Managing Across Borders: New Organizational Responses

IN A COMPANION ARTICLE (Summer 1987), we described how recent changes in the international operating environment have forced companies to optimize efficiency, responsiveness, and learning simultaneously in their worldwide operations. To companies that previously concentrated on developing and managing one of these capabilities, this new challenge implied not only a total strategic reorientation but a major change in organizational capability, as well.

Implementing such a complex, three-pronged strategic objective would be difficult under any circumstances, but in a worldwide company the task is complicated even further. The very act of “going international” multiplies a company’s organizational complexity. Typically, doing so requires adding a third dimension to the existing business- and function-oriented management structure. It is difficult enough balancing product divisions that bring efficiency and focus to domestic product-market strategies with corporate staffs whose functional expertise allows them to play an important counterbalance and control role. The thought of adding capable, geographically oriented management— and maintaining a three-way balance of organizational perspectives and capabilities among product, function, and area—is intimidating to most managers. The difficulty is increased because the resolution of tensions among product, function, and area managers must be accomplished in an organization whose operating units are often divided by distance and time and whose key members are separated by culture and language.

From Unidimensional to Multidimensional Capabilities

Faced with the task of building multiple strategic capabilities in highly complex organizations, managers in almost every company we studied made the simplifying assumption that they were faced with a series of dichotomous choices.1 They discussed the relative merits of pursuing a strategy of national responsiveness as opposed to one based on global integration; they considered whether key assets and resources should be centralized or decentralized; and they debated the need for strong central control versus greater subsidiary autonomy. How a company resolved these dilemmas typically reflected influences exerted and choices made during its historical development. In telecommunications, ITT’s need to develop an organization responsive to national political demands and local specification differences was as important to its survival in the pre- and post-World War II era as was NEC’s need to build its highly centralized technological manufacturing and marketing skills and resources in order to expand abroad in the same industry in the 1960s and 1970s.

When new competitive challenges emerged, however, such unidimensional biases became strategically limiting. As ITT demonstrated by its outstanding historic success and NEC showed by its more delayed international expansion, strong geographic management is essential for development of dispersed responsiveness. Geographic management allows worldwide companies to sense, analyze, and respond to the needs of different national markets.

Effective competitors also need to build strong business management with global product responsibilities if they are to achieve global efficiency and integration. These managers act as champions of manufacturing rationalization, product standardization, and low-cost global sourcing. (As the telecommunications switching industry globalized, NEC’s organizational capability in this area gave it a major competitive advantage.) Unencumbered by either territorial or functional loyalties, central product groups remain sensitive to overall competitive issues and become agents to facilitate changes that, though painful, are necessary for competitive viability.

Finally, a strong, worldwide functional management allows an organization to build and transfer its core competencies—a capability vital to worldwide learning. Links between functional managers allow the company to accumulate specialized knowledge and skills and to apply them wherever they are required in the worldwide operations. Functional management acts as the repository of organizational learning and as the prime mover for its consolidation and circulation within the company. It was for want of a strongly linked research and technical function across subsidiaries that ITT failed in its attempt to coordinate the development and diffusion of its System 12 digital switch.

Thus, to respond to the needs for efficiency, responsiveness, and learning simultaneously, the company must develop a multidimensional organization in which the effectiveness of each management group is maintained and in which each group is prevented from dominating the others. As we saw in company after company, the most difficult challenge for managers trying to respond to broad, emerging strategic demands was to develop the new elements of multidimensional organization without eroding the effectiveness of their current unidimensional capability.

Overcoming Simplifying Assumptions

For all nine companies at the core of our study, the challenge of breaking down biases and building a truly multidimensional organization proved difficult. Behind the pervasive either/or mentality that led to the development of unidimensional capabilities, we identified three simplifying assumptions that blocked the necessary organizational development. The need to reduce organizational and strategic complexity has made these assumptions almost universal in worldwide companies, regardless of industry, national origin, or management culture.

