Pittsburgh, November 1994 

Managing International Research and Development Networks

 by 

 Gerhard Barth, Daimler-Benz AG, Germany 
Barry Baum, Skil-Bosch Power Tool Company, U.S.A. 
Dino Bongini, Merloni Elettrodomestici spa, Italy 
Mario Cipriani, Merloni Elettrodomestici spa, Italy 
Theo Claasen, Philips NV, The Netherlands 
Ron Foltz, Hoechst Celanese Corporation, U.S.A. 
Jan Gerritsen, Clorox Corporation Inc., U.S.A. 
Ulrich Hackenberg, Audi AG, Germany 
Gezinus Hidding, Anderson Consulting, U.S.A. 
Gary Katzenstein, Carnegie Mellon University, U.S.A. 
Jurgen Idzko, Hoechst Trevira, Germany 
Volkmar Leutner, Robert Bosch GmbH, Germany 
Bruce McKern, Carnegie Bosch Institute, U.S.A. 
Joe Profeta, Union Switch & Signal Systems, U.S.A. 
Manfred Schmidt, Vermont American Corporation, U.S.A. 
Heinz Schulte, Carnegie Bosch Institute, U.S.A. 
Jan Bodo Sperling, Coverdale, Germany 
Ralf Speth, BMW AG, Germany 

Edited by Gary Katzenstein, Ph. D. Candidate, and 
Bruce McKern, President, Carnegie Bosch Institute (c) 1995 

Introduction

 Equally important in today's environment is a company's ability to internationalize successfully. With global competition increasing, a company's ability to compete successfully in global markets is crucial to its growth and even to its survival. Given these two important facts of modern corporate life, it's not surprising that many companies have been trying to wed the two successfully to manage international R&D to tap global markets that are driven by fast-emerging technologies. But managing international R&D is a challenge. Given the communication, cultural and coordination problems, managing any worldwide organization requires considerable discipline and responsiveness. Adding the uncertainties of high technology development and competition makes managing a complex hybrid--International Research and Development--particularly challenging. This report examines key issues in this important area, describing common patterns and problems, and suggesting solutions for some difficult dilemmas. 

 The report is based on a week-long seminar of international R&D executives held at the Carnegie Bosch Institute at Carnegie Mellon University. With its blend of input from both experienced managers and academic researchers, the report offers a realistic examination of the issues and challenges that the conference participants consider important in improving the performance of an international R&D network. 

Goals of this Paper

•  What organizational structures can improve the management of international R&D ?
• What organizational processes facilitate the management of international R&D? 

Structural Issues

• Centralized vs. Decentralized R&D Organizations 
• The relationship of R&D to Corporate Headquarters

Process Issues

• Role of R&D in Strategy Formulation
• Transfer of Personnel Between Laboratories
• Transfer of Ideas Between Labs and Business Divisions
• Duplication of Research Efforts Across Centers
• Standardization of Products
• Evaluation of R&D Effectiveness
• Speeding the Product Development Process

Research and Development

 Just as the problems and solutions relevant to research may not apply to development (and vice-versa), various prescriptions about R&D do not apply to all situations. Rather, numerous factors may affect the success of a given recommendation. They include: 

• The pace of change in the industry 
• The scope of business (global vs. national vs. regional)
• The size of the company
• The technological intensity of the industry 
• The market orientation (consumer vs. industrial)

Although this report does not cover all the implications of these distinctions, many will be mentioned. 

Overview of the Report: Key issues in managing international R&D

Much has been written separately about international business and R&D, but this report attempts to show how the two come together and what is needed to manage international R&D effectively. The report focuses on the key issues identified by the Forum group. A brief overview of these key issues is offered here. 

1) Why Internationalize?

Since companies could more easily maintain the R&D function in their home base, with lower expenses and coordination effort, there must be good reasons that firms choose to internationalize R&D. This section highlights some of those reasons. 

2) R&D and the Strategy Process

R&D projects often change the strategic direction of a company. Thus, it is fitting that R&D should play a major role in setting corporate strategy and helping implement that strategy. Key issues that this section examines are when R&D should be involved in setting corporate strategy, in what way, and who should take the key roles in the strategy-setting process. The role of R&D in identifying trends, evaluating technologies, and choosing and building on core competencies is also explored. 

3) R&D Organizational Structure

R&D is an organization within an organization, and its structural relationship is important. This section investigates briefly the advantages and disadvantages of centralized and decentralized R&D organizations. Since many organizations choose to decentralize, at least to some extent, this section also looks at problems that accompany decentralization, such as duplication of effort and the pressure to move away from standard product formulations. 

4) R&D Transfer in an International Setting

When R&D organizations decentralize, they must learn how to coordinate their various labs and transfer ideas and results throughout the organization. This section considers both the vertical transfer of results from basic research to product development, and the horizontal transfer of information among labs. In addition to examining the movement of ideas, this section discusses the transfer of people, another mechanism of organizational transfer that allows ideas to be exchanged and which strengthens the R&D organization. 

5) Operational Issues

Once the major structural and process choices have been made regarding international R&D, the company needs to monitor and measure its operations. This section examines how firms evaluate R&D effectiveness, and how companies can shorten their product development cycles. 

The sections that follow expand on each of these issues. 

I. Why do R & D internationally?

R&D activities need not necessarily be done internationally, even by companies that do business internationally. Rather, a company could choose to do all its R&D in one or more domestic sites. Doing so would avoid some of the disadvantages of doing research internationally: 
•  Multi-site research is more expensive and more difficult to coordinate.
• Duplication of effort and resources becomes more likely.
• Language and cultural issues prevent smooth communication and transfer of results.
• The need to allocate resources creates internal political and economic difficulties.

 Despite these problems, there are advantages to doing R&D internationally. 

These include: 
1) To be closer to government and those who set product standards 
Many products such as food, drugs, clothing, and electronics are subject to government standards and regulations that differ from country to country. It is often difficult to stay in touch with those standards if one's lab is overseas. Rather, doing research in the country in which the standards apply is a more reliable way to stay current with changing laws and regulations that eventually affect product development. This consideration applies as much to basic research as to development, since government regulations affect not just end products but the fundamental technologies used in those products. 

2) To allow connection to production sites 
Companies co-locate their development sites with their production site to ease coordination and improve communication between the two functions. Since production sites are often internationally sited (for example, to be able to take advantage of lower production costs), development labs also need to co-locate internationally. 

