: his first letter to Coke’s shareholders was titled “A Business in Its Infancy” and contained a section with the heading “Why is a billion [daily servings of Coke] just the beginning? A look at the other 47 billion.”³³ Ivester clung to the other elements of Goizueta’s global strategy as well: when a reporter asked him if Coke would change direction, his response was, “No left turns, no right turns.”

But Ivester’s strategy of staying the course quickly ran into roadblocks, many of them demand related. The world economy began to sag almost on the day he moved into the corner office, with Brazil and Japan—two of Coke’s largest overseas markets—taking economic nosedives. The aforementioned Asian currency crisis intensified with a vengeance in 1998. By 1999, the Russian operations were limping along at 50 percent capacity.³⁴ The same analysts who had previously assigned high values to Coke’s stock because of its global “presence” now marked it down for its global “exposure.”

Ivester discounted the growth shortfalls as short-run setbacks and refused to cut the 7–8 percent volume growth target that had been set—and achieved—under Goizueta, although he did trim the earnings growth target. But by late 1999, Coke’s stock valuation had declined by about $70 billion from its peak as a result of these problems and others, including fraying relationships with governments, particularly in Europe, and with bottlers. Regulators in the European Union resisted Coke’s attempts—directed out of its headquarters—to acquire Orangina and Cadbury Schweppes, and lags in tackling contamination problems in France and Belgium caused further strains. Bottlers had begun to find Coke overbearing as well. They were under profit pressures in many geographies and were particularly embittered by Coke’s attempts to stuff channels as its growth rate came under pressure. The last straw for them was Ivester’s attempt to sustain performance by imposing a 7.6 percent concentrate price increase. They put enormous pressure on Coke’s board, to which they had always had a back channel, to fire Ivester. It did.

Douglas Daft: Succumbing to Differences

Ivester’s successor was Douglas Daft, who had previously headed Coke’s Middle and Far East Group. Daft’s years in the field had imbued him with the belief that the way to win globally was to transfer strategic decision making to local executives. As he put it in January 2000, “No one drinks globally. Local people get thirsty and go to their retailer and buy a locally made Coke.”³⁵ He elaborated on this theme in a March 2000 newspaper article titled “Think Local, Act Local”:

As the century was drawing to a close, the world had changed course, and we had not. The world was demanding greater flexibility, responsiveness, and local sensitivity, while we were further centralizing decision-making and standardizing our practices, moving further away from our traditional multi-local approach . . . If our local colleagues develop an idea or strategy that is the right thing to do locally, and it fits within our fundamental values, policies, and standards of integrity and quality, then they have the authority and responsibility to make it happen.³⁶

This was more than just rhetorical froth for locals to feel good about. Daft made an abrupt about-face in terms of how Coke was to be managed. He ordered six thousand layoffs—most of them at headquarters in Atlanta—and launched a massive reorganization that, among other things, aimed to relocate decision making closer to local markets. Perhaps the single most surprising announcement—which sparked an exodus of top marketing talent—was that no more global advertisements would be made. Instead, ad budgets and creative control were placed in the hands of local executives, who were understandably delighted—but also underprepared. As a result, quality suffered even more than scale economies. A flood of homegrown ads hit the airwaves, ranging from people streaking naked across a beach (in an Italian ad) to an angry grandmother in a wheelchair leaving a family reunion when her granddaughter couldn’t come up with a Coke (in a U.S.-made spot). And the new umbrella themes proved short-lived as well: Enjoy lasted fifteen months, and Life Tastes Good five months (versus Always, which ran from 1993 to 2000).

Given this significant flailing around, it should come as no surprise that volume growth sagged. It averaged only 3.8 percent in 2000 and 2001, versus the 5.2 percent achieved in 1998 and 1999 under Ivester.

At a company that has traditionally cherished growth, this was unacceptable. In March 2002, the Wall Street Journal reported, “The ‘think local, act local’ mantra is gone. Oversight over marketing is returning to Atlanta.” A hundred or so marketers in Atlanta had been reconstituted as the apex of a global marketing group that would set strategy for core brands and agency engagement, develop marketing talent, and help local markets share best practices. But efforts to rebuild headquarters capabilities in this and other functions lagged because hiring and integrating people required much more time than firing them had. Meanwhile, the advertising churn continued. As a result, Coke’s volume growth rate recovered to only 4.7 percent over 2002–2003, well below the long-run target of 5–6 percent (to which Daft had lowered it in 2001), and its stock continued to stagnate. In February 2004, Coke announced Daft’s retirement.