• There is a widespread, often implicit assumption that roles of different organizational units are uniform and symmetrical; different businesses should be managed in the same way, as should different functions and national operations.
• Most companies, some consciously, most unconsciously, create internal interunit relationships on clear patterns of dependence or independence, on the assumption that such relationships should be clear and unambiguous.
• Finally, there is the assumption that one of corporate management’s principal tasks is to institutionalize clearly understood mechanisms for decision making and to implement simple means of exercising control.

Those companies most successful in developing truly multidimensional organizations were the ones that challenged these assumptions and replaced them with some very different attitudes and norms. Instead of treating different businesses, functions, and subsidiaries similarly, they systematically differentiated tasks and responsibilities. Instead of seeking organizational clarity by basing relationships on dependence or independence, they built and managed interdependence among the different units of the companies. And instead of considering control their key task, corporate managers searched for complex mechanisms to coordinate and coopt the differentiated and interdependent organizational units into sharing a vision of the company’s strategic tasks. These are the central organizational characteristics of what we described in the earlier article as transnational corporations —those most effective in managing across borders in today’s environment of intense competition and rapid, often discontinuous change.

From Symmetry to Differentiation

Like many other companies we studied, Unilever built its international operations under an implicit assumption of organizational symmetry. Managers of diverse local operating companies in products ranging from packaged foods to chemicals and detergents all reported to strongly independent national managers, who in turn reported through regional directors to the board. In the post-World War II era, the company began to recognize a need to supplement this geographically dominated structure with an organizational ability to capture potential economies and to transfer learning across national boundaries, To meet this need, a few product-coordination groups were formed at the corporate center. But the assumption of organizational symmetry ensured that all businesses were similarly managed, and the number of coordination groups grew from three in 1962 to six in 1969 and to ten by 1977.

By the mid-1970s, however, the entrenched organizational symmetry was being threatened. Global economic disruption caused by the oil crisis dramatically highlighted the very substantial differences in the company’s businesses and markets and forced management to recognize the need to differentiate its organizational structures and administrative processes. While standardization, coordination, and integration paid high dividends in the chemical and detergent businesses, for example, important differences in local tastes and national cultures impeded the same degree of coordination in foods. As a result, the roles, responsibilities, and powers of the central product-coordination groups eventually began to diverge as the company tried to shake off the constraint of the symmetry assumption.

But as Unilever tackled the challenge of managing some businesses in a more globally coordinated manner, it was confronted with the question of what to coordinate. Historically, the company’s philosophy of decentralized capabilities and delegated responsibilities resulted in most national subsidiaries’ becoming fully integrated, self-sufficient operations. While they were free to draw on product technology, manufacturing capabilities, and marketing expertise developed at the center, they were not required to do so, and most units chose to develop, manufacture, and market products as they thought appropriate. Thus functions, too, tended to be managed symmetrically.

Over time, decentralization of all functional responsibilities became increasingly difficult to support. In the 1970s, for example, when arch-competitor Procter & Gamble’s subsidiaries were launching a new generation of laundry detergents based on the rape seed formula created by the parent company, most of Unilever’s national detergent companies responded with their own products. The cost of developing thirteen different formulations was extremely high, and management soon recognized that not one was as good as P&G’s centrally developed product. For the sake of cost control and competitive effectiveness, Unilever had to break with tradition and begin centralizing European product development. The company has since created a system in which central coordination is more normal, although very different for different functions such as basic research, product development, manufacturing, marketing, and sales.

Just as they saw the need to change symmetrical structures and homogeneous processes imposed on different businesses and functions, most companies we observed eventually recognized the importance of differentiating the management of diverse geographic operations. Despite the fact that various national subsidiaries operated with very different external environments and internal constraints, they all traditionally reported through the same channels, operated under similar planning and control systems, and worked under a set of common and generalized mandates.

Increasingly, however, managers recognized that such symmetrical treatment can constrain strategic capabilities. At Unilever, for example, it became clear that Europe’s highly competitive markets and closely linked economies meant that its operating companies in that region required more coordination and control than those in, say, Latin America. Little by little, management increased the product-coordination groups’ role in Europe until they had direct line responsibility for all operating companies in their businesses. Elsewhere, however, national management maintained its historic line management role, and product coordinators acted only as advisers. Unilever has thus moved in sequence from a symmetrical organization to a much more differentiated one: differentiating by product, then by function, and finally by geography.