3) To take advantage of lower costs in other countries 
Often R&D costs are lower in other countries. An example is the burgeoning software development center established by American companies in Bangalore, India. Of course, one should move R&D overseas to reduce cost only if adequate skills are available; low quality R&D is not an option. However, this strategy has several disadvantages: 

•  Due to lower skills overseas, the company may need more people to do the task, even though overall it will still be less expensive than if done domestically. This requires increased coordination and management.
• The company needs to transfer the results back to the home country, which may increase costs, and add to coordination and perhaps transport problems.
4) To be close to customers so as to understand their preferences 
In certain industries, particularly those that produce consumer products, companies need to be in touch with the local culture to be able to develop products that will gain local acceptance. Proximity is more important for development activities than basic research, since the former are closer to the customer. For companies supplying products to industrial customers, the R&D center may need to follow the customer to foreign locations. 

5) To be able to take advantage of local knowledge resources and infrastructure 
Companies may locate in specific foreign sites to take advantage of local knowledge resources and infrastructure that do not exist in the home country. Some of these resources include: 

 
• Special skills needed for R&D that exist in a given region (e.g., taking advantage of integrated circuit design skills by locating in Silicon Valley) 
• University centers and other companies (ÒclustersÓ) doing related research. These might lead to alliances and other forms of technology transfer.
• Leading companies (including competitors) in an industry that enable firms to stay in touch with key trends and new developments.
6) Multiple sites allow more ideas to be generated 
Involving researchers with varied international backgrounds may generate more diverse ideas than if all ideas were generated within the same country or culture. These different ideas can then interact with and compete against one another, leading to better innovation. Nevertheless, the price paid for multiple sites is some duplication of personnel, equipment, effort, and costs. 

7) Lack of intellectual resources in home country 
A lack of adequate intellectual resources can cause companies in small countries to look elsewhere for the highly-trained minds needed for R&D. 

These are some of the strongest motivators for expanding a firm's R&D organization internationally. These basic themes will be elaborated throughout the report. 

II. R&D and the Strategy Process

One key role R&D can play at the corporate level is to help formulate and implement corporate strategy. R&D can offer key insights for strategy formulation because its work involves developing technologies and evaluating technologies and markets for new products. This section looks at R&D's role in both formulating and implementing strategy, both at the corporate and Strategic Business Unit (SBU) or divisional levels. 

Strategy Formulation at the Corporate and SBU Levels

Many of the issues associated with R&D's role in formulating corporate and SBU strategies are summarized by four related questions: 
•  What should the role of R&D be in setting strategy?
• How can R&D best integrate its input with the firm's strategy?
• How should the strategy for R&D be determined?
• How does one choose appropriate R&D projects?
Strategies need to be formulated at both the corporate (i.e. headquarters) and business unit levels. Necessary elements of corporate strategy include: 
• A clear mission statement: what the corporation as a whole wants to do
• A broad sense of the corporation's financial goals
• An appreciation of the corporation's strengths--its core competencies, and the contribution it can make to SBU or divisional performance.

 Business unit strategies require: 

• A mission statement for the business
• An analysis of the environment, including a clear view of the markets and market segments in which the firm will operate
• An understanding of the roles and capabilities of competitors
•  A decision about how global or local the business's operations and sales should be

II. A. Structural Links Between Corporate Headquarters and R&D

To foster meaningful dialogue between Corporate Headquarters and R&D, good communication channels must exist so that each group is aware of issues and developments in the other's environment. Surprisingly, R&D directors often do not know their companies' corporate strategies, or are not sure of the details. These deficiencies makes it difficult for R&D to contribute to the development of corporate strategy. First and foremost, therefore, the lines of communication between Corporate Headquarters and R&D must be clearly established. Most heads of R&D felt that they had a better sense of what strategies were being taken at the individual SBU level, enabling them to play a more informed and useful role in setting SBU strategies than in setting overall corporate strategies. While it is obvious that R&D in most companies is important to SBU success, there is a need to define how it can contribute to competencies that are common to many SBUs, which is a corporate matter. 

Another issue to be decided is how involved R&D should be in setting corporate or SBU strategies, and at what stage R&D's involvement is appropriate. This may depend on whether the company is mainly technology or market-driven. In technology-driven companies, involving R&D from the start and giving it a substantial role is desirable. For market-driven companies, R&D should become involved later in developing strategy that supports market decisions. One solution that more companies are using is the Òconcurrent model,Ó in which R&D works with other departments simultaneously to formulate a jointly-acceptable strategy. Indeed, in many cases, Corporate and SBU strategies influence R&D strategy and vice-versa. Because they are mutually dependent, business and R&D strategies should be consistent and set concurrently. 

Finally, in highly-diversified companies, individual SBUs cannot set their strategies in isolation because their goals may not be best for the company overall. Corporate strategy must be set at the top, with input from central R&D and from those R&D departments located within the SBUs. 

R&D Strategy Groups

For R&D to play a role in formulating corporate strategy, clear structural and process links between R&D and Corporate Headquarters must exist. One common and effective means of establishing these links is through an R&D Strategy Group. 

 In one company represented in this group's participants, the R&D Strategy Group's tasks are to allocate projects within the corporate research budget, to decide on and develop competencies among the labs, and to determine the size of those labs. The people who serve on this Group come from each of the company's labs and serve as members of the Strategy Group for three years. These people already know the research organization before joining the Group, and their Group service exposes them to the process of formulating a research strategy. The head of the Strategy Group is a Senior Director of the company. 

The following characteristics make for effective R&D Strategy Groups: 

•  They allocate R&D projects and funding.
• The head of the Strategy Group is often a Board member or reports to a Board member, making him or her an important link between the R&D Group and Corporate Headquarters .
• They act as liaisons with outside funding and regulatory agencies.
• These groups are small.
• They act as administrative control for current R&D projects, maintaining a database.
• They may be asked to monitor a given technological area, (such as Òvirtual realityÓ), by reading, going to conferences, and doing technology assessments. They can then evaluate that technology's potential as a possible core competency for the company.

Success Factors

Given the above considerations, the following are keys to successful integration of R&D into the Corporate and SBU strategy formulation processes: 
•  R&D strategy influences corporate strategy and vice-versa; their intertwined nature suggests that the two should be established concurrently.
•  How involved R&D is in formulating corporate strategy seems to be affected by R&D's representation on the Board of Directors or Board of Management. R&D should be represented either directly on the Board, or indirectly by some high executive if R&D's interests and input are to influence corporate strategy.
•  The Board of Directors (and top management) should be involved in setting R&D strategy. 

II. B. Formulating Corporate and SBU Strategies

Once the appropriate organizational structures and links are in place to allow direct and significant communication between Corporate Headquarters, SBUs, and the R&D department, the question arises as to what activities R&D can perform to provide useful inputs for strategy formulation. Some possibilities, which are discussed below, are: 
  
• R&D audits 
• Identifying technological trends 
• Analyzing core competencies

R&D Audits

To formulate a corporate strategy, the company needs to know what its technological capabilities are and what they could be, given the technology that is available in the market or that could be developed in-house. Answering those questions involves a technology audit. Because technology audits involve the organization in critically evaluating its R&D capabilities, a good R&D audit requires an honest, self-aware evaluation. 

