Funding from the Carnegie Bosch Institute for Applied Studies in International Management is gratefully acknowledged. 

The importance of managing organizational knowledge is increasingly gaining recognition from both scholars and practitioners. In the past decade and a half, seminal works have been published, such as Nelson and Winter’s (1982) examination of organizational routines, Teece’s (1977; 1982) work on technology transfer and proprietary knowledge, and Nonaka’s (1990, 1994) studies of knowledge creation. Notwithstanding this significant body of knowledge, however, there remain many aspects of organizational knowledge and knowledge transfer of which we know very little. In particular, the importance of context in shaping firms’ knowledge bases and transfer capacities is a key area requiring further study. Every organizational practice, routine, or piece of information is deeply embedded within its context. Firms develop specific capabilities in specific contexts. Although many studies of organizational knowledge note that context is important, few take context into account in their analysis (Szulanski’s 1996 study being an exception). The current study recognizes the importance of context in the development and utilization of knowledge in organizations. We develop a model to explain how context defines organizational knowledge and enables knowledge transfers. We then examine the model using data from three case studies involving international knowledge transfer. 

Multinational companies (MNCs) operate in an increasingly complex environment. We use a context-centered model to explain the difficulty MNCs experience while engaging in internal knowledge transfers. Replication of existing successful practices and routines enables multinational organizations to maximize their value (Bartlett & Ghoshal, 1989). MNCs use knowledge transfers to economize on their existing body of knowledge, to expand their competitive advantage base, and to make sure all their subsidiaries are at par with the most successful ones (Kogut & Zander, 1993). 

The sharing of knowledge across the organization has been recognized by organizational studies literature to be a critical driver of firm performance (Grant, 1991; Inkpen, 1996; Prahalad & Hamel, 1990). However, intra-firm knowledge transfers are just as problematic, expensive, and difficult as they are critical to organizational survival. Many large MNCs encounter great difficulties in transferring practices across organizational units. For example, a large consumer products company, in attempting to globally integrate various product management tasks, found that European managers resented what they saw as American managers with a limited understanding of strategy. In a case study, Cerny (1996) described a COO who recognized the need to encourage and facilitate the flow of knowledge across borders. However, because knowledge was not being transferred effectively, the company was in serious trouble. Thus, the research question raised by this paper is what factors influence the difficulty level of intrafirm, cross border knowledge transfers. More specifically, what role does context similarity between the knowledge source and recipient play in explaining transfer difficulty? 

The model developed in this paper suggests that various organizational units face different contexts. Every knowledge pocket is embedded within a specific set of contextual elements, which are critical to the firm’s ability to hold and utilize the knowledge. We divide the elements into five major dimensions: strategic, decision-making, environmental, cultural, and technological. When engaging in a knowledge transfer, firms attempt to take a knowledge pocket and replant it in a new, different context in the knowledge recipient. Our model suggests that difficulties in transfer stem from an absence of some or all of the critical contextual variables at the recipient. We propose that the more similar contextual variables present at the recipient relative to those at the source, the more successful the transfer will be. 

Organizational knowledge 

Viewing the firm as a body of knowledge has become a central theme in the organizational studies literature. The resource-based view of the firm (Barney, 1991; Wernerfelt, 1984), which focuses on a firm’s inimitable capabilities as a critical source for sustainable competitive advantage, places knowledge at the center of its latest theoretical developments (e.g. Conner & Prahalad, 1996). Similarly, organizational knowledge plays a role in other organizational research streams, such as organizational design (Sanchez & Mahoney, 1996) and strategic alliances (Inkpen & Beamish, 1997; Mowery, Oxley & Silverman, 1996). Increasingly, organizational research is centered around knowledge-based theories of firms (Foss, 1996) or around using knowledge as a basis for theory (Spender, 1996). 

Hedlund and Nonaka (1993: 117) defined knowledge as being constructed from "cognitive perceptions as well as skills and expertise embodied in products or services." They made the distinction between tacit, intuitive, non-verbalized knowledge and articulated knowledge. Other definitions refer to the explicit or tangible versus the implicit, or tacit parts of knowledge. Teece (1977: 243), for example, discussed a "capability to manufacture a product or process" and identified two forms of technology: the physical, or "embodied" and the "unembodied" information regarding effective utilization of the technology, such as methods, procedures, and controls. Kogut and Zander (1992) also divided knowledge into two types: information and know-how. Information, or knowing what something means, includes facts, axiomatic propositions, and symbols. To define know-how, or knowing how to do something, Kogut and Zander cited von Hippel (1988), who stated that know-how is the accumulated practical skill or expertise that allows one to do something smoothly and efficiently. 