Recently, within Europe, differentiation by national units has proceeded even further. Operations in “key countries” such as France, Germany, and the United Kingdom are allowed to retain considerably more autonomy than those in “receiver countries” such as Switzerland, Sweden, Holland, and Denmark. While the company’s overall commitment to decentralization is maintained, “receiver countries” have gradually become more dependent on the center for direction and support, particularly in the areas of product development and competitive strategy.

Figure 1 is a schematic representation of the different ways in which Unilever manages its diverse businesses, functions, and markets.2 The vertical axis represents the level of global integration, and hence of central coordination; the horizontal axis represents the extent of national differentiation, and consequently of the desired influence of subsidiaries in strategic and operational decisions.

The detergent business must be managed in a more globally integrated manner than packaged foods, but also needs a more nationally differentiated strategy than the chemicals business. But not all tasks need to be managed in this differentiated yet coordinated manner: there is little need for national differentiation in research or for global coordination of sales management. And even those functions such as marketing that exhibit the more complex simultaneous demands need not be managed in this way in all national markets. Marketing strategy for export sales can be highly coordinated, while approaches taken in closed markets like India and Brazil can be managed locally. Only in key strategic markets like Germany, the U.K., and France is there a need for differentiated yet coordinated marketing strategies. This flexible and differentiated management approach stands in marked contrast to the standardized, symmetrical approach implied in Unilever’s earlier blanket commitment to decentralized responsibility.

But Unilever is far from unique. In all of the companies we studied, senior management was working to differentiate its organizational structure and processes in increasingly sophisticated ways.3 For example, Philips’s consumer electronics division began experimenting with an organization differentiated by product life-cycle stage—high-tech products like CD players being managed with very different strategies and organization processes from those for stable high-volume products like color TVs, which, in turn, were managed differently from mature and declining products like portable radios. Procter & Gamble is differentiating the roles of its subsidiaries by giving some of them responsibilities as “lead countries” in product strategy development, then rotating that leadership role from product to product.4 Matsushita differentiates the way it manages its worldwide operations not on the basis of geography, but on the unit’s strategic role. (Single-product, wholly owned manufacturing units, the A Group, are managed differently from multiproduct, multifunction companies, the B Group, and from simple sales and marketing subs, the C Group.) L.M. Ericsson, which had centralized most of the basic research on its digital switch, is now decentralizing development and applications responsibilities to a few key country subsidiaries that have the capability to contribute.5

Thus, instead of deciding the overall roles of product, functional, and geographic management on the basis of simplistic dichotomies such as global versus domestic businesses or centralized versus decentralized organizations, many companies are creating different levels of influence for different groups as they perform different activities. Doing this allows the relatively underdeveloped management perspectives to be built in a gradual, complementary manner rather than in the sudden, adversarial environment often associated with either/or choices. Internal heterogeneity has made the change from unidimensional to multidimensional organization easier by breaking the problem up into many small, differentiated parts and by allowing for a step-by-step process of organizational change.

From Dependence or Independence to Interdependence

The limitations of the assumption of clarity in organizational relationships eventually confronted top managers in the Japanese soap and detergent company Kao. In the early 1980s they began to recognize that their foreign subsidiaries’ strong dependence on the parent company provided significant benefits of global efficiency only at the cost of less sensitivity and responsiveness to local market needs. For example, when investigating the reason for the company’s slow penetration of the shampoo market in Thailand despite offering a technologically superior product, headquarters managers found that the subsidiary had adopted the product positioning, packaging, and pricing policies developed for the Japanese domestic market. Since local management had been unable to make the necessary local adaptations, managers were brought in from headquarters to identify the source of the problem and to make necessary changes in the marketing mix.

In other companies we studied—Unilever and ITT, for example—clarity of organizational relationships was achieved by giving foreign subsidiaries substantial independence. But, as our earlier discussion of Unilever illustrated, such organizational clarity was achieved at the cost of substantial inefficiency; individual subsidiaries often reinvented the wheel or operated at suboptimal scale.