 Technology audits are particularly valuable for crucial technology make-or-buy decisions. These decisions strongly influence the company's corporate strategy, and mistaken make-or-buy decision can be very costly. 

 Should an R&D audit determine that buying a particular technology is the better strategy, the R&D group must evaluate potential sources of that technology. Often this involves identifying potential acquisition targets and then determining whether to buy entire organizations or merely the specific technology involved. 

 In conducting an R&D Audit, the following general questions should be kept in mind: 

 
• How often should an R&D audit occur? 
• What questions should be asked during an R&D audit? 
• What should be done as a result of an R&D audit?

Identifying Trends 

How can R&D help Corporate Headquarters identify social, industry, and technological trends that can be used to formulate corporate and R&D strategies? A commonly used method is: 
 
• Scenario planning 

Scenario planning involves asking targeted "what-if" questions to explore potentially critical circumstances that require planning. For example, one could ask the question ÒWhat technologies would be critical for us if the U. S. were to legislate 30% of cars to be electric by 2010?Ó Answers to that question could yield technology strategies that strategy planners might otherwise not consider. 

 A crucial part of scenario planning is how organizations generate the expansive and probing questions that force them to consider varied strategic options. Some organizations generate these questions with the help of psychologists, sociologists, as well as experts in scientific areas that may be important to the organization. Output from scenario planning analyses should be jointly examined by the Board, the Strategy Group, and the R&D department. 

Other means of identifying trends include: 

•  Delphi technique (an iterative process of questions and feedback)
•  Regulatory bodies, whose rulings may require product changes
•  Other industries whose technologies may be transferableThe rail industry, for example, often borrows technologies from other industries after the costs of those technologies have declined significantly. One instance of such lagged transfer is that railways now use vehicle suspension systems that have been used in cars for many years. Similarly, other advances in railways apply technologies developed in aerospace 25 years previously. This industry adopted technologies that were previously prohibitive in cost once the technology moved down the learning curve.

Core Competencies

A clear understanding of core competencies plays a critical role in both formulating and implementing strategies for the corporation, the divisions or SBUs, and R&D. Since strategies should reflect what the firm is currently good at, and what it wants to be good at, core competencies help define feasible and advantageous strategies for the company. Ideally, core competencies have high barriers for others to replicate and are sustainable. Depending on the industry and the firm, technology, marketing, and manufacturing can all be core competencies. 

Technical core competencies are fundamental because they define what is possible for the company. Identifying those competencies not only helps plan product development, it helps companies plan to acquire other companies or technologies. Given the importance of technical core competencies, what is the role of the R&D organization in this respect? 

•  R&D should identify the technical core competencies that make sense for the corporation. 

In doing so, R&D should consider the existing technical core competencies within the organization and look for synergies of those competencies across business units. Since technical competencies arise in business units, R&D can provide a valuable role in seeking ways to extend competencies to other existing or potential businesses. Another issue to consider are the core competencies of one's competitors. A competitor's skills may motivate the company to develop distinctly different core competencies, enabling it to compete from a different strategic platform. 
•  R&D can build or acquire the desired core competencies.

Core competencies require technical expertise, experience, and resources at the R&D level. Outright acquisition of technologies or companies that hold those technologies is one option for obtaining that expertise; licensing is another. Canon, for example, initially licensed the thermal copier technology in order to gain marketing competency in the copier business, while developing their own new process to circumvent the Xerox patent. Identifying needed future competencies leads a firm to take the actions required to develop or acquire them. 
•  The company can build on existing technical competencies to develop new products.

There are two ways to build on core competencies: 
•  Extend existing core competencies into new markets through product development.
•  Build new technical core competencies that logically follow from existing competencies. 

This strategy will eventually enable the company to develop new products. The new competencies must be assessed with care to avoid wasteful investment.
• Who should be responsible for determining extensions to competencies? Often the Marketing department of an organization is entrusted with determining the extensions to core competencies, although some companies assign that responsibility to R&D. One of the companies in the Forum gives individuals in R&D the additional role of Òmarketing expertÓ so that both marketing and technical knowledge might be brought to bear on these issues.

 Although all the above can help companies identify and use technical core competencies, a few other considerations regarding core competencies are important: 

•  Companies must also identify other competencies necessary for entering new markets.Komatsu, the Japanese heavy-equipment manufacturer, recognized in trying to overcome their competitor, Caterpillar, that product development competency alone was not sufficient for success. Developing manufacturing process competencies enabled them to build the high-quality products they needed to compete with Caterpillar. This example indicates the need for competencies to be complementary.
•  Core competencies may not be enough.

Even though organizations must rely on core competencies in formulating and implementing strategy, these competencies may not be sufficient for successful product development. One company in the group, for instance, had all the necessary technical competency for a particular class of products, but failed in its market entry attempts because its competitors had the financial power and market share to retaliate. Limited financing can hamper market entry, even if technical competencies are in place.

Conclusions: The Role of R&D in Strategy Formulation and Implementation

Given the complex information gathering and processing needed, and the difficult decisions involved in strategy formulation and implementation, there are some general activities by which R&D can support these processes: 
•  Corporate planning can raise technical questions, but R&D is responsible for finding the answers. To find those answers, R&D should:
•  Make clear the state of the art of technology (technology benchmarking)
•  Allocate resources to projects intended to strengthen core competencies
•  Identify needed future technical competencies
•  Establish linkages with other knowledge organizations (outside the firm)
•  Identify targets for research or technical alliances
•  Target potential acquisition candidates (for technology and other considerations)
•  In acquiring technology, decide whether to obtain an entire company or the desired technology only
•  Evaluate projects given input from strategic planning and marketing

III. R&D Structure: Centralization vs. Decentralization

The key question regarding the R&D department's organizational structure is ÒBy what criteria should a company decide whether its R&D structure should be centralized or decentralized, or take some intermediate form?Ó Centralization or decentralization of structure refers to where to locate and how to organize the functions of basic research, development, design competence, having in mind interfaces with manufacturing, marketing, and other functions. First, this section discusses the advantages and disadvantages of centralization and decentralization, and the factors that favor one arrangement over the other. Subsequently, the report examines the structure question from a technological perspective, relying on Jeffrey Williams' theory of cycle time in the product's environment. 

