Corporate strategy and the management of innovation and technology

Industrial and Corporate Change, Volume 11, Number 2, pp. 263–288

Corporate strategy and the management of innovation and technology Jens Frøslev Christensen

This paper explores the evolution of corporate organization with special attention to the organization of R&D. More specifically, the paper addresses the comparative long-term organizational dynamics of management of innovation and technology in two different types of technology-based industrial companies: the ‘related diversifier’ pursuing ‘synergistic economies’ and the ‘vertical integrator’ pursuing ‘vertical economies’. These types of companies are illustrated by case studies of two large Danish manufacturing companies. The analysis aligns the strategic management literature on strategy and structure in large companies with the literature on management of innovation and technology. It is argued that the organizational design for managing innovation and technology is contingent on both the overall strategy–structure profile and dynamics of the companies, and on key characteristics of their particular innovation and technology strategies.

1. Introduction This paper seeks to improve our understanding of the interdependencies between the overall strategy–structure dynamics of the large corporation and the particular dynamics associated with the management of innovation and technology. This is a neglected issue in both the corporate strategy literature and the innovation and technology literature. In the former, the organization of R&D is generally not treated in any detail, and in the latter, the role of corporate dynamics tends to be altogether excluded from the agenda. Throughout the paper, I make the following distinction between the analytical concepts of management of innovation and management of technology: management of innovation refers to the strategic and organizational context for the individual innovation processes with the objective to produce successful product and process innovations while economizing on time and resources. Management of technology includes the strategy and organization of the corporate or divisional technology base (the portfolio of existing and prospective technological capabilities underlying product and process innovation) with the objective of guiding technology diversification and accommodating technology integration. Management of innovation and technology is not only confined to what takes place in the firm’s R&D department. Thus, management of innovation typically involves elements of other business processes such as © ICC Association 2002

264

Jens Frøslev Christensen

marketing and manufacturing, and management of technology may involve managing process technologies which typically lies outside the domain of the R&D units. Therefore, I use the abbreviation I&T (innovation and technology) rather than the more institutionalized, but also more narrow term, R&D.1 The paper presents a comparative analysis of two large Danish manufacturing companies, Danfoss illustrating a ‘related diversifier’ company, and Grundfos illustrating a ‘vertical integrator’ company. The two companies demonstrate very different strategic and organizational trajectories both in their general corporate development and in their styles of managing I&T. The analysis is comparative in two other respects. First, it compares the role of the overall corporate strategy–structure versus the role of the particular innovation and technology strategies in determining the specific locus and structure of management of innovation and technology. Secondly, it compares the rationales behind similarities and differences in the development of management of I versus management of T. Sections 2, 3 and 4 provide the analytical framework of the paper. Section 2 gives a comparative review of the relevant literature on corporate strategy and structure with special emphasis on differences between related diversifiers and vertical integrators. Implications of the overall strategy–structure profiles for the organization of I&T are drawn. Section 3 takes the strategy–structure framework from the general corporate level into the domains of management of I&T with the objective to specify types of interdependencies between the strategy and structure of I&T. Section 4 proposes a contingency framework in which both the general corporate dynamics (as discussed in Section 2) and the specific dynamics of managing I&T are considered drivers determining the specific modes of managing I&T in technology-based related diversifiers and vertical integrators. Section 5 presents the two cases, and section 6 relates the case studies to the analytical framework. Section 7 concludes the paper.

2. Trade-offs in the development of corporate strategy and structure For a long time, it was widely assumed that the issue of economic organization of the growing and diversifying industrial corporation was settled by the innovation of the multidivisional (M-form) organization to replace the unitary/functional (U-form) organization. Chandler’s (1962, 1977) empirical and Williamson’s (1975) theoretical rationalization of the M-form as the appropriate governance structure for dealing with increasing corporate diversity provided key insights in how to organize a multibusiness corporation. Later research (Goold and Campbell, 1987; Hill and Hoskisson, 1987; Hill, 1988; Markides and Williamson, 1996; Riahi-Belkaoui, 1997) has pointed to the existence of a number of variants of the M-form corporation and indicated the relationship between these and different generic strategies. Thus, the decentralized or 1A more comprehensive critique of

the R&D concept can be found in Christensen (1995).

Corporate strategy and the management of innovation and technology

265

financial control oriented M-form (Goold and Campbell, 1987) accommodates unrelated diversification and seeks to pursue financial economies (Hill and Hoskisson, 1987). This form prevails in very diversified corporations in relatively low-technology sectors (Goold and Campbell, 1987; Chandler, 1991). According to Hill and Hoskisson (1987), the two other generic forms of diversification, related diversification and vertical integration, require more centralized coordination than does unrelated diversification. The related diversifier seeks to obtain ‘synergistic economies’,2 while the vertical integrator pursues ‘vertical economies’.3 Goold and Campbell (1987) propose a distinction between a strategic planning style M-form in which Headquarters is strongly involved in guiding the strategic development of the divisions, and a strategic control style in which Headquarters focuses on controlling the implementation of divisional strategies. These more centralized versions of the M-form tend to prevail in less diversified companies within more technology-intensive sectors (Chandler, 1991). This paper provides a comparative study of the two categories of corporate strategies that are most pertinent to technology-based companies, namely related diversification and vertical integration.4 As shown in Table 1, related diversifiers develop a broader range of businesses than vertical integrators, while the latter develop a broader range of vertically linked value chain activities related to their dominant business areas. Related diversifiers tend to become M-form companies with more or less centralized functions to promote coordination across divisions.5 Vertical integrators tend to rely on substantial top-level operational control to obtain vertical economies, and they tend to maintain a U-form structure. However, some vertical integrators establish a kind of M-form structure in which the different value chain activities are organized in divisions that can be allowed more or less autonomy to pursue business opportunities on the open market. Thereby, they tend to move closer to the related diversifier position and weaken the vertical economies. Empirical studies have indicated a decrease in financial performance and productive efficiency in vertically integrated firms following the implementation of the M-form (Hoskisson, 1987; Riahi-Belkaoui, 1997). Several statistical studies have found that diversified M-form firms have lower R&D intensities than less diversified, unitary firms and this may be due to managerial myopia 2Arising from exploiting economies of scope, i.e. economies from using inputs jointly by related activities (Teece, 1982), exploiting economies from eliminating transaction costs of generating synergy in the market place (Williamson, 1975), or economies from sharing know-how and capabilities (Porter, 1985). 3Arising

