In their struggle to drive down manufacturing costs, many companies look to relocation overseas as a possible solution. The immediate labor cost savings from such moves can be tempting. But they also involve great risks. Siting a plant at a distance from other functions can damage the vital relationships that sustain an organization's proficiency in those critical, hard-to-develop business processes that deliver value to customers. Though often underrated, geographical proximity is a powerful ingredient in the mix of factors that keeps a company flexible and responsive.

Many plant location decisions are tactical, stop-gap actions which result in short-lived benefits and create facility networks that are unbalanced, inefficient, or ineffective

The transition to global competition in many US manufacturing industries has made deciding on facility locations more complex than ever. Unfortunately, few companies have responded well to this challenge. Many plant location decisions made during the past decade could be characterized as tactical, stop-gap actions which resulted in short-lived benefits and created company facility networks that are unbalanced, inefficient, or ineffective. The financial performance of the plants themselves has frequently been disappointing. Typical management comments include:

"Our savings were significant initially, but wage rates at our overseas plants have risen consistently, and at current exchange rates our savings are almost insignificant."

"We save a lot on labor over there, but we have also added a lot of new people here to support the foreign plants and brought in additional finished goods to support our customers."

"The process of communicating with our [South American] plant is incredibly burdensome. It is almost never in synch on engineering changes, and timely new product releases are next to impossible."

When location decisions foster the development of critical capabilities, the resulting plants fit into well-conceived facility networks which solidly support a company's business

There have, however, been some exceptions to this pattern. A few companies have made location decisions that have fostered the development of critical capabilities and created lasting competitive advantage. The resulting plants fit into well-conceived facility networks which solidly support the companies' businesses. Not surprisingly, the decision processes utilized by these firms are quite different from those of most others. By understanding the nature of these processes, and the key ways in which they differ from more typical processes, managers can:

• Make better plant location decisions through recognizing the importance of such decisions to the firm as a whole.
• Improve decisions about which products, processes, or organizations to relocate overseas, and appreciate what the implications will be.
• Facilitate interactions both between internal functions (such as design, manufacturing, and marketing) and with the company's external suppliers and markets.
• Support critical capabilities in the company that are an ongoing source of competitive advantage.

The traditional approach

In 1990, Kodak's Business Imaging Systems (BIS) division found itself facing increasing pressure from foreign competition. Japanese rivals were experiencing growing success with small, low-priced microfilm capture and retrieval equipment. To engineers and managers at Kodak, the design of these machines seemed "cheap": their frames were of lower gauge metal and tended to flex; power supplies and other components were smaller; and functionality was much more limited.

But these apparent shortcomings did not deter foreign customers and price-sensitive domestic customers from switching to the cheaper models, thereby eroding BIS's market share. Marketing management responded by putting pressure on manufacturing personnel, who they felt were responsible for producing products that were "not competitive." This pressure quickly grew as the division general manager also became concerned about manufacturing's cost position.

Faced with this issue, manufacturing management determined that changes to production methods and sourcing—the traditional arenas for manufacturing cost reduction—could not yield the savings required to compete with the Japanese products. However, the manufacturing director happened to have recent experience of starting up a plant in Taiwan to produce optical components for another Kodak division.

Others in the team had similar experience, making them aware of the large difference in production hourly wage rates between their US and overseas facilities (approximately $12 versus $3). They knew too that materials costs could be reduced by 10 percent or more in some locations. As the Kodak optics facility in Taiwan had excess capacity, manufacturing some products overseas seemed the potential solution to creating short-term savings of the size BIS needed.

A high-level manufacturing manager was asked to assess the impact of moving some operations overseas. He began by looking for a framework to help in the selection of plant locations, and by examining methods that had been used to make past decisions. These consisted primarily of models that applied subjective weights to a long list of factors such as political risk, labor availability, union activity, and infrastructure. The locations with the highest scores were then considered most seriously.