III. A. Centralized R&D

The main factors that determine whether centralized R&D is appropriate are technology, site characteristics, and marketing considerations. They are each discussed below: 
•  Technology Reasons 
• Centralized R&D is preferred when the technical expertise and resources needed to develop a product require a synergy of diverse inputs and a critical mass of workers, machines, or know-how. This is particularly important when developing and maintaining a company's core competence. Thus, centralized R&D is recommended when:
•  Considerable technical competence is needed to develop a new product or the next generation of an established product. Under these conditions, a critical mass of skilled researchers is often needed.
•  A given technology is common across multiple products or business units, and that technology is a core competence that helps the company sustain its competitive advantage.
•  The company can continuously improve the quality of its products more easily through a single development or design center.
•  High technology or fast cycle products (such as computer chips, CDs, or the Sony Walkman) can be globally standardized.
• Site Reasons

Centralized R&D reduces the development cost of products because duplicated efforts are avoided, coordination and communication costs drop, and economies of scale may be realized through standardization and high usage of expensive equipment. It is economically preferable to use centralized R&D when: 
•  There is a single best site for doing research for that industry. In the U.S. automobile industry, for example, Detroit is advantageous because of its proximity to customers and suppliers, while chip design might best be done in Silicon Valley because technical skills and information are available locally.
•  It makes little difference, because of the product's characteristics, where research is done. In other words, when there are no special commercial or research centers for the item, multiple sites would be wasteful.
•  The costs of communicating and transferring research knowledge to operations are not substantial.
•  There is a single best site for manufacturing; ideally development and manufacturing should be located together to improve coordination and communication. Alternatively, if there are multiple manufacturing sites, but little difference in processes from site to site, and low technology transfer costs, R&D can be co-located at a single manufacturing site.
• Marketing Reasons

Centralized R&D works best from a marketing perspective when: 
•  There is a single best location for development due to a need to be close to major customers (who may themselves be centralized).
•  The product is not highly differentiated across markets, so development adaptations are not needed in each market.
•  Marketing can provide detailed information on the company's worldwide markets and there is a global product standard.
In short, the above suggests that firms centralize R&D: 
• To take advantage of a key site for reasons of technology, development, marketing, or manufacturing. 
• If technology, marketing, and manufacturing issues are standardized across the company's markets.
Although a centralized R&D organization can be beneficial, it must be managed to encourage diversity and fresh thinking. By blending different cultures and ideas within the R&D organization, the company can develop synergies borne of multiple approaches to a problem. One way to encourage diversity in a centralized R&D group is to hire people specifically with varied and complementary backgrounds. 

III. B. Decentralized R&D

Decentralized R&D is advantageous under the following conditions of technology, site and marketing: 
• Technology Reasons

Decentralized R&D is often used for reasons of technology when: 
•  The company wants to develop and maintain alliances with several specific sources of expertise outside the company. Centralized R&D organizations have only one local set of contacts; having multiple R&D labs enables the organization to develop a larger number and a richer variety of alliances with other companies, suppliers, customers, universities and institutes through which knowledge can be developed and exchanged.
•  There are advantageous local resources available. These resources may include special intellectual talents, lower cost researchers, or different required materials that may be locally available. In some cases, there may be access to special resources, such as local flora or fauna.
•  Local standards must be complied with. Government regulations, such as environmental or Òlocal contentÓ laws, may require local manufacturing and R&D sites. 
• Projects need more research staff than are available in a single location.
There is usually more justification for multiple development sites than there is for multiple basic research locations, given the economies of scale needed for effective basic research and the need for development to be closer to the customer. 
•  Site Reasons

Decentralized R&D is often used for manufacturing reasons when a company wants to take advantage of unique or advantageous manufacturing conditions in different locations. Some manufacturing facilities may be better suited for local production; co-location of R&D with those factories is often advantageous. 
•  Marketing Reasons

Decentralized R&D is often used for marketing reasons when: 
•  A need exists to be close to customers in different countries. Decentralizing R&D enables firms to tailor products to local tastes and stay in touch with local trends. For example, the user interface on an electronic product may be crucial to the product's success, yet preferences for that interface may vary from market to market. Differences in cooking techniques (appliances), driving habits (automobiles), or aesthetics may dictate R&D sites in several countries.
•  The value added for the product comes mostly from adaptiveness to the local market. When most of the value added comes from design features, service or packaging, decentralized R&D is advantageous.
•  Rapid new product development is a key marketing need that provides a competitive advantage. Decentralized R&D helps the company stay on top of markets in which customer requirements change rapidly. When concurrent engineering can bring products to market faster, development (but not research) of products should be located in multiple marketing and manufacturing locations.

Cycle Time of the Environment

Slow Cycle Products

Standard Cycle Products

Fast Cycle Products

Given these definitions and characteristics of products with different cycle times, the participants considered the following R&D structures: 

Basic
Research                    Preferred R & D
Structure     Slow
cycle                    Centralized     Standard
cycle               Centralized     Fast
cycle                    Centralized          Product
R&DSlow cycle products          DecentralizedStandard cycle
products          Mix of centralized and decentralizedFast
cycle products               CentralizedProcess R&DSlow
cycle processes          CentralizedStandard cycle
processes          CentralizedFast cycle
processes               Centralized

Success Factors/Conclusions 
To evaluate the centralization-decentralization decision, the following issues are relevant: 

 Marketing Factors 

• Need to be close to the customer
•  Differences in markets and product formulations across countries, versus opportunity for standardization 
• Speed of change in local markets

Site and Coordination Characteristics

• Cost of multiple sites 
• Communication and coordination ease and costs 
• Duplication of resources and effort 
• Costs of transferring findings 
• Need to link with manufacturing 
• Special characteristics at different manufacturing locations

Technological and Expertise Factors

• Need for specialized human resource skills 
• Specialized centers of product expertise 
• Uniformity of the company's products from a technological perspective 
• Special characteristics of the product 
• Need for critical mass of human and technological resources

General Conclusions:

• Centralization is usually less expensive, so there must be marketing, site, technological, or political reasons not to centralize. More specifically,
•  If there are special characteristics of sites for marketing, manufacturing, or technology, R&D should locate accordingly. If there is only one such site, R&D should centralize; if there are multiple special sites, R&D should decentralize.
• Technologies for globally standardized products are best developed through centralized R&D; regional or national product differentiation often calls for decentralized R&D.

III. C. Duplication of Effort

When companies establish several foreign research labs, duplication seems inevitable. This section looks at prime reasons for duplication, the disadvantages and advantages of duplication, and some remedies for its costs. 