from scale or integration economies, from increase in control over the firm’s supply/output markets (Pfeffer and Salancik, 1978), or from elimination of the transaction costs of using market exchange (Williamson, 1975). 4In practice, most industrial corporations integrate elements of more than one of these categories. Especially Hill and Hoskisson (1987) have studied the overall corporate structures associated with these different growth models. 5Henceforth, the notion of division will refer interchangeably to product divisions and strategic business units.

266


Table 1 Suggested structural differences between related diversifiers and vertical integrators

Related diversifier

Vertical integrator

Range of

Range of value

Corporate

Predominant

businesses

chain activities

structure

economies

Broad

Narrow

Varied across

M-form (some central

Synergistic

businesses

coordination

economies/scope economies

Extended

U-form (centralized)

Vertical economies/scale economies

and inefficiencies in diversified companies’ internal capital markets (Hoskisson and Hitt, 1988; Klein, 2001). This paper does not penetrate this particular issue, but only the related phenomenon of understanding the differences in modes of managing I&T between related diversifiers and vertical integrators. The argument will be developed through two analytical steps. The first, which will be discussed below, provides the link between the general strategy–structure profile and the organization of I&T. The second step, the agenda of Section 3, establishes the link between I&T strategy and the organization of I&T. The first step assumes that the context for managing I&T is fully determined by corporate management in order to assure coherence between the I&T strategy–structure profile and the corporate strategy–structure profile. Table 2 shows the most likely structural alignment between the two strategy–structure categories and the localization of management of I&T. Related diversifiers have been differentiated into two analytical sub-categories: technology-related diversifiers that diversify into product-markets in which they can exploit existing technological strengths, and market-related diversifiers that diversify into product-markets in which they can exploit existing market positions as based on i.e. distribution, sales and marketing assets. The related diversifier that undergoes the transformation from U-form to M-form will tend to move in the direction of still more decentralized (or rather divisionalized) management of innovation, in line with the management of other business areas. Especially among exclusively market-related diversifiers, the divisions will tend to maintain a high level of autonomy in management of innovation, while they may share distribution and marketing assets with other parts of the company. Management of technology may not exist as a distinct policy area, and if it does, it is most likely to be a divisional matter and more or less integrated in the management of innovation. Since the rationale for technology-related diversification is to reuse existing technological assets in new markets or to gain access to complementary technological capabilities, management of innovation in technology-related diversifiers is more likely


267

Table 2 Overall strategy-structure profile and the locus of innovation and technology management

M-form related diversifier

U-form vertical integrator

{

Locus of innovation

Locus of technology

management

management

Market-related

Divisional

Either absent or primarily divisional

Technology-related

Primarily divisional/ coordinated

Centralized/divisional

Centralized with one business function

Centralized

to involve coordination between divisions or between divisional and corporate level R&D. A well-known example of a technology-related diversifier is 3M, which has three types of R&D laboratories: divisional laboratories are responsible for most product innovation, but they often work in close cooperation with sectoral laboratories responsible for the technical development at the industry level, and with central laboratories responsible for longer term research (Christensen, 2000). A central level of managing technology may be required to assure the technology transfer or sharing involved in technology-related diversification (see for instance Coombs and Richards, 1993; Argyres, 1995; Tidd et al., 1997; Boutellier et al., 2000). In 3M ‘. . . this commitment to share is backed by a strong . . . policy that technology belongs to the company not the divisions’ (Goold et al., 1994, p. 203). The vertical integrator which remains a unitary organization will tend to adopt a more or less centralized mode of managing I&T in order to achieve the coordination necessary for obtaining vertical economies. This may be self-evident, given that these organizations do not comprise a divisional level. Still, these management functions may be located differently within the corporation. They may be physically and/or organizationally located close to, or within, one or more of the operational business functions (i.e. production or marketing) or they may be closely associated with headquarters. However, in contrast to related diversifiers, vertical integrators do not have a ‘natural’ inclination to move towards more decentralized managerial regimes. So far, we have considered management of I&T from the perspective of the corporate strategy–structure according to which the organization of I&T has to adapt to the general corporate structure (which, again, should adapt to the corporate strategy). In Section 3, we move beyond the corporate relationship and focus on some key features of innovation and technology development.

268


3. Trade-offs in strategy and structure of managing I&T The mode of managing I&T should not only reflect the overall strategy–structure constellation and trajectory of the company, but also the specific objectives of the company’s innovation and technology strategy. Based on empirical insights on the features of innovation and technology strategy, this section develops three propositions concerning the interdependency between the nature of innovation and technology strategy, on the one hand, and the organizational context for their implementation, on the other hand. Empirical research on the dynamics of technical innovation at the firm-level provides us with two robust sets of insights: A. Technological innovation is by nature interactive, involving cross-functional and/or cross-disciplinary learning and R&D at different levels (Pavitt, 1998). The firmspecific manifestation of these features both constrains, and is constrained by, the firm’s organizational context for managing innovation. B. The range of technical capabilities tends to grow relative to the range of products or business areas in technology-based companies. The firm-specific manifestation of these dynamics both constrains, and is constrained by, the firm’s organizational mode of managing technology. Below, these insights will be discussed and three propositions suggested.