In BIS's early response to its competitive situation are many of the characteristics of a "traditional" approach to facility location:

• Static "snapshot" analysis. The analyses that underlie plant location decisions typically tend to be focused on current wage rates and material costs. They make a good attempt at understanding the costs and benefits of a location at a given point in time, but ignore how these may change in the future. Further, there is seldom any ongoing effort to examine the viability of the company's facility network; the location analyses performed are individual, unrelated reactions to independent internal or external stimuli.
• Stop-gap response. Plant location decisions, particularly those to locate overseas, are often made as a short-term response to a competitive threat. Overseas wage rates can seem miraculously low ($1–2 per hour in some countries versus $10–15 in the United States), and can look very tempting to a manager facing foreign cost competition. A new plant located in a low factor cost country can seem like an attractive way to level the playing field, at least in the short term.
• Single-function focus. The plant location decision is addressed almost single-handedly by the company's manufacturing organization, usually with only minor outside functional input to address pricing and tax issues. This high degree of autonomy leads to three main weaknesses in the decision-making process:
  • Overemphasis on direct labor costs. Largely because of traditional cost accounting systems, the manufacturing function has traditionally been obsessed with direct labor costs. The result: numerous plant location decisions based primarily on direct labor costs.
  • Focus on options implementable by manufacturing alone. Rather than looking for reductions in cost across the entire business system, such as design changes or overhead reductions, the manufacturing function tends to focus on significant changes that it can implement directly. Once again, a plant relocation will appear to offer the potential for a "quantum leap" reduction in manufacturing costs.
  • Neglect of impact on other functions. Since the plant location is decided by individuals who are most familiar with manufacturing issues, they often underestimate the impact that a decision may have on the company as a whole.
• Focus on external environment and resources. Traditional decision analyses focus extensively and almost exclusively on the characteristics of the locations being considered: wage rates, material costs, utility costs, transportation costs, infrastructure, taxes, and so on. Analyses that address the impact of the facility location on the internal workings and "capabilities" of the company are much rarer.

The capability-focused approach

In the dynamic competitive environment facing most manufacturers today, products are quickly obsolete and static competitive positions are rapidly overtaken. In such an environment, the only real source of lasting advantage is the ability to respond consistently to changing markets with new products and ever-improving competitiveness. A firm can achieve this constant renewal by identifying, developing, and maintaining its critical "capabilities."

Capabilities are a company's proficiency in the business processes that allow it to distinguish itself continually in ways that are important to its customers. For example, in a high-tech industry, the ability quickly to develop new state of the art products builds an enduring advantage. In a commodity industry, it may be the ability constantly to reduce costs that creates the lasting competitive edge. Capabilities like these reside in the company's people and are supported by its procedures, culture, and infrastructure. Firms seeking sustainable success in today's competitive environment must focus on building and nurturing these capabilities.¹ Exhibit 1 presents examples of successful companies and the capabilities that support their success.

A capability clearly cannot create a competitive advantage if it is easily achieved by competitors. Thus the critical capabilities are those that are difficult to develop. Three common features contribute to difficulty:

• Complexity. Critical capabilities tend to be developed in business processes that are highly complex and involve patient "organizational learning" over a long period of time. A company with a head start in this learning process is difficult to overtake.
• Organizational diffuseness. Critical capabilities involve processes which nearly always cut horizontally across functions in the company, and frequently involve external groups (see Exhibit 2). As a consequence, they cannot be developed by hiring away a few key individuals.
• Well-developed interfaces. Critical capabilities depend as much on the way that individuals and organizations have learned to work with each other as they do on the particular expertise of the individuals and organizations themselves. A competitor could staff each function in a critical process with the most talented individual available and still not develop a capability.

Developing and maintaining critical capabilities demands dynamic and responsive relationships between all the groups involved in critical processes. Rich, constant, and often informal communication must replace the arm's-length, periodic, and contract-oriented relationship that frequently prevails among such groups. This type of association can be extremely difficult to conduct at a distance. Thus the configuration of a firm's facility network—the geographical proximity of each group to the others—can support or undermine its critical capabilities. If facility location decisions are to result in benefits that are sustainable, this need for co-location must be recognized and addressed.