Some reasons for duplication are: 

•  Inadequate Monitoring and Control

The simplest form of duplication occurs because of inadequate monitoring and control of lab activities. In such situations, labs do not know what research activities the firm's other labs are engaged in. Such duplication and the associated waste can be reduced by: 
•  Better oversight of R&D activities by the Strategic Planning Group for R&D.
•  Better dissemination to the labs of reports and other materials that document research activities and findings within the organization.
•  Better communication in general among labs. While convenient for disseminating information, communications technologies such as e-mail may not afford the needed security for highly secret projects. Similar security problems arise when information about confidential projects is made accessible in the corporate research files. Security must be balanced against the benefits of wide communication.
•  "Horse Races"

Another form of duplication might be called the "horse race." This is a situation where several labs compete with one another to generate the best solution to a given problem. Such "horse races" might be purposely established, or they may simply emerge because one group is unaware of a similar project in another lab. 
Although costly, such duplication is not necessarily bad, particularly if it induces productive competition among labs and generates multiple ideas. From these competing ideas, the best may be filtered and selected. 

 Nevertheless, "horse race" competition also has its drawbacks. For example, 

•  Inevitably, if several teams are all working on the same technology and only one team's solution is chosen, the company must manage the disappointments and frustrations of the losing teams, or risk low morale or turnover of these individuals. Some suggestions for managing the frustrations of the losing teams are:
•  Try to depersonalize the choice of technologies, focusing on the limitations of the technology itself as the reason for the preferred solution, not those of the individuals who worked on the research.
•  Reshuffle individuals on the losing team onto new teams so repeated "failure" becomes less likely and the team does not develop a "loser's mentality."
•  Portray the results of the "losing" team not as "failure," but as results which are as helpful in assessing technologies and making decisions as positive results are.
•  Set the rules of the competition well ahead of time so that the groups see the competition as fair. 
• Have one group develop several alternatives, without inviting inter-group competition. 
• Desire to work with the latest technology

Duplication also arises because each lab wants to work on projects involving the latest technology. As a result, battles emerge about resources because the allocation determines what can and cannot be researched in a particular lab. Nevertheless, not every lab can work on the same topic because certain crucial and expensive machines are often needed for major and advanced research projects. If technology and equipment are not issues, different labs might look at different aspects of the same problem. 

One reason to avoid technology-driven duplication is that it carries hidden costs with it. For example, each time an expensive or complex piece of machinery is purchased, there may be a tendency to expand the lab personnel to some minimum number needed to use the machine adequately. Thus, the hidden costs of duplicating machinery can be more costly than the initial purchase cost of the machine. 

• Involuntary Duplication: Government-Mandated Testing

Some forms of duplication are not voluntary. For example, if government regulations insist that tests be done in the country in which the product is to be sold, the company must set up similar labs to test the product in several countries. Examples of where this applies include: 
•  High Definition Television (HDTV) In the U. S., the F.C.C. requires that work done on televisions to be sold in the U.S. be tested in the U.S., while the E.U. has similar requirements for European testing. 
• Telecommunications
•  Safety requirements for automobiles and consumer products
•  Food and household products
•  Environmental mandates 
These requirements for multiple testing exist even though the formulation of the product may not differ across countries. Although this may make little rational sense, the best companies can do to reduce these restrictions is lobbying of the relevant government. 

III. D. Standardization of Products

When firms decide to decentralize R&D by establishing labs in several countries, the purpose is often to get closer to the customer and regulatory agencies. Quite often a concomitant goal of having local labs is to develop products that are tailored to local preferences and specific government regulations. Nevertheless, the more a company tailors its products to local environments, the more it gives up the advantages of having a small number of standardized products. This section examines the advantages and disadvantages of product standardization and tailoring. 

 Disadvantages of Excessive Product Tailoring to Local Markets Although tailoring products to local markets can help make those products more attractive to potential customers, excessive technical adaptation of products to local markets exposes the company to problems of: 

•  Differential quality across markets

If no company-wide standards exist for product quality, unanticipated and undesirable variations in product quality across labs and markets are possible. This can create variations in performance and problems for service and replacement parts. 
•  Differences in brand image

The company cannot maintain a global or regional brand image if every country's product is different. This limits the company's marketing strategy for the product, while increasing the costs and complexity of marketing. One solution is to maintain as much inward technical similarity as possible, while providing functional differences in relevant markets. 
• Safety testing

Safety testing to meet local regulations becomes costly because of the larger number of safety tests the company must conduct worldwide. 
•  Development problems

Development problems arise because the wider variety of products increases the chances that learning gained in another lab may not apply to a new and highly specific product variation. Also, development costs rise as such activities can no longer be centralized to take advantage of specialized equipment, skills, and economies of scale. 
 
Of course, different formulations are often needed to meet local government standards or consumer preferences. The corporate R&D office in one large company represented in the Forum has a policy on this issue: it requires its overseas subsidiaries to justify in advance any proposed changes from the standard formulation. In these situations, the burden of proof is clearly on the local groups to justify the change. More specifically, to make a special model, the local groups in that company need to justify their strategy with: 
• A large enough projected sales base to justify a new model
•  Consumer tests of the necessity of the change
•  Regulatory and legal drivers of the change
Selective Standardization 
Even though customer requirements and tastes can differ from country to country, market differences are not important for certain products or parts of products. For example, the styling of a product may indeed need to differ in different markets, but the underlying structure of the product (such as the motor or chassis of a car) can often be standardized. Once these universal products or parts are identified and standardized across products, their R&D can be centralized, saving costs and avoiding coordination problems. Indeed, if the costs are not prohibitive, the most stringent worldwide standard design can be chosen for the item. For example, in certain cases car manufacturers might be better off making all cars to U.S. (E.P.A.) emission standards rather than designing each foreign model to the different (but lower) emission standards of other countries. This assumes that the costs of using the highest standard worldwide are more than compensated by the savings in duplication of R&D and manufacturing costs. A careful cost calculation is needed here. 

IV. The Transfer of R&D Knowledge: Ideas and People

A major issue that firms face in managing international R&D is how to transfer the knowledge that is generated and stored in the company from one site to another, so that products can be developed world-wide and organizational learning can occur. Organizational transfer of knowledge occurs in two ways: 
• Transfer of ideas 
• Transfer of personnel

a) between R&D and SBUs or Divisions 
b) between R&D labs.
The issues associated with each are considered below. 

IV. A. Transfer of Ideas 
The transfer of ideas without the formal transfer of personnel is one means by which companies transfer R&D knowledge throughout the organization. Two types of transfer are involved, Vertical and Horizontal. 

1) Vertical Transfer: Basic-to-Applied  
Vertical transfer is the movement of research findings from basic research labs to applied research laboratories, to product development and production. Vertical transfer carries the efforts and expenditures of the R&D labs into new product development. 

2) Horizontal Transfer: Dissemination 
Horizontal transfer is the movement of research findings across the firm's various research and development sites around the world. Such dissemination is important to avoid duplication of effort and to develop research synergies across sites. 