Neville Isdell: Managing Similarities and Differences

Coke looked outside as well as inside for a successor to Daft and ultimately tapped a retired Coke executive, E. Neville Isdell, who signed on in May 2004. The story of Coke under Isdell is still being written, but his moves in his first two years seem consistent with his publicly expressed viewpoint that under his immediate predecessors, the “pendulum swung too far over.” Isdell has turned his back on the extreme localization that Daft initiated by continuing to rebuild headquarters capabilities and recentralize elements of marketing, with a particular emphasis on bigger, more universal advertising themes. Notably, however, this rejection of localization has not been accompanied by a reversion to Goizueta’s and Ivester’s one-size-fits-all approach with its emphasis on extreme standardization:

• Growth fever has receded as a result of Isdell’s reduction of Coke’s long-run volume growth target to 3–4 percent. Stock analysts, who no longer considered Daft’s 5–6 percent target credible, actually reacted positively.

• Economies of scale and selling a few established soda brands are no longer Coke’s primary focus: innovation, particularly in non-carbonated beverages, is.

• Statelessness is out. Early in 2006, Isdell reinstituted a position Goizueta dissolved a decade earlier: a head of all international operations outside North America. The intent was not just to improve possibilities for coordination overseas, but also to be realistic in recognizing the distinctive features and challenges of the home region. This is a far cry from the belief that for a company as global as Coke, there is no meaningful distinction left between home and abroad.

• Ubiquity has not been abandoned, but Isdell’s emphasis on “looking at where we are most profitable and then expand[ing]our offerings there” does point toward more nuanced resource-allocation decisions.

• Centralization and standardization have been moderated. The regional heads have more authority than they did under Goizueta and Ivester, and Coke’s strategies now exhibit more variation at the country level. In China and India, in particular, Coke has lowered price points, reduced costs by indigenizing inputs and modernizing bottling operations, and upgraded logistics and distribution, especially rurally. And as noted, there is more emphasis now on variety.

The last bullet point bears further discussion. What Coke headquarters seems to have realized, in the past few years, is that it may not make sense to compete the same way in all markets.

To be fair, this recognition really dates back to Daft: “I’m not saying that every market will turn into a mirror image of North America or Australia. What consumers in our most promising markets want from us may develop differently—maybe even very differently—than they have in our established markets.”³⁷

Unfortunately for Coke, Daft responded by letting a thousand flowers bloom. But such an approach naturally raises the question of why the whole is more than the sum of its parts. If there are no benefits to be derived from similarities across borders, why are the different country operations part of the same company in the first place?

Under Isdell, in contrast, Coke is trying to leverage ideas that have worked well in one market to rethink how to compete in others—in a way that does afford room for cross-border value addition. Particularly notable in this regard is Coke’s reliance on what it has learned in Japan (see the box “Coke in Japan”) to figure out how to become less cola driven in other markets. This is important in the United States, for example, because obesity has become a major concern, and also in China, where resistance to cola is compounded by an aversion to dark drinks. And in parallel, there is a newfound emphasis on globalizing noncola brands instead of simply treating them as localized add-ons.

In sum, Coke’s strategy under Isdell should be seen as an attempt to find a new and improved way of competing across borders, instead of settling for some compromise between Goizueta’s and Ivester’s extreme centralization and standardization and Daft’s extreme decentralization and localization—that would probably also compromise performance. How well this new strategy will work remains to be seen, but at least Coke is no longer seesawing between those two extremes. Instead, it is attempting to get off the seesaw altogether and compete in a way that neither ignores the differences across countries nor caves in to them entirely—that is, it recognizes the reality of semiglobalization.

Beyond Coke

It is time to extrapolate from the Coke case. This section begins by discussing why other companies may be subject to some of the same gravitational pull of one-size-fits-all strategies that Coke experienced under Goizueta and Ivester. The discussion then takes a cue from Coke’s lurch toward localization under Daft to look at the possibility of protracted weakness after the selection of such strategy, instead of a quick recovery from it. It concludes by using Isdell’s strategy as a springboard to a third way of competing across borders—a way that is more than a halfway house between the extremes of one-size-fits-all globalization and parochial local variation.

Broad Biases

The Coke story, although particularly colorful, is far from unique. Other examples of overstretch and retreat abound. Vodafone, in a faster-moving environment, managed to go through a similar cycle of overstretch-and-retreat in a much shorter time span. While it built up a substantial presence in Japan and the United States as well as its home region of Europe, differences in mobile telephone standards invalidated its attempts to achieve interregional economies of scale. And in the tenth year of the DaimlerChrysler merger, speculation abounds about a demerger. Whatever the eventual outcome, the intended results have clearly not been achieved.

One way of deriving some insight into the incidence of such cases is to consider in a more general context the biases associated with Coke’s failure, under Goizueta and Ivester, to take the differences across countries seriously.