A: The interactive nature of technological innovation A large volume of empirical research on the factors affecting success in managing innovation has pointed to the critical significance of cross-functional learning (i.e. between R&D, marketing and production) and cross-disciplinary learning (i.e. between mechanical engineering, software and optoelectronics), both within and across the boundaries of the firm (on cross-functional learning, see among others Burns and Stalker, 1961; Rothwell et al., 1974; Cooper, 1988; Wang 1997; on cross-disciplinary learning, see among others Miyazaki, 1995; Nonaka and Takeuchi, 1995; Iansiti, 1997). In other words, successful innovation cannot, or can only rarely, be properly understood as the outcome of an autonomous or unitary R&D function (cf. Christensen, 1995). A useful distinction has been suggested between upstream-linked versus downstream-linked innovation (Gerybadze and Reger, 1999). In the former inter-disciplinary technology and science interfaces are the most critical (i.e. in chemicals and pharmaceuticals). In the latter inter-functional interfaces prevail (i.e. in civil engineering and medical instruments). Upstream-linked innovations bear on R&D labs with network or matrix structures and easy communication between groups of engineers and scientists within different technical disciplines. Downstream-linked innovations require organizational focus on the interaction between R&D and other business functions such as marketing and production, and users/customers. These categories of innovation are related to the conventional notions of radical and


269

incremental innovation. To the extent that radical innovations combine significant technical novelty and depth, they tend to result from R&D laboratories with deep science-based capabilities and strong upstream linkages, while incremental innovations tend to result from smaller engineering units with strong application focus and downstream linkages.6 These insights translate into the first proposition phrased in a Chandlerian vein: the organizational structure for managing innovation reflects innovation strategy as characterized by the type of interface focus (upstream or downstream), and the depth and novelty of the R&D processes.

B: The dynamics of the corporate technology base. Technology strategy involves two critical functions: to guide technology diversification and to promote technology integration. During the last decade, several empirical studies (Pavitt et al., 1989; Granstrand et al., 1990, 1997; Oskarsson, 1993; Gambardella and Torrisi, 1998) have demonstrated a relatively stronger inclination in technologybased large companies for technology diversification than for product diversification.7 A likely explanation is that the costs associated with in-house knowledge accumulation and coordination tend to be lower than the transaction costs associated with external sourcing, especially with regard to evolving, complex and/or tacit knowledge.8 One critical task for technology strategy is to guide technology diversification, that is, to identify new technologies which can substitute or supplement existing technologies in innovative processes, and to choose between different ways of getting access to these technologies. The second proposition assumes that the choice of organizational design for managing technology is a function of the cost and time horizon involved in technology diversification,9 on the one hand, and the ease with which the existing R&D personnel can assimilate the new prospective technologies, on the other hand. The proposition maintains that technology diversification involving low R&D costs and relatively familiar or codified knowledge can be organized in small R&D teams with strong overlaps to ongoing innovation projects. In contrast, technology diversification 6However, in

some technology-based industries (i.e. medical instruments) even radical innovations are generated primarily through downstream linkages with lead users (von Hippel, 1988; von Hippel et al., 1999). 7Gambardella and Torrisi (1998) even found that the most successful large electronics firms over the past ten years were those that simultaneously broadened the technology base and narrowed their already diverse product range. 8Moreover, Cohen and Levinthal (1990) have demonstrated that the capacity to absorb external knowledge depends on the level and depth of internal knowledge in related areas. That would imply that multi-technology companies with ‘deep’ technological knowledge foundation in numerous technologies have lower coordination costs in both internal and external relations concerning technology development and integration than companies with a more narrow and/or less ‘deep’ technology base. 9We assume a positive relationship between costs and time span of the investments due to timecompression diseconomies involved in R&D (Dierickx and Cool, 1989).

270


involving costly and longer term R&D investments will have to be organized separately from ongoing innovation practices, either in existing R&D labs if these possess a minimum of ex ante familiarity with the new technology, or, if that is not the case, in separate units dedicated to the technology diversification in question. The trend towards increasing diversity of the company’s technology base implies rising demands on cross-disciplinary learning and coordination in innovation. Hence, categories of innovations that were exclusively downstream-based increasingly also involve upstream activities. Therefore, another and increasingly important task for the management of technology is to identify and sustain objectives for inter-disciplinary (and possibly inter-divisional) coordination and learning, or in short, technology integration (Iansiti, 1997). The contingencies favouring decentralized versus centrally coordinated modes of technology integration can be specified in the third proposition: Decentralized technology integration, which is closely linked to ongoing innovation processes, is feasible when the technologies to be integrated are mature, that is, well-documented and codified, or embedded in components for modular product architectures (cf. Sanchez, 1996). In contrast, a central team or unit with strong corporate support and authority is required to coordinate technology integration if the technological knowledge is very complex (possibly science-based) and tacit. To sum up, in order for technology-based large companies to stay competitive, they have to master an increasing range of technological capabilities and establish interdisciplinary points of integration between them. The basic insights of the second and third propositions are that the organizational context for managing technology is constrained by, and constrains, the technology strategy as reflected in specific features of technology diversification (R&D costs and level of prior familiarity with the prospective technologies), and technology integration (level of complexity and tacitness of the knowledge underlying the technologies).