Various different views of knowledge have been proposed, including knowledge as a competency (Amit & Schoemaker, 1993; Barney, 1991; Grant 1991); as the basis for creating competencies by combining knowledge (Clark & Fujimoto, 1991; Kogut & Zander, 1992; Nonaka & Takeuchi, 1995); and as an input for innovation (Bartlett & Ghoshal, 1989; Cohen & Levinthal, 1990; von Hippel, 1994). As noted by Foss (1996), all knowledge-based approaches to the theory of the firm explain why some firms out-perform others and they all agree that knowledge is socially embedded. However, the knowledge-related literature does not fully recognize the importance and consequences of the context embeddedness of knowledge. 

Our objective is to understand why some firms hold certain pockets of knowledge while others firms hold different pockets. We suggest a knowledge spectrum model that incorporates the importance of contextual factors in shaping the differences between knowledge states and knowledge spectrums of various firms. Moreover, the model seeks to explain the difficulty level experienced by firms that engage in knowledge transfers. Before developing the model in detail, we review the literature on knowledge transfer. The review clearly demonstrates the need for more focus on contextual elements to enable an enhanced understanding of organizational knowledge transfers. 

Knowledge Transfer 

While the literature discussed above focuses on the nature and purpose of organizational knowledge, its importance in large, complex organizations must also be addressed. Whether knowledge is a competency, an input, or an integration process, MNCs seeking to maximize its value must transfer and share knowledge across the organization. Nelson and Winter (1982: 121) argued that "if an existing routine is a success, replication of that success is likely to be desired." The internationalization of learning was identified by Bartlett and Ghoshal (1989) as a key dimension of a transnational firm. They defined international learning as the development and sharing of knowledge across national boundaries. Similarly, Kogut and Zander (1993) referred to the ability to transfer superior knowledge at the international level as a primary source for competitive advantage as well as further development and growth of the MNC. 

The importance of interdependencies and knowledge flows across organizational units of the MNC has been recognized and extensively discussed (Doz & Prahalad, 1991; Ghoshal & Bartlett, 1988; Gupta & Govindarajan, 1991; Hedlund & Nonaka, 1993; Simon, 1973). It is knowledge flows that enable the transmission of unique solutions from one subunit to another, the coordination of various connected units, and the collaboration among them. Simon (1973) professed that a highly interdependent structure across subunits enables organizations to economize on coordination costs. Information and knowledge have a critical role in managing interdependencies, especially in the international arena, where task uncertainty may be very high (Galbraith, 1973). 

More recently, empirical evidence demonstrates the importance of knowledge flows as a specific medium for unit interdependence. Darr, Argote and Epple (1995) found that interdependent organizational units, connected through knowledge flows, exhibited greater cost reduction than independent units. Ghoshal and Bartlett (1988) found that communications across organizational units facilitated efficient innovation. Gupta and Govindarajan (1991) similarly recognized the importance of knowledge flows in facilitating organizational control across organizational subsidiaries. 

Knowledge transfer, or what is often referred to as "best practice transfer" is the focus of this paper. Knowledge is not only a competency by itself, such as a manufacturing process or human resource practice, but is also a potential for achieving competitive advantages. As can be seen in Figure 1, holding knowledge will not necessarily lead to a competitive advantage. Only effective use of the knowledge through efficient integration (Grant, 1996) or combining new and existing knowledge (Kogut & Zander, 1992) will lead to a best practice. Best practice transfer was defined by Szulanski (1996: 28) as "replication of an internal practice that is performed in a superior way in some part of the organization and is deemed superior to internal alternate practices and known alternatives outside the company," while a practice is a routine use of knowledge (Nelson & Winter, 1982). Figure 1 shows that only effective use of knowledge may lead to a best practice. 

Teece (1977) and Nelson and Winter (1982) laid the critical foundation for understanding knowledge transfer. Teece (1977) focused on the implicit, or unembodied forms of knowledge. Teece focused mainly on technological know-how and suggested that the international transfer of technology enables the firm to accumulate a stock of knowledge that is applicable across borders. The primarily variables Teece studied were the level and the determinants of the costs involved in transfers. Nelson and Winter (1982) examined the replications of organizational routines and claimed that possessing the routine’s template enables far better replication within the organization than across organizations. Kogut and Zander (1993) examined the characteristics of knowledge that inhibit its transfer and thus influence a firm’s decision about internalizing the knowledge transfer. If the technology being transferred is codifiable and teachable, the firm does not possess a relative advantage in transferring it, and thus will not transfer it. 