• Growth fever: Even a company as internationalized as Coke averaged nearly ten times the penetration at home as it did overseas. For most companies, the differences between domestic and foreign penetration are even larger! A borderless frame applied to such differences in penetration levels runs an obvious risk of inducing growth fever about foreign markets, especially since most companies tend to cross borders after saturating their home market. To make matters worse, such biases can be exacerbated by advisers (e.g., investment bankers interested in doing deals).³⁸ As I write this, I’m looking at a slide from a major strategy consulting firm that superimposes offerings related to its “global strategy audit” on a stylized globe. On the North Pole is a label that summarizes the overarching objective of such audits: Growth.

• Economies of scale: Coke didn’t pull its obsession with economies of scale out of thin air; it was the logical consequence of failing to take differences across countries seriously. As Bruce Kogut pointed out long ago, in the absence of such differences, the answer to the question of “what is different when we move from a domestic to an international context . . . [is] simply that the world is a bigger place, and hence all economies related to the size of operations are, therefore, affected.”³⁹ There does, in fact, seem to be such an obsession with scale economies and, relatedly, increasing concentration. Thus, surveys that Fariborz Ghadar and I have conducted indicate that more than three-quarters of managers believe that increasing cross-border integration leads to increasing seller concentration—even though the eighteen global or globalizing industries for which we have compiled data show, on average, no such increase.⁴⁰ Incidentally, out of our sample, the soft drinks industry does exhibit the single largest increase in concentration. This suggests that a faith in scale economies would be even more misplaced in other contexts.

• Statelessness: Very few companies have gone as far as Coke did under Goizueta: proclaiming that they have no home base. Many managers do, however, seem to believe that a truly global company should strive to achieve such a state of statelessness. They run the risk of being severely disappointed since foreign companies don’t seem to be able to shake their foreignness (see chapter 2 for more on the liability of foreignness). This is obviously true for such U.S. icons as Coke in parts of the world where the United States is widely hated. But even companies from countries that generally maintain lower international profiles can face problems: consider the boycott of Danish products in the Middle East after a Danish paper ran cartoons of the Prophet Mohammed.

• Ubiquity: Very few companies are as ubiquitous as Coke, but many experience great angst because they are not—and would agree that a truly global company should compete everywhere. This logical consequence of believing in a borderless world seems to be reinforced empirically by an exaggerated conception of the number of countries in which the “typical” multinational operates. Thus, managers seem surprised to learn that U.S. multinationals typically operate in just one or two foreign countries and that for those that operate in just one, there is a 60 percent chance that this country is Canada.⁴¹ And again, managers can get bad advice in this regard; thus, the “global strategy audit” of the leading strategy consulting firm cited above frames global expansion as a question of when, rather than where.

• Centralization and standardization: Finally, if you (as the leader of a company) become convinced that borders don’t matter, you’re most likely to compete internationally the same way that you do at home, for reasons ranging from economies of scale to the sheer difficulty of grasping how different the conditions in foreign countries truly are. The likelihood of such an overemphasis on similarities is reinforced by the observation that firms that are successful at home are disproportionately likely to be the ones that venture abroad and, presumably, to be overly enamored of their own domestic business models. Furthermore, even if such a bias runs up against an unfriendly reality, that may not be enough to overturn it. Coke continued to emphasize centralization and standardization under Goizueta and Ivester, despite pressures for market responsiveness that had forced it to increase its number of brands from a handful in the early 1960s to more than four hundred today, and despite the idiosyncrasy of its most profitable major foreign market, Japan, as described earlier.

So while Coke is clearly unusual along certain dimensions, other more “typical” companies may experience similar biases toward adopting one-size-fits-all strategies. In some cases, they may even be under greater pressure to do so!

Calamitous Consequences

Recall that after Coke got carried away with a one-size-fits-all strategy under Goizueta and Ivester, the pendulum swung too far in the opposite direction during Daft’s first two years as CEO. In other words, not only does it take time to develop a strategy, detect problems with it, and devise an antidote, but all too often, the antidote is an overreaction.

One explanation for this kind of overshoot is emotional. If you get burned by an excessive faith in globalization, “globaloney”—Clare Booth Luce’s original riposte to Wendell Wilkie’s visions of One World more than half a century ago—is perhaps the natural, if irrational, reaction.

Another explanation is political. What happens in the wake of most revolutions? People settle old scores. After the peasants with pitchforks overrun headquarters—one characterization of what happened at Coke under Daft—it is easy to imagine headquarters’ capabilities being dismantled, even if local or regional substitutes aren’t yet in place.