New strategic demands make organizational models of simple interunit dependence or independence inappropriate. The reality of today’s worldwide competitive environment demands collaborative information sharing and problem solving, cooperative support and resource sharing, and collective action and implementation. Independent units risk being picked off one-by-one by competitors whose coordinated global approach gives them two important strategic advantages—the ability to integrate research, manufacturing, and other scale-efficient operations, and the opportunity to cross-subsidize the losses from battles in one market with funds generated by profitable operations in home markets or protected environments.6 The desire to capture such strategic benefits was one of Philips’s main motivations as it attempted to coordinate the competitive responses of historically independent national organizations.

On the other hand, foreign operations totally dependent on a central unit must deal with problems reaching beyond the loss of local market responsiveness described in the Kao example. They also risk being unable to respond effectively to strong national competitors or to sense potentially important local market or technical intelligence. This was the problem Procter & Gamble’s Japan subsidiary faced in an environment where local competitors began challenging P&G’s previously secure position with successive, innovative product changes and novel market strategies, particularly in the disposable diapers business. After suffering major losses in market share, management recognized that a local operation focused primarily on implementing the company’s classic marketing strategy was no longer sufficient; the Japanese subsidiary needed the freedom and incentive to be more innovative. Not only to ensure the viability of the Japanese subsidiary, but also to protect its global strategic position, P&G realized it had to expand the role of the local unit and change its relationship with the parent company to enhance two-way learning and mutual support.

But it is not easy to change relationships of dependence or independence that have been built up over a long history. Many companies have tried to address the increasing need for interunit collaboration by adding layer upon layer of administrative mechanisms to foster greater cooperation. Top managers have extolled the virtues of teamwork and have even created special departments to audit management response to this need. In most cases these efforts to obtain cooperation by fiat or by administrative mechanisms have been disappointing. The independent units have feigned compliance while fiercely protecting their independence. The dependent units have found that the new cooperative spirit implies little more than the right to agree with those on whom they depend.

Yet some companies have gradually developed the capability to achieve such cooperation and to build what Rosabeth Kanter calls an “integrative organization.”7 Of the companies we studied, the most successful did so not by creating new units, but by changing the basis of the relationships among product, functional, and geographic management groups. From relations based on dependence or independence, they moved to relations based on formidable levels of explicit, genuine interdependence. In essence, they made integration and collaboration self-enforcing by making it necessary for each group to cooperate in order to achieve its own interests. Companies were able to create such interdependencies in many ways, as two brief examples will illustrate.

• NEC has developed reciprocal relationships among different parts of its organizations by creating a series of internal quasi markets. It builds cooperation between the R&D function and the different product groups by allocating only a part of the R&D budget directly to the company’s several central laboratories. This portion is used to support basic and applied research in core technologies of potential value to the corporation as a whole. The remaining funds are allocated to die product groups to support research programs that reflect their priorities. In response to the product divisions’ proposed projects, each research group puts forward proposals that it feels will lead to the desired product or process improvements. What follows is a negotiation process that results in the product divisions’ “buying” some of the proposals put up by the laboratories, while different R&D groups adopt some of the projects demanded by the product managers. In other words, NEC has created an internal market for ensuring that research is relevant to market needs. (A similar process seems to have had comparable success at Matsushita.)8
• Procter & Gamble employs an entirely different approach to creating and managing interdependencies. In Europe, for example, it formed a number of Eurobrand teams for developing product-market strategies for different product lines.9 Each team is headed by the general manager of a subsidiary that has a particularly well-developed competence in that business. It also includes the appropriate product and advertising managers from the other subsidiaries and relevant functional managers from the company’s European headquarters. Each team’s effectiveness clearly depends on the involvement and support provided by its members and, more important, by the organizational units they represent. Historically, the company’s various subsidiaries had little incentive to cooperate. Now, however, the success of each team—and the reputation of the general manager heading it— depends on the support of other subsidiaries; this has made cooperation self-enforcing. Each general manager is aware that the level of support and commitment he can expect from the other members of the Eurobrand team depends on the support and contribution the product managers from his subsidiaries provide to the other teams. The interdependencies of these Eurobrand teams were able to foster teamwork driven by individual interests.