A rather different view is taken in a recent study by Szulanski (1996), who empirically investigated not only the characteristics of the knowledge being transferred, but also the situation or the context of transfer. He investigated factors that influence the "stickiness" of the knowledge, or the difficulty involved in the transfer, and found them concentrated mainly at the knowledge receiving unit. Lack of motivation at the recipient location, and its lack of absorptive capacity were found to be positively correlated with stickiness of knowledge transferred. 

The contributions of existing knowledge and knowledge transfer studies is significant. This work allows us to view the firm as a dynamic system that processes different types of knowledge (Spender, 1996) to support organizational renewal and sustainable competitive advantage (Prahalad & Hamel, 1994; Quinn, 1992). With the exception of Szulanski (1996), however, the importance of organizational context to enable this process is largely ignored. As a point of departure from prior studies, this paper focuses on the role of contextual factors in enabling effective knowledge transfer to create advantages such as strengthening the organizational knowledge base and better flexibility in responding to the firm’s environment. 

Knowledge Transfer Approaches 

Two approaches to knowledge transfer have been developed: the communication model and the knowledge spiral model. In the first, used by Szulanski (1996) to describe intra-firm knowledge transfer, the transfer is seen as a message transmission from a source to a recipient in a given context. More extensively analyzed by Dinur and Inkpen (1996), the process follows four stages: initiation, where transferred knowledge is recognized; adaptation, where knowledge is changed at the source location to the perceived needs of the recipient; translation, where more alterations occur at the recipient unit as part of the general problem-solving process of adaptation to new context; and implementation, where knowledge is institutionalized to become an integral part of the recipient unit. Four groups of related factors can be identified: source related factors, recipient related factors, factors relating to the relationship and distance between the two units, and factors related to the nature of the knowledge transferred. 

The second knowledge transfer model was suggested by Nonaka and Takeuchi(1995). Nonaka and Takeuchi asserted that four modes facilitate the conversion of knowledge from the individual to the organizational level. This process was described as the knowledge spiral. Inkpen and Crossan (1995) applied the knowledge spiral concept in a study of learning through international joint ventures. The four modes in the spiral are patterns of interaction between tacit and explicit knowledge. Through these processes existing knowledge can be converted into new knowledge. We suggest that these notions also can be used as an effective descriptive model for overall knowledge transfer. 

Using the knowledge spiral model of knowledge transfer, tacit knowledge can be transferred through two processes: socialization, which maintains the knowledge in its tacit form, and externalization, through which it is transformed into explicit knowledge. Explicit knowledge can be transferred through two other processes: combination, which retains its explicit nature, and internalization, a process through which explicit knowledge is converted into tacit knowledge. 

Embeddedness of Knowledge in Context 

Notwithstanding the importance and contribution of the various knowledge and knowledge transfer articles discussed above, and despite some references to the context embeddedness of knowledge and knowledge transfers, it is our contention that more attention should be given to the context in which firms in general, and MNCs in particular, utilize and transfer organizational knowledge. 

Firms are unique entities, composed of various unique organizational units. The circumstances of every unit, as well as the individuals that compose the units differ. Therefore, both the knowledge organizational units hold and the way they use it will also be unique (Tsoukas, 1996). The difference between organizational units stands at the core of understanding their ability to utilize and share knowledge. Every organizational unit operates in a context that is specific and unique. Organizational and sociological literature extensively discusses the context embeddedness of two distinguishable organizational layers: individuals and organizations. There is a societal, situational, and historical unarticulated background to every individual in an organizational role (Polanyi, 1962; 1975; Tsoukas, 1996). Individuals acquire this background through socialization (Taylor, 1993). A set of cognitive processes, physical repetition of a practices, as well as unintentional, embodied internalization, constitute an individual experience of unarticulated background (Hedlund & Nonaka, 1993; Moss, 1995; Taylor, 1993). Individuals, hence, operate within unique, individually, socially and organizationally embedded contexts. 