Like the traditional approach to facility location, a capability-focused approach works to exploit the costs and benefits of external conditions and factors, but it differs in also attempt- ing to foster the relationships, both internal and external, that sustain critical business processes. While the traditional approach takes a static view of conditions and benefits, the capability-focused approach tries to capture benefits that are dynamic and self-renewing. The traditional approach tends to view plant location decisions from the lone perspective of the manufacturing function, while the capability-focused approach takes the cross-functional business process view. Finally, the objective of the traditional approach is often tactical competitive improvement, whereas the capability-focused approach targets long-term strategic advantage.

Traditional analyses often include complex mathematical models that balance factor costs, transportation costs, taxes, and so on to determine the location which will minimize a firm's total cost. Exactly the same kind of balancing and optimization algorithm is required by the capability-focused approach, but different variables are included. For instance, certain benefits are derived by locating a company's functions near external entities like customers, suppliers, competitors, and governments. Other benefits are derived by locating individual functions near one another. The magnitude of the benefits differs by industry and by company.

Achieving capability-driven financial benefits may take several years, but once achieved they will endure

The timing of benefits is also important in determining the overall return. Achieving capability-driven financial benefits may take several years, but once achieved they will endure. The financial benefits from relocating to a low factor cost country are virtually immediate, but transient. The objective of the analysis should be to identify the network that combines these benefits in such a way that the net present value of future cashflows is maximized.

Two approaches contrasted

The cases of two high-tech companies from the past decade provide a good illustration of the traditional and capability-centered approaches to the plant location decision.

Seagate Technology, Inc.

Seagate Technology is recognized as the firm that made hard disks for personal computers affordable by bringing volume production and in-house component manufacturing capabilities to the industry. During the early 1980s these capabilities secured the company a dominant position in the market. By the summer of 1984, however, Seagate faced crisis. Falling PC sales had forced its customers to slash their orders. Seagate's largest customer, IBM, was insisting on a price cut, and had begun to manufacture some of its own disk drives. Seagate cut the price of its principal product by 45 percent in order to remain competitive.

The company's president, David Mitchell, responded with a series of actions which many at the time felt saved the company. Moving disk drive assembly offshore to Singapore was the most significant of these changes. The result was an almost immediate 30 percent cost reduction. In fiscal 1985 the company was able to show a small profit, even though Seagate's sales had dropped to $215 million from $344 million the previous year.

Exhibit 3 shows Seagate's profitability before and after this decision. Although the company did very well in the two years following the overseas move, subsequent performance has been disappointing.² The headline of an April 1991 Wall Street Journal article about the earnings of Seagate and a major competitor proclaimed "Connor Peripherals Says Profit Doubled; Seagate Earnings Fall." During summer 1991 Seagate dismissed 1,650 people, more than 20 percent of its non-manufacturing workforce, in an effort to cut costs in response to further price competition. In September 1991 David Mitchell abruptly resigned as president and chief operating officer, apparently as the result of pressure from the board of directors.

Many of the forces that brought about the sharp change of fortunes for Mitchell and Seagate from the mid-1980s to the early 1990s are more easily understood through an examination of the company's approach to the decision to move manufacturing overseas.

Static "snapshot" analysis. The primary driver of Seagate's relocation overseas was the competitive cost advantage that could be achieved through the significantly lower wage rates available there at the time. Unfortunately, other disk drive manufacturers soon followed Seagate to Singapore, reducing its initial advantage. The ensuing competition for production workers resulted in double-digit percentage annual wage rate increases and the need for employee benefits like childcare reimbursement and free medical care.³

In 1987 Seagate built a new disk drive assembly facility in Thailand, where wage rates were again much lower (starting wages were approximately 50 per hour). By the middle of 1988 this facility was assembling a majority of the company's drives, but the phenomenon of increasing wages was quickly repeated here too.

This pattern—a company moving to a low labor cost location, only to be forced to move again as costs rose—has been a common one in industries as diverse as disk drives, toys, textiles, and ball bearings. Many companies in these industries have made a series of decisions each geared to finding the location that minimized product manufactured cost at the time. When conditions changed, the advantage vanished.