For these reasons and others, many companies get their fingers burned by engaging first in misguided global standardization and then shifting abruptly to a localization strategy. Still other companies throw up their hands and terminate all their international operations. Why? For one reason, they don’t enjoy Coke’s huge border-crossing advantages. Some of these have already been cited: the world’s most valuable brand, relatively standardizable major products, and a consolidating industry. Other advantages include international operations that are more profitable than domestic ones, a broad and balanced geographic presence—Coke is one of only a dozen or so Fortune 500 companies that derive at least 20 percent of their sales from each of the three triad regions of North America, Europe, and Asia-Pacific—and a powerful network of bottlers that provides some counterweight to tendencies toward standardization.

Without these safeguards or strengths, the average border-crossing company can make even bigger mistakes—and be less able to recover from them. To assess your company’s proneness to such mistakes, answer the questions in the box, “Your Company’s Beliefs About Globalization: A Diagnostic.”

Rhetoric and Remedies

The challenges of beating the biased beliefs and avoiding the calamities just described are compounded by confused rhetoric. One vivid illustration is the slogan (mis)appropriated from the environmental movement, “Think global, act local.” This tagline has come to mean such different things to different people that it stands for nothing in particular. Thus, Goizueta pressed it into service to describe the extremely standardized and centralized strategy he adopted at Coke, particularly in regard to marketing. But in “Think Globally, Market Locally,” Orit Gadiesh, the chairman of Bain & Company, encouraged brand managers to “localize, localize, localize”—the exact opposite of Goizueta’s strategy.⁴² So one problem with the “Think global, act local” slogan is that it has been applied to strategic approaches ranging from the most localized to the most standardized, and has thereby been drained of specific content.

A second problem with “Think global, act local” is inherent to the basic idea of framing the challenge of global strategy as striking a balance between the extremes of local customization and global standardization. The problem is that these extremes do not so much span a strategy continuum as constitute two singularities in which cross-border complexities can be finessed and simple single-country approaches applied. To see this, note that if markets were totally segmented from each other, single-country approaches to strategy could presumably be applied country by country, and that if markets were totally integrated, there would be the equivalent of one very big country, and single-country approaches should, once again, suffice. These are not, therefore, the best reference points for a strategy that seeks to take cross-border complexities seriously.

Figure 1-2 reframes that point more positively, to emphasize that the intermediate levels of cross-border integration inherent in semiglobalization are what open up, over a very broad domain, the possibility of global strategy having content distinct from single-country strategy. In other words, the empirical diagnosis of semiglobalization in the first part of this chapter was more important than may have been apparent at the time. Semiglobalization is what enables the development of a distinctively global approach to strategy.

That opportunity, while enticing, isn’t easy. This book takes a layered approach to it, starting out by laying new foundations. Thus, this chapter—on borders mattering for strategy—is followed by one that looks at the substrate in which cross-border economic activity is embedded to understand why borders still matter so much. But to convince possibly skeptical readers that more than abstractions will result—and also because I have critiqued “Think global, act local” for vagueness—here is a preview of some of the specific recommendations from the chapters that follow:

• Determine which of a range of international differences—cultural, administrative, geographic, and economic—are key in your industry, and look for differences in differences: categorize foreign countries into those that are close to your home base along the key dimensions versus those that are far. This is the topic of chapter 2.

• Analyze increasing returns to scale or scope instead of assuming them—or assuming that they are absent—but also go beyond volume, growth, and scale economies to look at all the components of economic value in assessing cross-border alternatives. This is the topic of chapter 3.

• Stretch the responses to differences beyond tweaking the domestic business model—and also consider ways to profit from differences, instead of treating them all as constraints on value creation. The objective, expanded on in chapters 4 through 8, is to foster creativity in thinking about how to compete across borders.

Conclusions

The conclusions from this chapter are summarized in the box “Global Generalizations.” It should now be clear why semiglobalization is more than just a middle-of-the-road conclusion of middling—or even zero—interest about the state of the world. In particular, it is essential to the possibility of global strategy having content distinct from single-country strategy.

It is useful to conclude this chapter by beginning to shift attention from the importance of semiglobalization toward the implications of taking it seriously. Note, in particular, that semiglobalization involves integrated consideration of localized interactions and cross-border interactions—of the barriers and the bridges between countries—instead of a focus on just one or the other. In other words, to take semiglobalization seriously is to infer that business decisions cannot be made on either a country-by-country basis or on the one-size-fits-all-countries basis. What must be grasped, instead, is a business reality that lies in between “one (insular) country” and “one (integrated) world.” This isn’t easy, but the rewards include a richer sense of the strategic possibilities than afforded by the extreme perspectives of zero or complete integration. Semiglobalization can therefore be liberating as well as challenging.