In observing many such examples of companies building and extending interdependence among units, we were able to identify three important flows that seem to be at the center of the emerging organizational relationships. Most fundamental was the product interdependence that most companies were building as they specialized and integrated their worldwide manufacturing operations to achieve greater efficiency, while retaining sourcing flexibility and sensitivity to host country interests.10 The resulting flow of parts, components, and finished goods increased the interdependence of the worldwide operations in an obvious and fundamental manner.

We also observed companies developing a resource interdependence that often contrasted sharply with earlier policies that had either encouraged local self-sufficiency or required the centralization of all surplus resources. Systems such as NEC’s internal quasi markets were designed to develop a greater flow of funds, skills, and other scarce resources among organizational units.

Finally, the worldwide diffusion of technology, the development of international markets, and the globalization of competitive strategies have meant that vital strategic information now exists in many different locations worldwide. Furthermore, the growing dispersion of assets and delegation of responsibilities to foreign operations have resulted in the development of local knowledge and expertise that has implications for the broader organization. With these changes, the need to manage the flow of intelligence, ideas, and knowledge has become central to the learning process and has reinforced the growing interdependence of worldwide operations, as P&G’s Eurobrand teams illustrate.

It is important to emphasize that the relationships we are highlighting are different from the interdependencies commonly observed in multiunit organizations. Traditionally, MNC managers have attempted to highlight what has been called “pooled interdependence” to make subunit managers responsive to global rather than local interests. (Before the Euroteam approach, for instance, P&G’s European vice president often tried to convince independent-minded subsidiary managers to transfer surplus generated funds to other more needy subsidiaries, in the overall corporate interest, arguing that, “Someday when you’re in need they might be able to fund a major product launch for you.”)

As the example illustrates, pooled interdependence is often too broad and amorphous to affect day-to-day management behavior. The interdependencies we described earlier are more clearly reciprocal, and each unit’s ability to achieve its goals is made conditional upon its willingness to help other units achieve their own goals. Such interdependencies more effectively promote the organization’s ability to share the perspectives and link the resources of different components, and thereby to expand its organizational capabilities.11

From Control to Coordination and Cooption

The simplifying assumptions of organizational symmetry and dependence (or independence) had allowed the management processes in many companies to be dominated by simple controls—tight operational controls in subsidiaries dependent on the center, and a looser system of administrative or financial controls in decentralized units.12 When companies began to challenge the assumptions underlying organizational relationships, however, they found they also had to adapt their management processes. The growing interdependence of organizational units strained the simple control-dominated systems and underlined the need to supplement existing processes with more sophisticated ones. Furthermore, the differentiation of organizational tasks and roles amplified the diversity of management perspectives and capabilities and forced management to differentiate management processes.

As organizations became, at die same time, more diverse and more interdependent, there was an explosion in the number of issues that had to be linked, reconciled, or integrated. The rapidly increasing flows of goods, resources, and information among organizational units increased the need for coordination as a central management function. But the costs of coordination are high, both in financial and human terms, and coordinating capabilities are always limited. Most companies, though, tended to concentrate on a primary means of coordination and control—”the company’s way of do ing things.” (At ITT it was through “the system,” as Harold Geneen used to call his sophisticated set of controls, while at Kao it was primarily through centralization of decisions.) Clearly, there was a need to develop multiple means of coordination, to rank the demands for coordination, and to allocate the scarce coordinating resources. The way in which one of our sample companies developed its portfolio of coordinative processes illustrates the point well.

During the late 1970s and early 1980s, Philips had gradually developed some sophisticated means of coordination. This greatly helped the company shape its historically evolved, nationally centered organization into the kind of multidimensional organization it needed to be in the 1980s. Coordinating the flow of goods in a global sourcing network is a highly complex logistical task, but one that can often be formalized and delegated to middle and lower-level management. By standardizing product specifications and rationalizing sourcing patterns through designating certain plants as international production centers (IPCs), Philips facilitated goods-flow coordination. By making these flows reasonably constant and forecastable, the company could manage them almost entirely through formal systems and processes. These became the main coordination mechanisms in the company’s attempt to increase the integration of worldwide sourcing of products and components.