4. A Contingency framework Based on the analytical discussion in the two previous sections, the relations between strategic focus and organizational locus in, respectively management of innovation and management of technology shall be synthesized in a contingency framework. A decentralized mode of managing innovation can reflect the overall organizational structure of the company (i.e. an M-form unrelated or market-related diversifier) and/or, as illustrated in Table 3, an innovation strategy that gives high priority to incremental innovation with a primary concern for downstream, inter-functional relations and engineering-based R&D. Likewise, a centralized mode of managing innovation may reflect a more centralized corporate organization (i.e. a U-form vertical integrator or an M-form technology-related diversifier) and/or an innovation strategy focusing on radical innovation with primary focus on upstream or inter-disciplinary relations and in-depth or science-based R&D. A decentralized mode of managing technology (technology diversification and


271

Table 3 Management of innovation: strategic focus and organizational focus Focus of innovation strategy

Category of

Interface relations

Nature of R&D

Incremental

Primary focus on

Engineering-based

innovation

downstream, inter-functional

application focus

Organizational focus for management of innovation

innovation

Decentralized

relations Centralized

Radical

Primary focus on

‘Deep’ or

innovation

upstream interdisciplinary

science-based R&D

relations

integration) can reflect the overall organizational structure of the corporation and/or, as illustrated in Table 4, a technology strategy in which the R&D costs are relatively low, the technologies to be acquired or integrated relatively mature, and the knowledge structures embodied in the new technologies are similar or related to those of existing technologies. A centralized mode of managing technology can reflect a more centralized overall corporate structure and/or a technology strategy that involves large and concerted R&D investments in complex and tacit technologies, and the development of new technical knowledge that is not too dissimilar from the existing knowledge. Furthermore, a centralized structure is, if equipped with sufficient political authority, better positioned than a decentralized structure to exercise control over a highly distributed and decentralized corporate technology base (Combs and Richards, 1993). Finally, technology diversification, possibly associated with radical innovation, may be located outside the existing R&D organization in a special unit if the competence requirements are highly dissimilar to existing competencies and routines. In contrast, the subsequent integration of such technologies into the existing technology base will generally require organizational integration of the activities in the special unit into the existing R&D organization.10 In Sections 5 and 6 the two company cases will be presented and analysed.

10However, in case the newly acquired technology represent a new dominant technology with competence-destroying impacts on the existing technology base (Tushman and Anderson, 1986), the special unit is more likely to take over the role as the primary R&D organization of the company.

272


Table 4 Management of technology: strategic focus and organizational focus

Organizational focus for management of technology

Focus of technology strategy

Decentralized

Technology diversification

Technology integration

Low-cost R&D/similar

Codified/mature technologies

knowledge structures Centralized

High-cost R&D/not too

Complex and tacit knowledge

dissimilar knowledge

structures

structures Special unit

High-cost R&D/highly

Unlikely to be provided in a special

dissimilar knowledge structures

unit

5. The company cases: Danfoss and Grundfos11 5.1 Corporate strategy and structure The two companies, Danfoss and Grundfos, show both distinct differences and distinct similarities (see Table 5). Danfoss (founded in 1933) is a diversified company operating in thousands of products within refrigeration controls, motion controls, and heating and water controls. Grundfos (founded in 1945) is a focused company operating within pumps and pump systems. One category of pumps, circulator pumps, accounts for a dominant share of the company’s total turnover. During the last decades, a consistent contrast in their corporate strategy–structure development has surfaced. Danfoss has followed a strategy of related diversification (both market- and technology-related) assisted by an increasingly decentralized M-form structure which has also been associated with increasing emphasis on outsourcing. Grundfos has pursued a strategy of vertical integration (and practically no business diversification). Contrary to most other pump manufacturers, Grundfos has in-house development and manufacturing of electromotors, machinery and electronic controls dedicated to its own pumps. The vertical integration strategy is assisted by a quite centralized U-form structure. Until the early 1980s, economic growth in both companies was organic. But since then, Danfoss has primarily grown through an aggressive acquisition strategy, while Grundfos has remained a predominantly internally growing company. Thus, in terms of overall growth strategy and governance structure, the companies are close to being 11This section is based on comprehensive longitudinal case studies on the two companies. See Appendix on methodology.


273

Table 5 Similarities and differences between Grundfos and Danfoss

Danfoss

Differences

Similarities

Related diversification

Entrepreneurial history

Outsourcing Acquisition strategy

Family-based ownership Growth orientation Institutional continuity

Grundfos

Narrow business scope Vertical integration

Mass production Profile of technology base

Internal growth strategy

Industrial market base

ideal types of the related diversifier (Danfoss) and the vertical integrator (Grundfos) categories discussed in Section 2. However, in several respects the two companies share striking similarities. Both grew out of entrepreneurial organizations with strong founding personalities. Both have been highly growth-oriented for more than fifty years and have developed from being small artisan firms to becoming large and global companies while still remaining predominantly family-owned. Both have experienced long institutional continuity (Cantwell and Fai, 1999).12 Both have grown out of a mechanical engineering and craft-oriented base and have exploited a strategy of quality-oriented mass production. Moreover, both are dedicated to technological innovation and possess a technology base comprising broadly similar capabilities in mechanical and electronic engineering (mechatronics), new materials, fluid mechanics and software engineering. Finally, they are both today world-class companies with a primary focus on industrial markets. Within circulator pumps, Grundfos is a global market-leader controlling about half the world market. Danfoss is a global market-leader within radiator thermostats accounting for about 10% of Danfoss’ total turnover. In most of the many other product lines in which Danfoss operates, the global markets are generally smaller than the radiator thermostat market (and much smaller than Grundfos’ circulator pump market). Even if Danfoss is among the major global players in many of these product markets, it rarely holds a dominant position. In 2000, Danfoss had a turnover of 14.8 billion DKK and employed 17 000 people, while Grundfos had a turnover of 9.5 billion DKK and close to 11 000 employees.

12 Institutional continuity denotes companies with a long history without sudden radical reconfiguration of the corporate identity due to for example major mergers or wars (Cantwell and Fai, 1999).