Where one company can relocate for a wage advantage, others can follow; the result is a self-perpetuating cycle

As these companies have discovered, where one company can relocate for a wage advantage, others can follow; the result is a self-perpetuating cycle. For example, maquiladoras—assembly plants near the US/Mexican border—have seen wage rates rise as companies compete for local skilled labor. Currencies in Taiwan, Singapore, and Korea have all shown sustained appreciation (between 25 and 50 percent) in the second half of the 1980s, increasing local costs when they are converted to US dollars. These changes have been in part the natural result of the economic success that foreign investment and increased national income have brought. By building plants in low factor cost countries, US companies alter the very conditions that initially attracted them, while contributing to the development of these countries' economies.

By building plants in low factor cost countries, US companies alter the very conditions that initially attracted them

Stop-gap response. In moving manufacturing overseas to Singapore, Seagate responded to immediate and pressing competitive threats. It was rewarded with a quick cost reduction. At the same time, however, the company was affecting its ability to compete long term. In particular, it was inhibiting its ability to develop and introduce innovative new products in a timely manner.

Single-function focus. Seagate's decision to move was based almost exclusively on manufacturing concerns and issues. The decision was driven by David Mitchell, who had prior experience managing the Asian plants of Fairchild Camera & Instruments Corp., and who was described in a 1987 Fortune article as "a manufacturing man." As mentioned earlier, he gave little consideration to the impact of the move on the other functions important to the company's long-term success, such as the vital link between product and process technologies.

Focus on external environment and resources. For Seagate, the benefits of moving manufacturing to Singapore derived entirely from the economic conditions and resource costs there; the impact of the move on the company's internal capabilities was scarcely considered. Rather than enhancing Seagate's capabilities in low-cost manufacturing, the move only gave it access to lower labor costs—a benefit instantly available to any other company that followed suit. Evidence suggests that the move weakened, or at least exacerbated an existing weakness in, Seagate's capability to maintain the pace of technological development that success in this industry required.

Throughout the late 1980s and early 1990s Seagate has been months—and in some cases over a year—later to market with significant new technologies than its competitors. This problem is the primary cause of Seagate's recent poor financial performance, and remedying it is the focus of the company's current improvement efforts.

Applied Materials, Inc.: An alternative approach

In the late 1970s, James Morgan, the CEO of Applied Materials—a US producer of semiconductor manufacturing equipment—became dissatisfied with his company's progress in the Japanese market.⁴ At the time, his company was using the Japanese trading company Kanematsu Electric to sell in Japan, with unimpressive results. As he analyzed the situation, Morgan came to the conclusion that succeeding in Japan would be critical to his company's long-term worldwide success. In the process of learning to be successful in the demanding Japanese market, the company would, Morgan felt, develop capabilities with global benefits.

Semiconductor manufacturing equipment is highly complex, involves state-of-the-art technology, and is closely geared to customer requirements. These requirements must be quickly and thoroughly understood and then satisfied using the latest available technology. Success in Japan was thus based on two factors, technology and relationship.

Applied's limited success in Japan at that time depended entirely on technology. With its trading company arrangement, Applied would never be able to develop the critical customer relationships needed for success in the market. A long-term commitment to the Japanese market through local direct operations was required in order to build lasting relationships with Japanese customers.

Acting on these concerns, Morgan established Applied Materials Japan (AMJ) in late 1979. Initially, he set up only a sales operation staffed with local nationals, but the benefits were significant. In its first year, AMJ doubled the sales achieved the year before by the trading company. By 1983, AMJ provided one-third of Applied's total sales, and Morgan was ready to take the next step.

Applied began construction on the Narita Technology Center in order to bring both R&D and manufacturing operations to Japan, thus improving the level and responsiveness of its support to the market through the use of customer-focused design teams and the development of in-depth customer knowledge. The company as a whole also had the opportunity to benefit technically from direct exposure to the world's most demanding customers and the industry's latest breakthroughs.