Coordinating the flow of financial, technical, and human resources, however, was not so easily routinized. Philips saw the allocation of these scarce resources as a reflection of key strategic choices and therefore managed the coordination process by centralizing many decisions. The board became heavily involved in major capital budgeting decisions; the product divisions reasserted control over product development, a process once jealously guarded by the national organizations; and the influential corporate staff bureau played a major role in personnel assignments and transfers.

But while goods flows could be coordinated through formalization, and resource flows through centralization, critical information flows were much more difficult to manage. The rapid globalization of the consumer electronics industry in the 1970s forced Philips to recognize the need to move strategic information and proprietary knowledge around the company much more quickly. While some routine data could be transferred through normal information systems, much of the information was so diverse and changeable that establishing formal processes was impossible. While some core knowledge had to be stored and transferred through corporate management, the sheer volume and complexity of information—and the need for its rapid diffusion—limited the ability to coordinate through centralization. Philips found that the most effective way to manage complex flows of information and knowledge was through various socialization processes: the transfer of people, the encouragement of informal communication channels that fostered information exchange, or the creation of forums that facilitated interunit learning.

Perhaps most well known is the company’s constant worldwide transfer and rotation of a group of senior managers (once referred to internally as the “Dutch Mafia,” but today a more international group) as a means of transferring critical knowledge and experience throughout the organization. Philips also made more extensive use of committees and task forces than any other company we studied. Although the frequent meetings and constant travel were expensive, the company benefited not only from information exchange but also from the development of personal contacts mat grew into vital information channels.

In other companies, we saw a similar broadening of administrative processes as managers learned to operate with previously underutilized means of coordination. Unilever’s heavy reliance on the socialization of managers to provide the coordination “glue” was supplemented by the growing role of the central product-coordination departments. In contrast, NEC reduced central management’s coordination role by developing formal systems and social processes in a way that created a more robust and flexible coordinative capability.

Having developed diverse new means of coordination, management’s main task is to carefully ration their usage and application. As the Philips example illustrates, it is important to distinguish where tasks can be formalized and managed through systems, where social linkages can be fostered to encourage informal agreements and cooperation, and where the coordination task is so vital or sensitive that it must use the scarce resource of central management arbitration.13

While the growing interdependence of organizational units forces the development of more complex administrative processes, the differentiation of roles and responsibilities forces management to change the way it uses the new coordination and control mechanisms. Even though they recognize the growing diversity of tasks facing them, a surprising number of companies have had great difficulty in differentiating the way they manage products, functions, or geographic units. The simplicity of applying a single planning and control system across businesses and the political acceptability of defining uniform job descriptions for all subsidiary heads were often allowed to outweigh the clear evidence that the relevant business characteristics and subsidiary roles were vastly different.

We have described briefly how companies began to remedy this situation by differentiating roles and responsibilities within the organization. Depending on their internal capabilities and on the strategic importance of their external environments, organizational units might be asked to take on roles ranging from that of strategic leader with primary corporatewide responsibility for a particular business or function, to simple implementer responsible only for executing strategies and decisions developed elsewhere.

Clearly, these roles must be managed in quite different ways. The unit with strategic leadership responsibility must be given freedom to develop responsibility in an entrepreneurial fashion, yet must also be strongly supported by headquarters. For this unit, operating controls may be light and quite routine, but coordination of information and resource flows to and from the unit will probably require intensive involvement from senior management. In contrast, units with implementation responsibility might be managed through tight operating controls, with standardized systems used to handle much of the coordination —primarily of goods flows. Because the tasks are more routine, the use of scarce coordinating resources could be minimized.

Differentiating organizational roles and management processes can have a fragmenting and sometimes demotivating effect, however. Nowhere was this more clearly illustrated than in the many companies that unquestioningly assigned units the “dog” and “cash cow” roles defined by the Boston Consulting Group’s growth-share matrix in the 1970s.14 Their experience showed that there is another equally important corporate management task that complements and facilitates coordination effectiveness. We call this task cooption:. the process of uniting the organization with a common understanding of, identification with, and commitment to the corporation’s objectives, priorities, and values.