274


5.2 Managing I&T in Danfoss Until the death of its founder, Mads Clausen, in 1966, Danfoss was a U-form company with top-down management structures. Since then Danfoss has undergone a series of organizational changes. In the 1970s towards a centralized M-form close to the strategic planning style (cf. Section 2); later, in the 1980s towards a more decentralized M-form broadly in accordance with the strategic control style, and in the 1990s some elements from the financial control style were integrated. This development has taken place as the company became more diversified and the product divisions grew increasingly self-contained.13 During its steep growth in the 1960s, management of I&T was carried on in a conventional R&D lab, referred to as Corporate Technology and Research (CTR). CTR was especially focused on long-term venture projects sometimes associated with quite radical innovation. There was no separate function dedicated to management of technology. However, significant processes of technology diversification and integration took place in the context of several venture projects. Through the 1970s, however, the newly established and expanding product divisions gradually developed their own capacity for managing innovation. From the early beginning, these activities were almost exclusively downstream-focused and incremental. Divisional innovation management gradually came to control resource allocation and project formulations in CTR. This resulted in increasing numbers of small short-term projects without much coherence. Even though CTR, due to strong leadership, regained significant autonomy in the 1980s by focusing on a more limited number of strategic R&D and venturing projects, this happened at the cost of increasing decoupling from the downstream and incrementally focused innovative activities in the divisions. As the product divisions gradually came to manage their own innovation processes, other activities than R&D-projects came to play an increasing role within CTR: technical extension services to the divisions, management of patents, standardization and certification, management of quality systems, and cross-divisional management of technology (see below).14 Thus, CTR experienced its originally clear and coherent R&D function gradually deteriorating, and increasingly CTR became the repository for a collection of tasks and services among which there were no, or only few, synergies. Along with this development, the prestige of the CTR declined, both among divisional and top management, and in 1996, it was decided to close CTR which by then had a staff of 150 people. The objective was not to downsize overall R&D but rather to spur divisional management to take full responsibility for R&D, and more thoroughly 13This tendency towards self-contained divisions has been sought countervailed through organizing the

10 product divisions into three divisional groups. 14While

most R&D in Danfoss was carried out in CTR in the 1960s and early 1970s, the CTR-based R&D in the early and mid 1990s only covered about 20% of total corporate R&D. By then, around onefourth of total costs in CTR was directly financed by the divisions. Divisional R&D expenditure varied from 2 to 10% of divisional turnover.


275

integrate technology and business strategies. Top management argued that CTR had become too much of an excuse for not building sufficient technological capabilities at divisional levels. However, already from the late 1980s, top management was concerned that divisionalization of R&D would lead both to increasing fragmentation of the corporate technology base, and to an increasing incrementalist bias in innovative activities. Concerning the former problem, a systematic approach to management of technology was initiated in 1989, when the director of CTR and some divisional R&D managers began to explore the opportunities for promoting cross-divisional sharing of technologies. A technology portfolio tool was developed with the objective to stimulate the creation and diffusion of technological capabilities. This technology portfolio contains a selection of technologies in which Danfoss can (or wants to) claim world-class expertise. These ‘key technologies’ have significant value for more than one division.15 In other words, this management of a technology tool implies a continuous reflection of the strategic prospects and priorities for the corporate technology base. At the same time, the tool promotes, on a voluntary basis, inter-divisional technology integration and concerted technology investments.16 After the closure of CTR, the technology portfolio activities were maintained at corporate level. These technology management efforts have exerted some guiding influence on technology diversification and some stimulus for cross-divisional technology exchange. However, by only emphasizing cooperation and consensus building and not applying any incentive mechanisms, the overall effects seem to have been limited. To countervail the latter problem of creeping incrementalism, two measures were initiated after the closure of CTR. First, the various corporate venture and business development projects, so far distributed in different parts of the corporation, were organized in a new corporate department with an exclusive focus on venturing.17 The strategic ambition behind this decision was to revitalize the yearlong tradition in Danfoss for entrepreneurial business venturing, a tradition that had suffered during the period of divisionalization. Data are still too preliminary to judge the impacts of this attempt to strengthen and coordinate venturing activities. Secondly, in order to create stronger bottom-up commitment to more radical and up-stream innovation and concerted technology development, a sponsoring project scheme was established. In this scheme, Headquarters may fund 50% of projects if the proposals (i) are backed up by at least two divisions, (ii) have a long-term explorative perspective, and (iii) do not have a natural home base within one of the divisions. One 15For technologies that are only important to one division, the division in question is expected to take full responsibility. 16Associated to the technology portfolio tool is a technical advisory group, numerous cross-divisional committees that regularly review the content of the portfolio, and several gatekeepers responsible for monitoring and promoting developmental activities. 17Examples of

these ventures are water hydraulics and analytical sensors.

276


current example is a programme for upgrading software competencies relating to seven product lines. A small team from the corporate level coordinates a number of software projects in order not only to develop specific software components, but also to accumulate and share capabilities in software development. Generally, however, the sponsoring scheme has not been widely used and known in the corporation, and innovative efforts at divisional level have remained predominantly downstreamfocused and incremental.

5.3 Managing I&T in Grundfos Grundfos has remained a quite centralized U-form company and has never seriously considered taking the step into the M-form. From the 1980s and onwards, central control has been challenged by the need for internationalization of sales and still larger parts of manufacturing. Even if internationalization has paved the way for some delegation of responsibility to regional sales and production centres, Headquarters in Bjerringbro still maintains a firm strategic control. In the pump industry, Grundfos is perhaps the most high-profiled in terms of innovation and R&D. In 1998, R&D accounted for over 4% of turnover. R&D is primarily focused on product and process development, and product development is targeted at making pumps smaller, increasing their performance, making them more user-friendly, making them consume less energy, and producing pumps for new applications. Until the mid 1980s, product and process development activities were assembled in one unit in Bjerringbro. This created a very integrative approach to product and process development. However, in 1985 all development activities were transferred to the individual factory sections dedicated to particular pump categories. This was done to create closer integration between development activities and manufacturing operations. Product innovation indeed became more downstream-linked to manufacturing and, as a side-effect, also more incremental because product changes primarily had to adapt to manufacturing requirements. Some new problems, however, emerged. Since development tasks were carried out by employees who normally were engaged in operational tasks, the development projects tended to become excessively long and costly. Another problem was the lack of coordination between especially tool and machine construction across the individual factory sections, implying, for example, that several almost identical tools were constructed every year. During this same period several competitors began to catch up with Grundfos’ traditional leader position by imitating unique Grundfos designs. Acknowledgement of these problems gradually created momentum in favour of centralizing R&D and establishing a project organization. In 1987, top management decided to merge the different machine and tool construction sections, as well as the research department with its associated materials laboratory, forming a new centre to be termed the Technology Centre (TC) close to Headquarters’ main building. The objective was to build a strong and concerted technological platform that would help maintain and expand the company’s leading position in vital fields such as stainless