By 1990, a time when most US competitors, still relying on local trading partners to sell in Japan, were having very little success in this growing market, Applied Materials was the Japanese market leader in most of its technologies. It was also the only non-Japanese company in the top five semiconductor equipment suppliers in the world. All five had been US companies in 1980.

This approach to facility location is altogether different from Seagate's. For Applied Materials, the decision was based on its view of the strategic market direction, not current competitive issues and resource prices. It was a response to a projection of future conditions, not a snapshot of the current situation. It was made by a CEO focusing on marketing, product, and operational issues, not by a manufacturing executive concentrating on direct labor rates. Finally, Applied Materials' decision was based on the capabilities that conditions in the new location could help the company to build, not on the conditions themselves.

The goal of a strategic approach to facility location is to build a network of capabilities, rather than just a network of facilities

The goal of this strategic approach to facility location is to build a network of capabilities, rather than just a network of facilities. This network is specifically designed to support the company's business and market strategies on a continuing basis. It is structured as much for what it can "teach" the firm as to take advantage of external location resources. It is dynamic and maintained as an ongoing element in a company's strategic plan.

Key steps to a network of capabilities

Exhibit 4 outlines an approach for first identifying a company's critical capabilities and then developing a facility network that will support them. Step 1 is to examine the company's distinctive competitiveness and the nature of the markets in which it competes. The objective is to identify the value that the company intends to deliver to its customers: low-cost, leading-edge technology, extensive customer service, or whatever.

Step 2 then seeks to identify the business processes in the company that can give it an advantage over its competition in delivering this value. These are the critical processes in which capabilities can exist.

Step 3 examines these processes and maps out where they cross the boundaries between internal functions or with external groups. The relationships that exist at these boundaries are then scrutinized to determine whether they are critical to the process, and if they are, whether the relationships require geographical proximity for success.

In step 4, alternative plant locations and sequences of locations are pinpointed, with the objective of best balancing the need for proximity of each of the critical relationships.

Finally, in step 5, a return on investment model is utilized to assess the financial implications of each alternative.

1. Identify key sources of customer value

A manager addressing the facility location decision should begin by listing the products or product families supported by the facilities under consideration. For each product, he or she should identify the targeted customer groups and their key buying factors—the product attribute dimensions which differentiate one product from another. In a true commodity market, for example, the only dimension is cost. For a buyer of disk drives, dimensions could include cost, capacity, speed, size, and reliability.

The manager should then assess the relative importance of each dimension to the target customers, and analyze the positioning of the company's products relative to the competition on each one. He or she should thus determine the attributes which dis- tinguish the product in the eyes of the customer.

In the case of Applied Materials, the value delivered to customers was a leading-edge technology design combined with a high degree of product customization. For Seagate customers, although the factors associated with a leading technology design were important, the distinguishing attribute was cost. The value Seagate delivered in the early 1980s was a drive that met all the form and function requirements at the lowest cost. As the market matured, Seagate was no longer able to meet its customers' most important requirements as quickly as competitors, and consequently a significant proportion of its value for customers disappeared.

2. Identify critical capabilities

This step should begin with the identification of the business processes that enable the company to create and deliver the product attributes which are important to customers. Particular, though not exclusive, attention should be given to the attributes that distinguish the company's products from the competition and influence customers to select them.

The processes that create lasting competitive advantage may be different from those that seem most important today

It is vital to focus on the processes that create lasting competitive advantage; these may be different from the processes that are seemingly most important today. For example, if customers buy a particular product because it is the least expensive, one might conclude that its efficient, automated manufacturing process is the source of the value for the customer. However, when this product is obsolete, or when the competition imitates the manufacturing line, the process, though still intact, will no longer be a source of value.

On the other hand, a design process for low-cost products, or a capability in developing automated manufacturing processes, or even organizational processes that drive continual cost reduction, can deliver products with a competitive cost advantage on an enduring basis.

Competitors should not be able to match a critical process by hiring away key individuals or through heavy investment

Once the processes that create value have been identified, the "critical" processes should be isolated—those that cannot be easily duplicated by the competition. Competitors should not be able to match the process by hiring away key individuals or through heavy investment.

