Ten years ago, globalization seemed unstoppable. Today, the picture looks very different. Even Coca-Cola, widely seen as a standard-bearer of global business, has had its doubts about an idea it once took for granted. It was a Coke CEO, the late Roberto Goizueta, who declared in 1996: "The labels 'international' and 'domestic' no longer apply." His globalization program, often summarized under the tagline "think global, act global," had included an unprecedented amount of standardization. By the time he passed away in 1997, Coca-Cola derived 67 percent of its revenues and 77 percent of its profits from outside North America.
But Goizueta's strategy soon ran into trouble, due in large part to the Asian currency crisis. By the end of 1999, when Douglas Daft took the reins, earnings had slumped, and Coke's stock had lost nearly one-third of its peak market value—a loss of about $70 billion. Daft's solution was an aggressive shift in the opposite direction. On taking over, he avowed, "The world in which we operate has changed dramatically, and we must change to succeed. No one drinks globally. Local people get thirsty and buy a locally made Coke."
Unfortunately, "local" didn't seem to be any better a description of Coke's market space than "global." On March 7, 2002, the Asian Wall Street Journal announced: "After two years of lackluster sales the "think local, act local" mantra is gone. Oversight over marketing is returning to Atlanta."
If the business climate can force Coke, which historically was (and is) more profitable internationally than domestically, to seesaw back and forth on globalization in this way, think of the pressures on the typical large company, for which international business is usually much less profitable than domestic business.
Why is globalization proving so hard to get right? The answer is related in part to how companies frame their globalization strategies. In many if not most cases, companies see globalization as a matter of taking a superior (by assumption) business model and extending it geographically, with necessary modifications, to maximize the firm's economies of scale. From this perspective, the key strategic challenge is simply to determine how much to adapt the business model—how much to standardize from country to country versus how much to localize to respond to local differences. Recently, as at Coke, many companies have moved toward more localization and less standardization. But no matter how they balance localization and standardization, all companies that view global strategy in this way focus on similarities across countries, and the potential for the scale economies that such commonalities unlock, as their primary source of added value. Differences from country to country, in contrast, are viewed as obstacles that need to be overcome.
It's possible to apply different strategies to different elements of a business.
Correctly choosing how much to adapt a business model is certainly important for extracting value from international operations. But to focus exclusively on the tension between global scale economies and local considerations is a mistake, for it blinds companies to the very real opportunities they could gain from exploiting differences. Indeed, in their rush to exploit the similarities across borders, multinationals have discounted the original global strategy: arbitrage, the strategy of difference.
Of course, we're all familiar with arbitrage in its traditional, and least-sustainable, form—the pure exploitation of price differentials. But the world is not so homogeneous as to have removed arbitrage from a company's strategic tool kit. In fact, many forms of arbitrage offer relatively sustainable sources of competitive advantage, and as some opportunities for arbitrage disappear, others spring up to take their place. I do not claim that arbitrage to exploit differences is any more a complete strategic solution than the optimal exploitation of scale economies. To the contrary: If they are to get their global strategies right in the long term, many companies will have to find ways to combine the two approaches, despite the very real tensions between them.
Reconciling Difference And Similarity
One would think companies that try to exploit differences would not find it easy to exploit similarities as well. And indeed, a large body of research on the horizontal versus the vertical multinational enterprise has shown that there are fundamental tensions between pursuing scale economies and playing the spreads. (See the table "Conflicting Challenges.") The data indicate that there is some merit to classifying companies according to the primary way they add value through their international operations over long periods of time. But that either/or characterization of globalization strategies is very broad. Finer-grained analysis of case studies—particularly of companies that have in various ways been global innovators—suggests that it is possible to combine the two approaches to some extent.
For a start, it's possible to apply different strategies to different elements of a business. CEMEX pursued a financial strategy of arbitraging capital cost differences even as it implemented a standardized operational strategy. It set up complete, uniform production-to-distribution chains in most of its major markets, reinforced by cross-border scale economies in such areas as trading, logistics, information technology, and innovation (in the broadest sense of the term). Mixing and matching was possible in this case because, to a large extent, CEMEX can choose how to raise capital independently from the way it chooses to compete in product markets.
Some companies have gone further. Consider the case of GE Medical Systems (GEMS), the division that Jeffrey Immelt built up between 1997 and 2000 before he was tapped to take over from Jack Welch as CEO. Immelt pushed for acquisitions to build up scale because, for the leading global competitors, an R&D-to-sales ratio of at least 8 percent represented a significant source of scale economies. But he also implemented a production strategy that was intended to arbitrage cost differences by concentrating manufacturing operations—and, ultimately, other activities—wherever in the world they could be carried out most cost effectively. By 2001, GEMS obtained 15 percent of its direct material purchases from, and had located 40 percent of its own manufacturing activities in, low-cost countries.
Even the best management can only go so far in melding the two strategies.
Like CEMEX, GEMS was able to pursue both approaches because it could organize its operations into relatively autonomous bundles of activities (like product development) in which economies of scale and standardization were essential and those (like procurement and manufacturing) where arbitrage economies were being pursued. What's more, it was able to coordinate its widely dispersed operations by applying centrally developed learning templates. In particular, Immelt applied the "pitcher-catcher concept," developed elsewhere in GE, in which for each move, a pitching team at a high-cost existing plant works with a catching team at a low-cost new location, and the move is not considered complete until the performance of the catching team meets or exceeds that of the pitching team. As a result, GEMS (and GE) seems to have managed to move production to low-cost countries faster than European competitors such as Philips and Siemens while also benefiting from greater scale economies.
But even the best management can get only so far in melding the two strategies. Acer, one of the world's largest computer manufacturers, supplies a cautionary tale of what can happen when companies go too far. Acer entered early into the contract manufacturing of personal computers, operating in low-wage Taiwan, and made good money with that arbitrage play. But in the early 1990s, it began to push Acer as a global brand, particularly in developed markets. This two-track approach turned out to be problematic. The branded business grew to significant volumes but continued to generate losses because the competitive environment was particularly tough for a late mover. Meanwhile, customers for Acer's contract-manufacturing arm worried that their business secrets would spill over to its competing line of business. They also feared that Acer could cross-subsidize its own brand with profits from its contract-manufacturing operations and so undercut their prices. In 2000, the strategy blew up when IBM canceled a major order, reducing its share of Acer's total contract-manufacturing revenues from 53 percent in the first quarter of 2000 to only 26 percent in the second quarter of 2001.
Acer responded by making some hard choices. Contract manufacturing has remained focused on customers in developed countries—and will gradually be spun off into a separate company. Meanwhile, sales of its own branded products have been focused on the East Asia region, particularly Greater China, where the contract customers cannot sell at a low enough price to compete. With this move, the company has acknowledged that the logic of exploiting similarities often calls for targeting countries similar to a company's home base, whereas the logic of arbitrage involves exploiting one or more of the differences inherent in distance.
The future of the globalization process is by no means obvious. Markets may integrate further once economic conditions improve. But some argue that the process could actually shift into reverse, toward even greater economic isolation, if the experience between the two World Wars is any precedent. Whatever the ultimate direction, though, the differences that make arbitrage valuable as well as the similarities that create scale economies will remain with us for the foreseeable future. That spells opportunity for those companies that have the imagination to see the full range of possibilities.