277

steel, new materials (e.g. composite materials), surface treatment, and the development of process technologies. The centre was inaugurated in 1990 and represented the largest investment in the history of Grundfos. By 1998, TC employed around 350 R&D employees and was financed predominantly from the Group. Not only did TC reflect a centralization of all process innovative activities, it also signified the integration, within one building, of management of process innovation and the management of technology (for both product and process innovation). While TC was being built, the organization of product development was discussed at a series of strategy meetings. The product development departments were perceived to have become too self-contained within their respective production sections and reluctant to engage in close coordination. This was considered necessary not only to obtain economies of scope in product design and component sourcing, but also to sustain the inter-disciplinary learning required for integrating a growing number of complex technologies used in product development. The discussions resulted in a decision to join all product development departments in one Development Centre. A few years later, in 1993, product developers and product managers were transferred to the newly built Development Centre (DC) that was organized according to project- and matrix-based principles. Thus, DC not only represented a centralization of all management of product innovation activities, it also represented an organizational renewal more favourable to upstream and radical innovation. This new approach is also reflected in the interior design of DC, where an open work environment facilitates communication and exchange of knowledge. Moreover, the close juxtaposition of TC and DC makes direct communication among the altogether nearly 700 employees very easy. Prior to the decisions to centralize and strengthen R&D, top management had come to recognize that control electronics, a field in which Grundfos like most of its competitors had no expertise, would very likely become a critical part of pumps in a not too distant future. Consistent with the general vertical integrator philosophy, Grundfos decided, in the mid 1980s, to become a first-mover with respect to integrating electronics into pumps. The starting point was the ambition to develop a robust and low-price micro-frequency converter that makes possible the continuously variable control of speed in the pumps. A venture project was initiated at a location in Bjerringbro, separate from the existing R&D activities. This unit developed a strong entrepreneurial culture quite distinct from that of the well-established mechanical engineering culture. In cooperation with large foreign semiconductor companies, Grundfos developed and learned to produce ASICs (Application Specific Integrated Circuits), and in 1991 Grundfos could launch an effective micro-frequency converter dedicated to its own pumps. This venture was a highly successful process of technology diversification, and it was furthermore a radical technical innovation giving Grundfos a competitive lead for several years. Today, approximately 400 people work in the electronics factory. In 1995, a Business Process Reengineering (BPR) analysis of the overall organization

278


of R&D resulted in a decision to create a unitary R&D division with all R&D activities gathered in Bjerringbro. This first of all implied an organizational coordination of TC and DC. The main principle behind the structure of the R&D division was to align a technological competence and a commercial application perspective. When the R&D division was implemented in 1996, it was estimated that the electronics expertise was still so immature in Grundfos that it would not yet be desirable to separate electronics development from production. However, just about one year later, the development activities of the electronics factory were organizationally and physically integrated in the main R&D organization. Electronics and mechanical engineering development is now considered of equal value for development work in Grundfos and with huge scope for inter-disciplinary learning.

6. Analytical discussion The objective of the analysis of the company cases is threefold. First to clarify to what extent the mode of governing I&T is determined by the general corporate strategy– structure or by the specific innovation and technology strategies. Secondly, to appraise the pertinence of the three propositions on the interplay between strategies and governance structures of I&T. Finally, to reflect on the possible causes behind the very different trajectories of the two companies in terms of general corporate strategy and governance, and management of I&T.

6.1 The role of corporate strategy and structure The case studies demonstrate that both companies have been subject to distinct and consistent trajectories in their strategy–structure profiles. It seems as if the two companies, from their early years, had been genetically coded for specific trajectories of corporate development. In the Danfoss case, the trajectory of related diversification associated with an M-form with some centralized features, and in the Grundfos case, the trajectory of vertical integration associated with a centralized U-form. Moreover, since Danfoss began to implement the M-form in the early 1970s, the two companies have pursued very different organizational modes of managing I&T. Danfoss has witnessed divisionalization of practically all management of innovation. This development is not a result of specific changes in innovation strategy, but the outcome of the general forces underlying the divisionalization process, namely the incentives for expanding the autonomy and financial accountability of the product divisions. Management of technology was in the ‘golden era’ of CTR integrated in more radical innovation and venture activities, and as divisions gradually took over the responsibility for innovative activities, management of technology tended to become reduced to add-on status in relation to innovation projects. This implied that technology diversification and integration predominantly came to involve quite mature technologies that could be acquired in the market-place. At corporate level, a few more radical venture and technology projects were maintained and attempts of stimulating cross-divisional technology development and sharing were made by means of the


279

technology portfolio tool. Also these tendencies in the practice of management of technology are rather the outcome of the overall corporate changes than of specific changes in technology strategy. In the Grundfos case, management of I&T has not to the same degree been superimposed by the overall corporate dynamics. While the corporation as a whole has maintained its quite centralized U-form structure, the organization of I&T has undergone quite substantial changes from decentralization and integration with the manufacturing function in the mid-1980s to centralization and relative autonomy from operations in the 1990s. Even if vertical integration has also for decades been a powerful philosophy in the management of I&T, the different modes of organizing I&T has been much more influenced by specific fluctuations in innovation and technology strategies.