In businesses like software design, the work of a few talented individuals may create nearly all of the value that the customer perceives in the product. Here the critical capabilities that the company possesses may be the processes and systems that create the working environment and the rewards that keep these individuals motivated and content. Such a capability provides an enduring competitive advantage for the company over competitors who would not be able to retain these essential employees. Processes like these are complex, involve individuals throughout the organization, and are not easily imitated.

For Applied Materials, the process that delivered value to customers was the product development capability, which includes working with customers to understand their requirements thoroughly before pushing technology to provide a state of the art solution.

3. Highlight key links

Having identified the critical processes around which capabilities exist, the manager can identify the key interfunctional and external links required to support them. First the manager should construct a simple flow chart that depicts the major steps in the process at a high level, and includes the material and information flows between these steps. The next task is to identify the locations where the process crosses boundaries either between functions or between the company and outside groups.

Particular attention should be given to links with demanding outside groups which can give the company insight into future competitive requirements in the industry. These might be leading-edge customers whose requirements anticipate the future needs of other companies in their industry (for example, Japanese semiconductor manufacturers for Applied Materials). They might be a government agency in a regulated environment, or key technology suppliers in a fast-moving high-tech industry (for example, Intel for a personal computer manufacturer).

The manager must then assess how critical each interfunctional and interorganizational link is by identifying the major types of failure a link might exhibit—lack of communication, communication delay, introduction of errors, loss of information, and so on—and the probability of failure under normal circumstances. The impact of each failure on the value that the process delivers to the customer should then be assessed. If a significant impact is at all likely, the link must be considered critical.

Common critical links include that between development and manufacturing for a high-tech product and that between development and marketing for a consumer product. Companies in different industries have different critical links. For Applied Materials, the link between R&D and the demanding Japanese customer base was essential for global success. Similar critical links can exist with key suppliers, government agencies, financial institutions, and so on.

The need for geographical proximity in the critical relationships should then be evaluated. Several factors influence the level of need:

• Complexity of information. The more complex the information that must be regularly communicated between groups, the greater the need for proximity. By its nature, complex information is difficult to communicate. Communication may entail multiple cycles, several different media, or even eye contact to ensure that the message has been understood. Complex information is also more easily communicated incrementally over time than it is in large batches. Proximity facilitates all these practices.
• Required level of interaction. Communication that involves significant interaction needs proximity more than communication that does not. Factors affecting the level of interaction include the predictability of the communications content (can similar information be communicated in a similar format month after month, or does it exhibit more variability?) and the direction of information flow (one-way or two-way?)
• Similarity of background and expertise. It is much easier for individuals with similar backgrounds and experience—people who "speak" the same language figuratively or literally—to conduct a relationship at a dis- tance. Where differences do exist, it can be difficult for one individual to know what information is important to the other, and what is not. When people are distant, they are forced to shape the information they communicate to fit the communication opportunities available. They may unknowingly filter out important information.
• Requirement for trust-based relationship. In many cases, particularly where external parties are concerned, effective communication means that individuals must trust one another. Typically there is no obligation on the part of one individual to communicate to the other; there may even be disincentives. This type of relationship is nearly impossible to develop and maintain at a distance.
• Concreteness of information. Some information is difficult to verbalize. Customer feelings and the subtleties of a foreign culture are good examples. In cases like these, there really is no substitute for being there.

In matters of communication it is tempting to think that technology provides a solution. In reality, electronic mail, facsimile, and telephone communication work well only after initial relationships have been established. Even then they cannot support the rich and constant communication that is needed in the relationships which support capabilities.

For Applied Materials, the complexity of the interaction between several internal functions (sales, design, manufacturing) and with its most demanding customers required geographical proximity to be effective.

4. Identify future progressions

The manager should plot each of the major links between important internal and external groups on a two-dimensional map. The dimensions are current level of proximity and the need for proximity identified in step 3. Clearly the ideal area is the upper right-hand one, and the worst is the lower right.