A clear example of the importance of cooption was provided by the contrast between ITT and NEC managers. At ITT, corporate objectives were communicated more in financial than in strategic terms, and the company’s national entities identified almost exclusively with their local environment. When corporate management tried to superimpose a more unified and integrated global strategy, its local subsidiaries neither understood nor accepted the need to do so. For years they resisted giving up their autonomy, and management was unable to replace the interunit rivalry with a more cooperative and collaborative process.

In contrast, NEC developed an explicitly defined and clearly communicated global strategy enshrined in the company’s “C&C” motto—a corporatewide dedication to building business and basing competitive strategy on the strong link between computers and communications. For over a decade, the C&C philosophy was constantly interpreted, refined, elaborated, and eventually institutionalized in organizational units dedicated to various C&C missions (e.g., the C&C Systems Research Laboratories, the C&C Corporate Planning Committee, and eventually the C&C Systems Division). Top management recognized that one of its major tasks was to inculcate the worldwide organization with an understanding of the C&C strategy and philosophy and to raise managers’ consciousness about the global implications of competing in these converging businesses. By the mid-1980s, the company was confident that every NEC employee in every operating unit had a clear understanding of NEC’s global strategy as well as of his or her role in it. Indeed, it was this homogeneity that allowed the company to begin the successful decentralization of its strategic tasks and the differentiation of its management processes.

Thus the management process that distinguished transnational organizations from simpler unidimensional forms was one in which control was made less dominant by the increased importance of interunit integration and collaboration. These new processes required corporate management to supplement its control role with the more subtle tasks of coordination and cooption, giving rise to a much more complex and sophisticated management process.

Sustaining a Dynamic Balance: Role of the “Mind Matrix”

Developing multidimensional perspectives and capabilities does not mean that product, functional, and geographic management must have the same level of influence on all key decisions. Quite the contrary. It means that the organization must possess a differentiated influence structure—one in which different groups have different roles for different activities. These roles cannot be fixed but must change continually to respond to new environmental demands and evolving industry characteristics. Not only is it necessary to prevent any one perspective from dominating the others, it is equally important not to be locked into a mode of operation that prevents reassignment of responsibilities, realignment of relationships, and rebalancing of power distribution. This ability to manage die multidimensional organization capability in a flexible manner is the hallmark of a transnational company.

In the change processes we have described, managers were clearly employing some powerful organizational tools to create and control the desired flexible management process. They used the classic tool of formal structure to strengthen, weaken, or shift roles and responsibilities over time, and they employed management systems effectively to redirect corporate resources and to channel information in a way that shifted the balance of power. By controlling the ebb and flow of responsibilities, and by rebalancing power relationships, they were able to prevent any of the multidimensional perspectives from atrophying. Simultaneously, they prevented the establishment of entrenched power bases.

But the most successful companies had an additional element at the core of their management processes. We were always conscious that a substantial amount of senior management attention focused on the individual members of the organization. NEC’s continual efforts to inculcate all corporate members with a common vision of goals and priorities; P&G’s careful assignment of managers to teams and task forces to broaden their perspectives; Philips’s frequent use of conferences and meetings as forums to reconcile differences; and Unilever’s extensive use of training as a powerful socialization process and its well-planned career path management that provided diverse experience across businesses, functions, and geographic locations—all are examples of companies trying to develop multidimensional perspectives and flexible approaches at the level of the individual manager.

What is critical, then, is not just the structure, but also the mentality of those who constitute the structure. The common thread that holds together the diverse tasks we have described is a managerial mindset that understands die need for multiple strategic capabilities, that is able to view problems from both local and global perspectives, and that accepts the importance of a flexible approach. This pattern suggests that managers should resist the temptation to view their task in the traditional terms of building a formal global matrix structure —an organizational form that in practice has proven extraordinarily difficult to manage in the international environment. They might be better guided by the perspective of one top manager who described the challenge as “creating a matrix in the minds of managers.”

Our study has led us to conclude that a company’s ability to develop transnational organizational capability and management mentality will be the key factor that separates the winners from the mere survivors in the emerging international environment.