6.2 The role of innovation and technology strategies According to the first proposition (see Section 3), the organizational structure for managing innovation reflects the innovation strategy as characterized by the type of interface focus and the depth and novelty of the R&D processes. In the Chandlerian tradition, this translates into a causal chain in which structure follows strategy. However, in the case of Danfoss the causal change is the reverse: divisionalization of technological innovation has had a strong impact on the de facto innovation strategy. Even if a full account of the specific styles of managing innovation in each of the ten divisions is not available, our general impression from the case study is that divisionalization of innovation has implied an increased focus on inter-functional linkages as related to the limited business scope of the individual divisions. In contrast, the focus on inter-disciplinary linkages, in order for instance to integrate new technologies into the products, generally plays a secondary role. Furthermore, divisional R&D tends to be primarily concerned with applied R&D addressing the specific requirements of ongoing product development, and less concerned with fundamental R&D required for building new distinctive technological capabilities (see below on management of technology). This means that by far the most innovative efforts are dedicated to incremental and custom-specific product development.18 This is not only due to divisionalization per se, but rather to the associated tendency towards controlling the divisions through financial performance indicators. Prior to Danfoss’ embarkation on the divisionalization trajectory, the central R&D lab was responsible for most innovative activities. The lab was congruent with the then functional corporate structure. Even if data concerning the objectives of the lab’s innovation strategy in the 1960s and early 1970s are sparse, the long-term venture activities seemed to play a prominent role. During the era of divisionalization, the central lab increasingly showed incongruity vis-à-vis both the overall tendency towards divisional autonomy and accountability, and the emerging divisional innovation strategies. This incongruity was reflected in increasing difficulties in aligning the 18In contrast, some of the venturing projects, especially those organized in the corporate venture function, possess substantial elements of quite radical technical renewal.

280


long-term horizon of the central lab and the short-term commercial inclination of the divisional R&D units. In the wake of the closure of the central lab, attempts were made to counterbalance the incrementalist bias in divisional innovative efforts by a sponsoring scheme which allows the Group to co-fund more radical and cross-divisional innovative investment projects. However, this has so far only to a marginal extent contributed to change the short-term orientation of divisional R&D. This illustrates the difficulties of many diversified companies in creating efficient internal capital markets for longer-term R&D (cf. Klein, 2001), especially under a corporate governance more prone to financial control style than to strategic planning style (cf. Goold and Campbell, 1987). In Grundfos, the dynamics of management of I&T is much more in accordance with the conventional ‘structure follows strategy’ sequence. Grundfos has not experienced a similar departure from the functional structure, and R&D has therefore not been exposed to strong pressures for decentralization incurred by the overall strategy– structure development, as has been the case in Danfoss. However, the increasing proliferation and internationalization of, in particular, the manufacturing operations, have given rise to considerations on how to improve the coordination between product development, process development and manufacturing operations. The relative decentralization of R&D to the individual factories in the early 1980s reflected both a deliberate attempt to establish tighter coordination of innovative and operational activities in order to obtain better ‘design for manufacturability’,19 and the relatively strong political position of the production side within the corporation. Referring to the first proposition, the new organization reflected a strategic emphasis on interfunctional relations between development and operations. However, the increasing recognition of inefficiencies incurred by the co-location of R&D and operations contributed to create momentum for the turnaround in innovation strategy in the late 1980s, implying a strong determination to regain the position as the technological and innovative frontrunner of the pump industry. This strategy underlies the re-centralization during the 1990s of all R&D within the co-located Technology Centre and Development Centre. Even if the centralized nature of technological innovation can be said to match the company’s narrow product focus and functional organization, the more specific organizational set-up for innovative activities reflects well-digested objectives in the company’s innovation strategy. Grundfos has traditionally been a strongly production- and technology-driven company with a focus on the inter-functional interfaces between product development, process development and manufacturing operations. The previous decade’s substantial investments in fundamental technology development and increasing focus on inter-disciplinary linkages across technologies, while maintaining close interaction between product and

19This

decision was probably influenced by a strong consultancy wave during the same period in Denmark favouring so-called ‘integrated product development’ implying, among other things, integration of product development and manufacturing operations.


281

process development20 demonstrates a concerted commitment to be a first-mover with respect to implementing quite radical technical renewal of products and processes. Turning to the issue of management of technology, both companies have expanded their portfolio of technological capabilities, and even if the diversification process has followed similar paths (for instance to build expertise in electronics, software and materials technologies), the specific proliferations have differed. Probably Danfoss has a more diverse technology base when specified at a more disaggregate level due both to its larger size and to the more diverse product range. While technology diversification in Danfoss takes place within the divisions without any significant corporate coordination, technology diversification in Grundfos is centrally controlled. Thus, the decisions to build strong capabilities within materials research and control electronics were taken by top management in conjunction with the R&D management. The second proposition (see Section 3) suggests that the organizational context for managing technology reflects the strategy of technology diversification in terms of the cost and time horizon of R&D, and the organization’s prior familiarity with the prospective technologies. In the Danfoss case, the divisional structure imposes a low-cost and short-term technology diversification strategy which implies a technology base consisting of relatively mature technologies. By contrast, the up-front technology strategy in Grundfos gives rise both to a unified organization and to new separate units (for control electronics and materials research) for dealing with radical technology diversification requiring competencies very different from those possessed by the established organization. With respect to technology integration, we also see contrasting dynamics. In Danfoss, technology integration, involving cross-disciplinary learning, is subject to substantial barriers to the extent that this learning also involves cross-divisional dialogue and exchange.21 Particularly the decentralization of corporate R&D has contributed to this development. Even if corporate measures such as the crossdivisional committees (including the technology portfolio tool) and sponsoring projects to some extent promote such learning, the measures are voluntary by nature and not assisted by any incentive mechanisms and therefore rarely lead to distinct cross-divisional projects. In Grundfos, cross-disciplinary learning is systematically dealt with in the matrix-like organizational set-up of the R&D division, in which professional and functional boundaries are sought eliminated. The recent attempts to organizationally and physically integrate the electronics development activities into the existing R&D organization reflect the management’s judgement that the electronics knowledge assets within Grundfos have by now matured to an extent making it feasible to integrate the development part into the main R&D organization. The third 20However,

the interface between product and process development, on the one hand, and manufacturing operations, on the other hand, is likely to have been weakened as a consequence of the centralization of innovation processes. 21One exception is that one of the divisional groups, Refrigeration Controls comprising three divisions, has set up a common R&D unit to deal with cross-divisional R&D-projects.