IV. A.1. Vertical Transfer

Without vertical transfer, R&D, and basic research in particular, has little justification for existence. Vertical transfer occurs primarily by one of two models, Sponsored Research or R&D-Driven Research. With a Sponsored Research model, the SBUs or Divisions request that specific research projects be carried out for them by R&D. In the R&D-Driven Research model, R&D generates ideas that it must then sell to SBUs or Divisions. Each mechanism will be discussed and evaluated in more detail below. Inherent in the discussion and debate between these two fundamentally different models are the four following questions: 
• What are the advantages and disadvantages of these two models?
•  What organizational characteristics work better with one or the other model?
•  What structures of control and accountability are needed for each model to work well?
•  What is the role of each in facilitating transfer?
Sponsored Research Model 
Under a Sponsored Research model, research is done only when a sponsor or champion in an SBU or Division supports it; R&D does not start projects autonomously. The identity and location of the sponsor may vary from company to company, so that sponsors may be the SBU itself, or they may be outside the SBU. 

Most often an SBU or Division comes to the Research Group and tries to convince it that the SBU's proposed research has economic or strategic value. This procedure is used because corporate, not SBU, money usually pays for such research. Thus, accountability in this system rests with both the SBU or sponsor, which must convince Research of the project's value, and with Corporate Headquarters, which pays for the work. This model's strong emphasis on accountability may explain why resource-constrained companies prefer it. In some companies, the SBU may provide the majority of the funds, which tends to result in focus on solutions to short-term development needs. 

With this model, only ideas migrate from the SBU to R&D; the people stay put. In one company in the Forum, however, the SBU lends people to R&D and then later regains these people. In this situation, the transferred individuals are more likely to take a project back with them from R&D because they have become "owners" of that project. They have most likely formed a group that will sustain itself outside R&D, allowing their idea to take root in the organization. If their project is successful, the individuals involved are more likely to be promoted, further increasing the chances that the idea will be used in the organization. This is an effective means of transferring ideas through people, but the practice is not a universal solution because its continual use can weaken the core research group. 

Advantages of Sponsored Research 

• Less waste of resources than with non-sponsored research.
•  More accountability for resources used in R&D.
•  SBU sponsorship facilitates interest and makes it easier to disseminate research.
•  A more appropriate model for development than for basic research, as specific products need to be generated.
Disadvantages of Sponsored Research 
• Companies that use this model are less likely to generate unanticipated discoveries.
•  Dilemmas arise if a sponsoring division loses interest in a project or runs out of money and cannot convince Corporate Headquarters to fund the project any longer. One option is to have Corporate Headquarters continue funding the project if it seems sufficiently interesting for the company overall. That is, sometimes R & D sustains an abandoned project because it and Corporate Headquarters believe the project is applicable in ways that the sponsor either didn't see, need, or have the funding for. Indeed, some R&D managers believe that the R & D organization can be ÒsmarterÓ than the business units involved because R&D's expertise with the technology may give it a better perspective on the research's ultimate utility. Thus, R&D may continue projects for which SBUs do not see the value. Nevertheless, this is always a difficult decision and finishing the project with corporate money eventually has to be defended before upper management. A project's contribution to core competencies can be a reason for continuation.
R&D-Driven Research Model 
In the other model of vertical transfer, R&D-Driven Research, the key is that ideas and research are generated independently by R&D without specific input or sponsorship of the SBUs or Divisions. Once these ideas develop or research results are available, these ideas must then be sold to SBUs. In some firms, top management must endorse ideas before work can begin on them, but R&D clearly generates these ideas. 

 Under this model, Corporate Research takes responsibility for disseminating potentially useful results to the SBUs. In one company this is done by talks, papers, reports, videos, and workshops. The regular participants at these workshops are researchers from both the company's other research labs and from development groups in the SBUs. In addition, this company schedules yearly exhibitions and research conferences where the labs display yet-to-be-applied findings that could potentially be used throughout the organization. 

Advantages of R&D-Driven Research: 

• This model allows a company to develop technologies and competencies that might not be immediately needed, but which could have strategic importance in the future. The skilled selection of those projects whose potential results could be applied in many places in the organization can make this a productive, viable strategy. In some companies such projects focus on key technologies identified as crucial for the future.
•  For the above reasons, this model may more useful for basic research than for development.
Disadvantages of R&D-Driven Research: 
• R&D has to sell an idea to the SBUs after it is developed; R&D may not be close enough to the needs of the customer at an early enough stage.
•  This model may result in considerable waste because none of the SBUs uses the results of some research projects. Of course, there is always some ÒwasteÒ of research efforts, particularly of basic research. Thus, organizational attitudes towards risk (relative to available funds) may determine the acceptability of this model.
Accountability for Research 
Sponsorship is one decision that must be made regarding R&D; another is who pays for the research. In some organizations, the sponsor and paying customer are the same unit, while in others these roles are divided. 

In many companies, Corporate research is free. While this encourages the SBUs to take advantage of R&D resources, not surprisingly, there are always long lines of SBUs waiting for Corporate R&D to do their research. Given this backlog of requested work, how does Corporate R&D choose its projects? The company's overall Corporate strategy should be the main guideline. In addition, projects whose results are more likely to benefit several SBUs are more likely to be chosen. 

In most organizations, results generated by Corporate research for one research project are freely available to other business units. For example, in one firm, the International Group regularly adapts and improves on products developed for the U.S. market, which are often perceived as the standard formulations. 

 Success Factors for Vertical Transfer of Ideas: 

•  The appropriateness of using sponsored or non-sponsored models may depend on whether the work is "basic research" or "applied development" work. Basic research may be more successfully funded using an R&D-Driven model in which Corporate research is the sponsor. Applied development work, for which specific goals exist, may be better sponsored by the SBU that is responsible for the project.
•  The appropriate sponsorship model also depends on the risk preferences of the organization, the methods of accountability and control, and the nature of the industry and projects.
•  Other factors that may determine how much research should be sponsored and paid for by Corporate Headquarters include the size of the company and type of technology.
•  In some companies, a combination of approaches is used with the Corporate share ranging from 10 percent to 50 percent of the total.
•  Good communication and discussion of research priorities between central R & D, dispersed R & D centers and SBUs is critically important.

IV. A.2 . Horizontal Transfer

Horizontal transfer is the productive transfer of research ideas and results among an organization's research and development labs around the world. Firms can facilitate such transfer either structurally, through the configuration of their labs, or procedurally, by transferring personnel. These methods are considered in the next two sections. 

Organizational Configurations 
Although numerous organizational structures are used by different companies, some are more common than others. In particular, "Network" and "Star" configurations occur frequently. Their advantages and disadvantages are presented below. Organizational structure for R & D is discussed in the 1993 Report of the Institute's R & D Forum, "Managing International Research and Development", November 1993. 