Organizational contexts are a basis for differentiating between organizational units. Organizational context differs from one physical unit to another and from one operational unit to another. The constructs used by organizational scholars to discuss context usually parallel organizational practices. Examples are the use of firm policies and standards of operations and performance (Szulanski, 1996) or work ethics such as standards, objectives, individual involvement level, and managerial support (Ghoshal & Bartlett, 1994). Other knowledge-based models primarily use individual, organizational, and environmental factors. For example, Leonard-Barton (1992: 114) discussed a knowledge system constructed of four dimensions, the central one being the "value assigned within the company to the context and structure of knowledge." The other three dimensions were technical systems within the organization, managerial systems of control and creation, and individual skills and knowledge. Gupta and Govindarajan (1991), in a framework of knowledge flow patterns in international subsidiaries, identified key context variables as task environment, structural characteristics, and required behavior, which could be inferred as organizational culture. 

The question we target is part of the organizational boundary question, first raised by Penrose (1959). However, our interest goes beyond boundaries of physical nature to boundaries of aptitude. Somewhat similar is Nelson and Winter’s (1982) notion of a firm bounded by its rules and routines. Spender (1996: 51) identified the need to ask "What is the minimum set of genetic material or organizational routines necessary to create life or the viable firm?" We address a closely related question: What defines and confines the ability of a firm to hold and effectively utilize knowledge? Effective utilization means the potential to turn knowledge into a competitive advantage-yielding capability (Grant, 1996). Considering the importance of knowledge in the value chain (Denison, 1968), the boundaries around the ability to utilize knowledge may be the boundaries of maintaining firm viability. 

In summary, the review in this paper suggests that contextual elements stand at the core of knowledge utilization and transfer. The specific context of each organizational unit provides it with specific tools as well as specific limitations. The use of context as an explanatory factor for firms’ operations and performance is not new. From Porter’s (1981) basic model of industry structure, conduct, and performance, it is clear that both environmental and internal factors influence firms. External factors such as industry entry barriers or demand elasticity and strategic factors such as a firm’s choices of price, quality or capacity both affect the innovativeness of a firm and its allocative and technological efficiencies. More contextual elements are discussed by Miller (1990), who claimed that firms lock in strategic, cultural and structural configurations. Once set, these configurations are extremely difficult to change, and will affect future directions the firm can take. Certain configurations may cause momentum towards viability-dissenting directions. The core idea shared by our model is that a firm’s situation limits its spectrum of available action or knowledge. 

The model in this paper uses the term knowledge spectrum as a core concept. Knowledge spectrum consists of the total knowledge an organization may be able to utilize. Organizational routines, technologies, processes, and procedures used by an organization are all confined within a five dimensional space, next to other routines and processes yet to be discovered. Knowledge spectrum is both the feasible and actual organizational knowledge: what is actually used by the firm and what could potentially be used. We argue that five contextual dimensions shape the spectrum and define it: environmental, cultural, strategic, decision-making, and technological (Table 1). The structural dimension is captured by the decision-making dimension, which incorporates the individual level of hierarchical position and power, as well as organizational characteristics such as communication style and incentive structures. The common thread of the dimensions is that they enable and restrict knowledge transfer. Each dimension contributes to the knowledge spectrum’s shape and size by affecting different organizational variables. 

The five dimensions influence a firm’s decision-making, behavior, skills, and view of the world. Their effect on a firm’s ability to hold knowledge is tightly connected to such influences. An organizational unit with experience in manufacturing cars will not usually come up with the latest drug, unless critical changes take place. A firm with a defender strategy will not easily think about venturing into the latest fad market. A firm operating in India is not likely to have the knowledge required to operate a steak house. Thus, the model suggests that every knowledge pocket that can be found in a firm must be considered in the light of the contextual elements that enable it to exist. Context shapes a firm’s ability to recognize knowledge, to utilize it, and to derive competitive advantage from it. As such, in order for a specific knowledge pocket to be successfully transferred from one organizational unit to another, similar or at least related contextual elements must exist in both locations. 

Three types of knowledge pockets exist (Figure 2). The first is within the knowledge spectrum and outside the knowledge state. This pocket has the potential to become a competency but is not yet recognized, identified, or discovered by the organization. The second is within the knowledge state. It has the potential to become a competency and is recognized and utilized by the organization. The third type of knowledge pocket exists outside of the knowledge spectrum. It is incompatible to the firm’s knowledge base and has very little potential of becoming used in a way that will yield a competency. The dimensions represent the specific contextual elements that are an integral part of the ability of the firm to utilize knowledge. A knowledge pocket that exists within the knowledge state is said to be in alignment with its context. This alignment creates the potential for a competitive-advantage-yielding competency utilizing this pocket of knowledge. Note that we are not suggesting that knowledge spectrums are rigid and unchangeable. However, the model does suggest that changing the boundaries that shape the organization’s abilities and knowledge is not an easy task. It requires effort, time, and resources. 