The ideal progression of moves can now be identified. There will probably be a large gap between the current network and the ideal network that co-locates all links with a high need for proximity. Some groups cannot be moved easily, either because they are external to the company or would be expensive to relocate. Since it is unlikely that they can be moved, these groups act as anchor points—the fixed points around which the network must be built.

At this stage, the manager should identify high-leverage moves around anchor points—ones that create maximum strategic value and move toward the ideal network. They should be considered as an evolution over time. At Applied Materials, James Morgan elected initially to move sales to Japan (see Exhibit 5). This initiated the development of direct relationships with the local customer base, and sales in the market grew dramatically as a result.

The links between development/manufacturing and the Japanese market were also critical, but the market was too small to support their presence initially. After the market growth driven by the move of sales to Japan, however, the problem disappeared, and the relocation of development and manufacturing soon followed.

It is important that the analysis should extend beyond the next move to include potential sequences of moves that could take place over several years. The move after next (or even the move after that) may influence what the best next move should be. If a company has a long-term objective of entering a previously untapped market with sales and manufacturing operations, it might make sense to locate a small design team there in advance of the other groups to develop a customized set of products.

Optimizing a sequence of moves will drive results superior to those achieved by considering moves individually

This "next move" makes sense when considered as a step in a sequence, but not in isolation. Tax laws, political considerations, changing economic conditions, and learning requirements can all create situations where optimizing a sequence of moves will drive results superior to those achieved by considering moves individually one after the other.

5. ROI modeling and sensitivity analysis

A return on investment analysis should be used to quantify the financial implications of alternative facility networks, and should incorporate as accurately as possible their costs and benefits. When functions or groups that require a high degree of interaction are distant, inefficient communication and higher costs will result. Conversely, strengthening the network of capabilities may lead to increased margins or sales, and these financial impacts may also be modeled.

While such figures can be included in the ROI calculation, a careful qualitative analysis should be used as a "reality check" of the results. To be accurate, this ROI analysis must meet several requirements.

• The ROI model must take a cross-functional perspective, including costs and savings to the company as a whole. For instance, an overseas plant will generally cause costs to rise for the distribution function, as a company with predominately US customers has to pay for additional duties and longer shipping distances, and fund a larger inventory pipeline. The company must also consider the substantial additional costs that will be incurred by almost all functions because of added complexity. These costs include currency management, adherence to local regulations, and expatriate support, plus all other costs created by the introduction of distance, including differences in time zone, language, and culture in both communications and logistics channels.
• The model must also account for changes in local economic conditions. The manager must forecast future exchange rates by using a simple model based upon purchasing power parity, interest rate parity, or return to real exchange rate equilibriums. This information must be used to assess both the company's and its competitors' situations over time; even purely domestic producers have exchange rate exposure since their overseas competitors' cost structures will be affected by exchange rate shifts.⁵ Political risk assessment must also be incorporated into the analysis.
• Wage rates, taxes and transfer pricing, and overhead savings must be handled correctly. These aspects of the analysis present three complexities. The first is the choice of wage rate to use in the model. Raw local labor costs are deceptive, since they do not account for the lower productivity of workers in low factor cost countries relative to US workers. Accuracy entails adjusting labor costs for the relative productivity within the relevant economic sector of the country. Changes in local wage rates over time should also be estimated, since there have been sharp rises in many developing countries in response to foreign investment.

The second area of complexity is the choice of effective tax rate and transfer price. This is important to the final result, and a successful manager will need to involve tax staff early to understand and fully exploit all opportunities. Finally, in evaluating overhead savings, only those savings that are relevant for the company as a whole should be included. Fixed costs such as buildings and systems do not disappear directly when a plant moves overseas, although the company's cost accounting system may at first suggest otherwise.

One end result of this analysis should be an understanding of the external conditions under which a network does or does not make sense. Another should be the identification of the level of benefits (for example, improved future sales or margins) which must be derived from improved internal capabilities to justify the costs of the alternative locations. Management can then make a judgment on each alternative's likelihood of success and level of risk. It can also make an informed decision on the short-term network configuration and its planned future development. If the results of the analysis are not satisfactory, management can revert to the previous step to search for a network that improves the level of high-leverage co-location and produces good returns on investment.