"Network" ConfigurationVarious companies organize their labs as networks in which little central control or direction of the labs occurs. In many cases, each lab may concentrate on a particular product or set of products. This arrangement has several advantages: 

•  Labs maintain their autonomy and can carry out research programs that best meet their needs and the needs of the respective business units.
•  This arrangement works well when there is a natural distribution of projects for each lab because of local market differences or big technical differences between SBUs justifying technical centers of excellence. When this occurs, competition for projects is not a problem.
The disadvantages of a network, however, may be: 
•  The lack of a central or "leader" lab raises issues concerning the coordination of projects. One company that uses this configuration has had trouble coordinating labs on the East and West Coasts of the U.S. Their experience is that such "leaderless" split assignments are not a good idea, at least during the initial stages of a project. This problem is more acute when both labs are working for the same or similar businesses.
•  Disparities in the labs' sizes may develop with no means of equalizing them. In one company, one of its labs is disproportionately large, because of history and the nature of the projects being undertaken there. Under those circumstances, the smaller sites worry about the power of the larger site, and yet have no easy recourse for attaining equality of size, power, and resources. To remedy this situation, the company gives each group global assignments, but with a certain amount of local development for particular sites. Still, smaller labs worry about not having sufficient global responsibility.
•  There is a greater chance that duplication of efforts will occur across uncoordinated labs.
"Star" Configuration 
An alternative configuration for organizing research labs is a "star" model, in which a central lab is surrounded by satellites whose role is mainly to localize products developed elsewhere. One large company prefers this arrangement to "leaderless networks" because it more easily exploits economies of scale and yields better coordination. This arrangement also avoids the leadership problems of the general network configuration, but sacrifices some of its flexibility and autonomy. This company finds it workable because international transfer for them means adapting findings from the central lab to local conditions. In this case, the company is not highly diversified. Reverse communications and transfer from the edges of the star to the center are less common in their case, and this may be a weakness. 

Critical Factors for Horizontal Transfer 

•  The structure of the labs' international configuration and the distribution of leadership across these labs play key roles in the process of horizontal transfer.
•  Power must be equalized across labs to maintain political satisfaction with the importance of the work each unit does.
IV. B. Transfer of Personnel 
The transfer of personnel can be an effective way for companies to enhance the breadth and coordination capabilities of their R&D organizations, and the skills and knowledge of their researchers. Also, because ideas reside in peoples' heads, transferring people between labs or between labs and divisions is another way to transfer ideas throughout the organization. Several types of transfers, each of which is discussed below, might be used: 

  

• Among R&D labs 
• From the R&D labs to the Divisions or SBUs 
• From the SBUs or Divisions to the R&D labs
Transfer Among R&D LabsThe international transfer of employees from one research or development lab to another is a common means of broadening a researcher's experience and exposure to the company. These transfers are often for periods of several years and culminate with the return of the individual to his or her home lab. 

 Advantages: 

•  International transfers allow personnel to gain technical expertise unavailable in the home country.
•  International transfers enable particular technical skills and experience to be transferred to another site.
•  International transfers allow personnel to build personal connections in other parts of the organization, facilitating communication and knowledge.
•  Foreign assignments may allow the individual to play a more important role in the foreign subsidiary than he or she would have at home. This enables the person to build valuable managerial and technical skills.
•  As a result of the above experience, transferees are often promoted upon returning to their home countries, enabling individuals to advance their careers.
Disadvantages 
Despite the above advantages, international transferees may be disadvantaged in several ways: 
•  They can lose touch with the home R & D center and mainstream activities there. This can be mitigated by frequent reports, home office visits, and a designated contact at the home office who keeps the transferee up-to-date on home activities and changes.
•  These transferees may be guaranteed a position upon their return, but may not be able to return to this position because opportunities have changed in their absence.
•  Individuals who return to less than satisfactory conditions in the home office may be more likely to leave the organization, negating the benefits of the transfer in the first place. Incentives such as promotions, salary increases, and other motivational and recognition measures may be needed to complete a successful transfer and retain the transferee. 
Critical Success Factors: 
• Make sure transferees stay in touch with the home office during their time overseas. A specific person should be designated to take responsibility for keeping the transferees in touch with mainstream developments in the home office.
•  Do not design or portray overseas positions as minor positions, as they often have been perceived. By developing substantial overseas positions and valuing such experiences, companies can induce more talented people to request or accept such transfers, and at the same broaden the firm's experience base.
•  Link the duration of transfers to specific objectives and the nature of the position involved. Transferees require a certain amount of time to adapt to new environments and become productive; two years seems the minimum time needed for a successful transfer assignment.
Transfer from the Divisions or SBUs to R&D 
Some organizations value having people transfer from particular divisions or functional areas to the R&D labs and back again. One organization rotates technically-qualified people from marketing to R&D and then returns those people to marketing positions. Selecting individuals with the necessary technical skills is of course a critical prerequisite for such a transfer. 

 In another organization, people who have sufficient experience in various technical or business groups work on R&D projects for a set period, although they formally remain within their business units. After two or three years of such experience, these individuals can formally become members of the R&D department. 

 One advantage of lending people to R&D and then returning them to their original units is that they are more likely to take new ideas they have worked on or encountered back to their units when they return. Given that they may receive promotions when they return, the transfer experience increases the chance that ideas developed in R&D will take root in the company. 

Transfer from R&D to the Divisions or SBUs The other direction in which transfer can occur is from R&D to the divisions or SBUs. This can yield some of the same technical and organizational cross-fertilization that occurs when transfers go from the divisions to R&D. One organization informally requires that employees spend time in corporate R&D before entering the production function or other divisional R&D centers. This company does so partially because there are no substantial and directly applicable university programs where people can obtain the knowledge needed to perform research and development on the company's products. Thus, their transfer program fulfills an important training role. 

 One firm has a program whose goal is to send 8 percent of the R&D personnel through their business units every year. Another firm's goal is to hire new research graduates into central R & D and then transfer them after five to ten years to production units of business groups. Their goal is to transfer 15 percent of their R&D employees from research labs to business units each year. 

 While certainly beneficial, the disadvantage of this process is that people who come out of R&D sometimes do not adjust to the practical realities of the production environment. They may prefer to experiment and look for optimal solutions, rather than get things done according to strict deadlines. 

 In addition, downsizing can hinder transfer programs because it limits the number of positions in the divisions or functional areas to which R&D personnel might be sent. Some organizations facilitate transfer by moving personnel within a specific project, sometimes from the pre-production environment to development. Cross-project movements, however, are more susceptible to the pressures of downsizing for both economic and political reasons. 

 Finally, while moving personnel from R&D to SBUs or Divisions is a good way to transfer ideas, this cannot be used excessively as it would weaken the core R&D group. 

Success Factors: 

• Rotations either from R&D to the Divisions or in the other direction can be valuable, but only if the people involved have the necessary technical competence and interest.
•  Institutionalized rotation programs may reduce resistance to these practices and make them more successful.