Knowledge Transfer within the Model 

When a transfer of a specific knowledge pocket takes place, it is extracted from its "natural" context and moved to new context. This new context may or may not be in alignment with the knowledge. Every organizational unit faces different restricting conditions across the five dimensions. Thus, every unit will develop different abilities in identifying, developing, and utilizing knowledge. If two units face somewhat similar levels of the restricting dimensions their knowledge spectrums will partially overlap. The more distinct the restricting conditions of every unit, the smaller will be the overlapping area of their knowledge spectrum (see Figure 3). 

Figure 3 illustrates two knowledge pockets within the source unit’s spectrum: pockets X and Y. Since both are candidates for transfer to the recipient unit, both are currently utilized by the source unit and are located within its knowledge state. However, when turning to the recipient unit, it is evident that while pocket Y is within the recipient’s knowledge spectrum, pocket X is in dissonance with the knowledge spectrum and is located outside the spectrum. Our model suggests that the transfer eventfulness of pocket Y will be significantly lower than of knowledge pocket X. Eventfulness was first empirically studied by Szulanski (1996), who used the term stickiness of knowledge (von Hippel, 1994). As measured by Szulanski, eventfulness has three primary attributes and all relate to how noticeable is the event. A knowledge transfer will be considered more eventful when it costs more than expected, takes longer than expected, or leaves the parties unsatisfied from any other reason. Note that eventfulness is a process rather than an outcome. Even through a transfer outcome may be successful - the knowledge pocket is incorporated into the recipient unit - such success may be preceded by a transfer that deviated from cost, time, and other specified limitations. Eventfulness is thus not necessarily a measure of success but rather of difficulty. 

To examine the proposed model and develop new insights in the knowledge transfer area, case study data were collected. Three case studies involving different knowledge transfer challenges were carried out (see Table 2 for characteristics of the cases). The most comprehensive case study involved a U.S. MNC’s establishment of a new manufacturing and marketing organization in a Latin American country. For this case, on-site interviews in three different cities in the Latin American country were carried out. In total, 12 managers, including the country manager, were interviewed over the course of a week. The second case involved a U.S. MNC’s efforts to transfer its technology and R&D process to various locations throughout the world. Interviews were conducted with 7 middle and senior managers, including the vice president of technology, at the headquarters of the U.S. MNC. The third case involved a U.S.-based MNC and its efforts to transfer its manufacturing processes to several new locations. Interviews with 10 senior and middle level managers were conducted at the headquarters of the U.S. MNC. 

Alpha. In the first case study, the U.S. MNC, which we call Alpha, had four strategic objectives associated with its knowledge transfer and internationalization: 1) re-establish its brand name in the Latin American country; 2) build an efficient manufacturing facility; 3) export to other Latin American countries using the new production capacity; and 4) establish the new production facility as a worldwide benchmark for efficient production. Thus, there were clear short-term objectives for the transfer of knowledge to the new organization and longer term objectives for the transfer of manufacturing best practices to other parts of the organization. In that sense, the new organization was intended to be both a knowledge receiving location and a knowledge source location. This study focuses on the short-term objectives since the outward transfer of knowledge will take additional years to become established. 

The technology transfer involved both process and product technology from other parts of the organization. One of the products will be about 70% U.S. technology and about 30% from another Latin American country. The other main product will be largely European technology. The manufacturing process technology was transferred from another plant in Latin America and a plant in Europe. Transferred knowledge also included both marketing knowledge and the overall Alpha culture since the objective was to build a new, sustainable organization that would survive for many years. Some of the specific non-manufacturing process best practices that Alpha was attempting to transfer from other parts of the organization included: a customer satisfaction measurement program, distribution network, team building (unfamiliar to most of the local employees), market research methods, and various human resource practices such as an internship program, cell structure in the plants, plant dress codes, and a collective bargaining process. 