A change of approach for BIS

The manager charged with the BIS plant decision completed his review of past site selection methodologies and judged them inadequate. They seemed to him excessively subjective, since they failed to identify the impacts of various site alternatives on BIS's business, and ignored their effect on key business processes. The manager set about finding an approach that would address the real impacts of potential plant locations on BIS and on Kodak as a whole.

He decided to undertake an ROI analysis of the proposed foreign plant locations to understand the full economic picture. What he discovered surprised him: the products BIS produced did not have the scale or direct labor content required to make such moves pay off. There were three main reasons:

• Labor rates. Although raw labor rates were low, productivity was as little as 20 percent of US levels. Thus, for example, although the basic wage rates in Singapore in 1988 were US$3.09 per hour, the productivity-adjusted wage rates were US$6.40 per hour.

• Overhead structure. Only about 35 percent of the total allocated overhead dollars represented hourly direct labor; the rest were either fixed costs, such as tooling and information systems, that were not affected by an overseas move, or salaried manufacturing support (typically, experienced engineers). Overseas wage differentials were much smaller for engineers than for assembly workers.

  In Malaysia, for example, experienced engineers earn about M$4,000 (US$1,600) per month and typically also receive use of a car and housing subsidies. Hourly employees often earn as little as M$400–800 (US$160–320) per month. Savings on overhead costs driven by engineering wage rates were thus much smaller than those driven by the wage rates of hourly assembly workers.

• Cost increases over time. In many of the proposed sites, projected currency appreciation over time resulted in increases in labor and material costs. Even in countries with no such appreciation, the results were affected by what Ken Ohmae refers to as the "expense of cheap labor." As he notes, "The chip-makers have learned first-hand what CTV and textile industries discovered earlier: inexperienced labor must be trained and, once trained and experienced, labor does not stay cheap very long."⁶

Following these insights, BIS began to develop their approach from a fairly traditional one to one that was more capability centered. As the team delved into the details of what a Taiwanese plant would mean to their business, they found:

• The designs of the low-end products had excessive functionality and ruggedness, both of which drove excess cost. The products were also too large to match global customer requirements well. Moreover, a large segment of domestic customers was changing its orientation from features to value.
• Successful competitors' products in the US market were not manufactured in low factor cost countries.
• The core capability set implied by the division's competitive strategy required that manufacturing and engineering be co-located, and a transplant engineering outpost supporting the foreign plant did not seem economically feasible.
• Cost advantages that were at least equal in magnitude to the benefit of low foreign wages, and were more durable, could be achieved without moving, through improving the manufacturability of product designs.⁷

On the basis of these findings, the Kodak facility location team grew to encompass multiple functions including engineering and marketing. The team decided that although a low factor cost manufacturing site for current products made little sense, an Asian presence was important to improve the company's understanding of the requirements of this significant high-growth market. They also felt the competitive Asian environment and skilled engineering workforce could help Kodak develop its internal low-cost manufacturing capabilities.

BIS eventually elected to move a few engineers to Taiwan to coordinate the development of a low-end product for the local market. The development effort itself was staffed with local engineers. Design is proceeding on schedule. A small-scale manufacturing operation dedicated to this one product will be progressively initiated as it is needed to support the engineering effort and eventual production volumes. In parallel, BIS is working to improve the design of its current US-manufactured product line to increase cost competitiveness and ensure a better match to global customer requirements.

A network of capabilities will support a company's competitive advantage on an enduring basis

A capability-focused approach to the facility location decision can be used to create a network of capabilities that includes plants as an integral part. The resulting networks will support the company's competitive advantage on an enduring basis. Broadening the analysis to include non-manufacturing issues will actually help to enlist the manufacturing organization as a true contributor to the company's strategic capabilities.

This capability-focused approach is far more difficult to implement than the traditional one, and the financial improvements may not appear as quickly. The end results, however, are lasting and cumulative. They can transform a company's operations from a reactive, defensive stance to a key element in a premeditated strategic plan.