V. Operational Issues

Having examined major structural aspects and processes involved in managing international R&D, one further set of issues involves operations. This section highlights both the means by which companies evaluate the effectiveness of their R&D operations, and the processes that can help companies develop products more quickly. 

V. A. Evaluation of R&D Effectiveness

Evaluating R&D effectiveness is a subjective, difficult process. Evaluating basic research is more difficult then evaluating development efforts because the results and long-term utility of basic research are often unclear. Although managers have developed measures by which to measure basic research, their accuracy is often exaggerated. 

 Given these characteristics of the R&D process, (particularly basic research), quite often top management does not evaluate research groups globally. Rather, periodic evaluations are done on a project-by-project basis by the SBUs they serve, and by upper management. Based on these evaluations of whether customer and corporate interests are being served, corporate executives determine whether an R&D project is productive and should be continued. 

 Some criteria by which development efforts are often evaluated include: 

•  Transferability and applicability of the end result 
• Quality of the research in scientific terms
•  Time required to complete the project
•  Products developed per 100 R&D personnel
•  Cycle time needed to bring product to market. This measure is imprecise because many factors outside the control of R&D affect cycle time.
Although conventional wisdom suggests that successful R&D groups are those that receive good reviews from the SBUs they serve, alternative opinions exist. That is, some feel that R&D should not get overly high or low reviews from the SBUs. Overly positive evaluations would mean that R&D is too focused on short-term development projects and is not doing enough basic research to sustain the company in the long term. Overly negative evaluations would also be dangerous because this means that R&D is not serving the SBUs properly and may be impeding product development. Thus, R&D evaluations should be carefully designed to reflect in detail the firm's goals for short- and long-term growth. 

V.B. Speeding the Product Development Process

An axiomatic fact of life in most companies is that they want to speed the product development process. How important shorter product development cycle times are to a company depends on the nature of competition and cycle time in its industry. While reducing product development time is an important goal, this task is generally easier said than done. 

Developing a product consists of three phases: deciding what is demanded, evaluating feasible solutions, and then actually designing and testing the product. The key to reducing cycle time is often not compressing the product development time itself, but reducing the time spent in the first two phases--determining need, and evaluating product feasibility. Some solutions that target these earlier phases are: 

•  Reduce product variety without sacrificing market presence. Marketing departments often want additional versions or variations of a product to satisfy the demands of different market segments. However, companies can be overly responsive to consumer needs and demands, generating and supporting too many products. This over-responsiveness creates a situation whereby the product development process becomes swamped, developing too many new products, thus lengthening the cycle time for all new products. In fact, overloading the product development system slows it down disproportionately (that is, the inefficiency generated is larger than the additive load each one places on the system). One powerful remedy to this tendency towards over-variety is to resist proliferation of products. This has several benefits: 
• The product development process can progress more quickly when it is not overwhelmed. 
• Supporting fewer products is less expensive, faster, and requires lower inventory, documentation, and coordination costs throughout the entire organization. Many benefits of supporting fewer products are not immediately visible, but are real and substantial.
To reduce "product swamping" several tactics are offered: 
•  Ideas should be evaluated and filtered to prevent the system from bogging down. 
• The design department may need to assert more political influence to direct and manage its own process, rather than respond to all the wishes of marketing.
•  When the company sets its product strategies, the product development process should be viewed not only in economic terms, but also as a political process.
In addition to reducing product variety, some other solutions for reducing new product development cycle time are: 
• Radically change the design process through process redesign methods.
•  Use the best people in the development cycle.
•  Use modular building blocks and reusable parts. For example, one company uses software platforms that can be reconfigured for use in other environments.
•  Control each project with its own time schedule.
•  Have gatekeepers to make sure that team members adhere to schedules.
•  Use a permanent team whose role is continually to examine and analyze the development process activities and find ways to eliminate or shorten process steps.
•  Develop several generations of a product in parallel.

Appendix A: Structures of R&D Departments

Participants of the Forum provided outlines of the characteristics of R&D and its organization in their companies. Based on this information, what are the common dimensions of R&D structures? Some of them are: 
• Location of R&D departments as independent units or within SBUs or Divisions.
•  Reporting relationships.
•  Location of Development labs as distinct from Research labs.
•  Percent of total research done by:
 
• Corporate research 
• Divisions or SBUs 
• Contract research 
• Which group funds the research within the company
•  Representation of R&D on the Board of Managers
•  Existence of laboratories that specialize in a given technology
•  Relationship of R&D to Marketing, Production, and other departments
•  Means by which the R&D departments are evaluated

Appendix A: R&D Structures of the Participants' CompaniesPhilips

Basic Structure 
• · Independent Product Divisions with individual development departments
• · Corporate Research reports directly to Board of Management.
• · Corporate research is organized according to technological disciplines. It is comprised of five laboratories, each in a different country.
Split of Research Load 
• Corporate research 30% 
• Contract research 70% 
Ý  

Appendix A: R&D Structures of the Participants' CompaniesDaimler-Benz

Basic Structure 
 
• Four main "research directions"/disciplines 
• 4 SBUs
Split of Research Load 
 
• Corporate research 50% 
• Contract/Divisional research 50% 
• SBUs do not use outside research centers to do the research, or resell research done at Daimler-Benz outside the organization. Nevertheless, Corporate research may contract with outside research facilities such as universities or other non-profit organizations.
Ý 

Appendix A: R&D Structures of the Participants' Companies

BMW 
• Applied research
Basic Structure: 
• Uses a centralized R&D center in Munich; 5,000 individuals in the same building
•  Production, marketing, and other functions are completely independent of this research center.
•  R&D Center is organized as a matrix with technology groups in one dimension and three products on the other. For example, technology groups specialize in motors, suspension, or other particular technological areas. 
• Projects are managed as profit centers
•  Project leader directs the product development effort; requisitions the technological resources from the needed technology areas.
•  Given this structure project leaders compete for the best resources in each technology organization. 
• Rover organization is separate from the overall structure.
•  Issue: How to manage two R & D centers for two different product lines (Rover and Other Automobiles).
 

Appendix A: R&D Structures of the Participants' Companies
Union Switch and Signal Inc.

Background 
• Company designs and manufactures rail/transit signaling and control systems.
Basic Structure 
• Three operating units: France, Italy, USA 
• Each division has its own R&D department 
• Advanced Technology Group (ATG) reports to US & S Corporate directly
•  ATG has the following charter: Pursue strategic core technologies within a three-to-five year planning horizon that can be transferred and adopted by the Divisions to support their practical, product-oriented missions.
Split of Research Load: 
• Corporate research 25% 
• Contract/Grants 75%
 
Williams, J. R., California Management Review, Volume 34, Number 3