Alpha was relying on various knowledge transfer mechanisms, including: 1) a large team of expatriates from various parts of the world; 2) an advisory system of visitors from other Latin American and international locations; 3) local and international training for managers and line workers; 4) videos from other locations; 5) extensive email communication; 6) visits outside the Latin American country for Latin American managers and supervisors; 7) international coordination of the manufacturing plant construction; and 8) company manuals. 

Beta. Although about 40% of Beta’s sales were outside the United States, product technologies largely originated in the United States. Thus, the firm was faced with two major challenges: 1) transferring new product technologies to other regions of the world and 2) transferring the technology development, i.e. R&D, process outside the United States. The first objective was critical to firm success and was generally accepted within the organization. The second objective was more controversial within the firm because there was no agreement that R&D should be distributed throughout the world. One argument was that as the firm’s non-U.S. sales grew, it was natural to begin the process of distributing R&D and early product development to other parts of the world. The argument was stated as follows: It is impossible to develop customer solutions if you do not understand customer problems. Different parts of the world should logically be the focus for problems unique to their area. Remote R&D facilities would facilitate technology transfers outside the United States. 

The counter to the above argument was that the culture and tacit knowledge of the centralized R&D unit was so unique it would be impossible to replicate it. Moreover, trying to replicate a smaller version of central research and development would fail. The complexity of Beta products and systems, it was argued, was such that a large team of R&D people was needed in one location to ensure interaction occurs between scientists. Decentralization would reduce interaction and personal contacts. Moreover, one of the main functions of Beta’s central R&D was to move ideas around the world. Decentralizing R&D would jeopardize this central dissemination function. 

Gamma. Gamma, a successful U.S.-based manufacturer, recognized the need to expand internationally if the firm was to remain a major supplier in its industry. Without international growth, Gamma risked losing its high quality status as a supplier. A major challenge for Gamma revolved around the question of how to transfer what was perceived to be a unique organizational culture. Among senior management there was a consensus that the foundation of Gamma’s success was its fundamental culture and values. As a result, the vision for expansion was that all Gamma facilities inside and outside the United States would have the Gamma underlying value system based on trust, respect, and integrity. 

The specific knowledge transfer that was examined for this study was the establishment of a manufacturing plant in Mexico. Explicit knowledge transfer was quite straight forward. The Mexican plant was only involved in manufacturing; there was no product development. The assembly line was a replication of one in Gamma’s U.S. facilities with about 75% of the equipment the same. Some of the automation was designed out of the process to take advantage of the lower cost of labor in Mexico. The transfer of the manufacturing knowledge involved a variety of linkages between Mexico and the United States, including training in Mexico and the United States and various specialized technical personnel from other Gamma facilities. 

The transfer of tacit knowledge associated with the organizational culture was much more ad hoc. There was no specific planning or discussion about how to transfer the culture; the startup objectives were strictly based on the strategic rationale for establishing a Mexican plant. A key factor in the successful transfer of the culture was that the manager running the Mexican operation was very experienced and therefore, there was an assumption that "he knew the culture and had the fit with Gamma’s way of doing things." As well, some specific, albeit unplanned, actions helped to facilitate knowledge transfer. The Mexican HR manager and financial manager visited the U.S. headquarters to learn how things were done. The workforce hired in Mexico were young and "moldable" and did not have much manufacturing experience. Also, the customer was unwilling to settle for anything less than the same quality standards as Gamma delivered in the U.S. The result was that the Mexican quality and delivery standards were as good as in the United States. However, the cleanliness and plant organization were not yet up to Gamma standards and people development processes were lagging. 

The three case studies provide an opportunity to study various types of knowledge transfers and knowledge transfer attempts. The following sections provide a linkage with the earlier discussion of the knowledge transfer model. 

Knowledge Spectrum and Contextual Dimensions 

We suggested earlier that when knowledge is extracted from its "natural" context and moved to a new context, this new context may or may not be in alignment with the transferred knowledge. Contextual alignment in turn is a function of five dimensions. The nature of the dimensions is that they enable and restrict knowledge transfer. In our case studies, we found that explicit knowledge, usually associated with manufacturing processes, was generally transferred successfully with low eventfulness. In the case of Alpha, the parent firm established very clear goals for the transfer of manufacturing knowledge and established a very receptive culture for the knowledge. The recipients in the Latin American country had the opportunity to visit the United States as well as other Latin American operations. These visits were designed to establish cultural similarity and fit with the new knowledge. The decision making structure adopted for manufacturing was a replica of one in another Latin American country. One of the plant managers was a long time Alpha employee from a neighboring country. He was deliberately chosen to manage the manufacturing knowledge transfer. The other plant manager, a European, was in a more difficult position because the scale of the plant was larger and the incoming knowledge more complex. In the case of Gamma, the manufacturing knowledge was also transferred successfully, primarily because of very tight coordination between the Mexican plant and the United States headquarters. 

With Beta, a continual challenge was transferring technology developed in the United States to international units. To create a successful technology transfer, there had to be evidence that the technology was relevant and could provide customer solutions. Crossing borders with the technology was always difficult because of a historical lack of trust between headquarters R&D and international operations. In recent years, the distrust was breaking down but it persisted in some areas. Drawing on our earlier discussion, successful transfer requires similar or at least related contextual elements in the source and recipient locations. With Beta, the contexts were often very different, which restricted the amount of knowledge that could be transferred. 

Although the explicit knowledge transfer in Alpha was proceeding smoothly, the more tacit knowledge transfer involved much more adaptation and experimentation. For example, there was great skepticism about the ability to adopt Alpha’s European system of supplier management. According to one manager, the environment was too dissimilar. This manager cited the poor transport system in the Latin American country, lack of understanding at the supplier level, need for warehousing of inventory, and no competition at the supplier level as evidence of environmental differences. Our observation is that a lack of understanding in Alpha’s Latin American operation about the rationale for adopting the system was an impediment to successful knowledge transfer. The more tacit marketing knowledge was not yet fully transferred. Our prediction is that significant difficulties will arise in this area. 

Knowledge Pockets and Transfer Success 

When knowledge overlaps the knowledge spectrums at both the source and recipient, successful knowledge transfer is more likely. Because Alpha was creating a new organization where one did not previously exist, initially there was no source knowledge spectrum. This spectrum had to be created from scratch. To do so, the knowledge spectrum created became an amalgam of Alpha’s knowledge from around the world. Sales and marketing knowledge came from two Latin American countries plus the United States; materials and logistics knowledge came from Latin America, United States, Europe, and a subsidiary company; finance knowledge came from Latin America and the United States; human resources knowledge came from the United States, Europe, and Latin America; and the general manager came from Europe via Latin America. These initial transfers, although not without problems, were largely successful. As this spectrum solidifies, knowledge transfer will become more difficult. Thus, we expect that that over time in Alpha, the overlapping area within both the source and the recipient unit’s knowledge spectrums will shrink, making future knowledge transfers more difficult. 

In Beta, the problem was that the knowledge to be transferred typically existed outside of the recipients knowledge spectrum. The new technology had to exist outside the recipient’s knowledge spectrum or there would have been no need to transfer it. Thus, to ensure successful transfer required a type of negotiation between the central R&D managers and the international subunit managers. As a manager indicated, "tech transfer is about interaction and begins with trust between two people." Applying our model to this statement suggests that the negotiation was designed to penetrate the recipient’s knowledge spectrum and create an overlap between the spectrums. This was done through the internal network of Beta and the personal contact-based system. 

Gamma was able to create a recipient unit with overlapping knowledge spectrums through the use of the specific knowledge transfer mechanisms discussed earlier. Gamma was also able, through their management selection and hiring practices, to create a motivated recipient. The challenge for Gamma in the future will be to transfer more complex knowledge. Plans were underway for a significant European presence. With Gamma, there was a great deal of concern about the transfer of organizational culture and specifically, human resource practices. 

This study represents an initial attempt at understanding knowledge transfers and knowledge management by focusing on the knowledge spectrum and the context within which knowledge exists. Our initial expectation was that context similarity would be the key factor in the success of knowledge transfer. To context similarity we would add the nature and extent of the knowledge transfer mechanisms employed. Although we were aware of the importance of knowledge mechanisms, and our prior research (Inkpen & Dinur, 1998) in the alliance area revealed this, we were surprised by the extent of the different mechanisms associated with different types of knowledge. In particular, in the Alpha case, we identified more than 50 types of knowledge and more than 15 different knowledge transfer mechanisms. Even in the Gamma case, where the knowledge transfer was relatively straightforward, a variety of very different knowledge mechanisms were used. 

We also found that explicit knowledge was easier to transfer than tacit knowledge, which suggests that knowledge complexity is a key variable. Therefore, we propose Figure 4 to illustrate the interrelationships between the variables that impact knowledge transfer success. Knowledge transfer mechanisms is shown in the model, along with context similarity, and knowledge complexity.