282


proposition (see Section 3) suggests that the organization of technology integration reflects the level of complexity and tacitness underlying the technologies. The decentral organization in Danfoss corresponds to technology integration closely related to innovative activities and involving relatively mature technologies, while the centralized organization in Grundfos corresponds to technology integration involving more complex and tacit technological knowledge. Also here, we see that in Danfoss structure determines integration strategy, whereas in Grundfos the reverse is the case. To conclude, the drivers of the modes of managing I&T are highly different in the related diversifier M-form case from the vertical integrator U-form case. The related diversifier tends to impose an organization on management of I&T that severely constrains the strategic scope for innovation and technology strategy. Even if these dynamics can be countervailed by different measures, such measures probably have to be assisted by strong incentive systems or cultural ethos (as the case for many years has been in 3M, see Christensen, 2000) in order to play more than a marginally countervailing role. In the Danfoss case, management control of the divisions primarily focuses on financial performance. This is reflected in the divisional approach to managing innovation and technology, namely a strong concern for relatively short-term commercialization and amortization of R&D investments. In other words, economic incentives drive divisions away from explorative, high-risk and long-term orientation in R&D projects and towards incremental low-risk projects. The vertical integrator may be less constrained by the overall strategy–structure constellation when choosing its mode of managing I&T. It is easier for such a company to choose its distinct innovation or technology strategy and design an organization that is appropriate for the implementation of the strategy. Compared to the related diversifier, this is a particular strength when more radical innovation and technology strategies are required. In the 1990s, Grundfos has shown a commitment to large explorative investments in high-risk projects with a long time horizon. A similar commitment would have been difficult to mobilize in Danfoss. These investments are funded by the corporation and not, to the same degree as R&D investments in Danfoss, subject to financial management control of their commercial pay-off. As a diversified company with strong financial performance incentives, Danfoss has better opportunities for amortizing R&D investments within the individual divisions and product lines than Grundfos where a larger share of the R&D investments has been guided more by managerial judgement of long-term options than by the prospects for short-term commercial pay-off. On the other hand, Grundfos has better options, when needed, to mobilize large corporate resources in concerted efforts to transcend technological discontinuities and jump unto new technological trajectories. This, however, also incurs the risk of over-investing in promising options that later prove to be technological dead ends.

6.3 Why did the companies come to pursue different overall strategies? What explains the radically different strategies pursued in the two companies that


283

originated with very similar features? The case studies seem to indicate that highly robust strategy–structure trajectories of different industrial companies can be explained especially by three mutually reinforcing factors: The idiosyncratic inclinations of the founder/entrepreneur or top managers in the early stages of a company’s history, the nature of the first successful products, and the market prospects of these products. At least in the two cases presented in this paper, these factors seem to have mobilized a sort of firm-specific genetic codes inducing the firms to follow distinctly different trajectories. Both founders were classical entrepreneurial personalities possessing technical ingenuity for product innovation and a drive for risk-taking and commercial enterprise. Moreover, after the formative years of their respective companies, both conducted a paternalistic management style which somehow matched the quite centralized U-form structure that emerged. More specifically, both were highly occupied with implementing ‘rational’ mass production systems which were also supported through the American Marshall program for economic reconstruction of Western Europe after World War II. However, they differed with respect to their concern for new ventures. While Poul Due Jensen (Grundfos) quite early on decided to stick to pumps and focus on constructing improved and new types of pumps and implementing mass production, Mads Clausen remained virtually obsessed with searching for new business ventures. Thus, when Mads Clausen died in 1966, Danfoss was already quite diversified, and without the paternalistic entrepreneur at the helm, the inherent tensions between the unitary form and the increasingly diversified business practices, as reflected in severe administrative overload at the Headquarters, became manifest. In contrast, when Poul Due Jensen died in 197722 the unitary structure was basically well-matched to accommodate the overall corporate strategy. The nature of the early product markets may similarly have contributed to form the different strategic routes on which the two companies eventually embarked. The first products launched by Mads Clausen were quite simple valves imitated from an American company and dedicated for specific cooling applications with initially quite limited market prospects. However, the valve principle was subsequently adapted and further developed for other applications and markets. Thereby a related diversification dynamics emerged, and the new application and market areas again stimulated the building of new capabilities (in e.g. hydraulics or electronics) that could provide yet new diversification options, and so forth. Poul Due Jensen, on the other hand, very early on developed an automatic pump for a local water utility, and this and subsequent pump constructions proved to be relatively complex product innovations that contributed to drive the hand-pump out of the market. In other words, while the early Danfoss valves were imitations or incremental improvements of components for limited market segments, the pump innovation, especially the centrifugal pump from 1950, was a more radical innovation with much larger market prospects. 22Both entrepreneurs

died in their sixties as highly active top managers.

284


These imprints of the founding fathers’ strategic inclination, the innovative nature of their early products, and the market prospects of these products have had longstanding impacts on the strategies and structures of their companies, even several decades after the death of the founders. In many respects, the two companies have been operating within the same general types of technological trajectories (from mechanical engineering to mechatronics, from craft production to mass production) which, compared to more science-based trajectories, generally require stronger downstream relations in R&D but not necessarily central R&D labs with a strong science foundation. In other words, the different strategic and organizational trajectories of these particular companies cannot be explained by differences in the technological trajectories they have pursued.

7. Conclusion Most of the literature on the strategy and structure of large companies has dealt with the conventional M-form. Contingency theory argues that this M-form is especially appropriate for companies that diversify along unrelated paths. During the last two decades, there has been increasing interest in investigating companies that follow other strategies (some kind of related diversification or vertical integration) and are governed by other governance structures than the conventional M-form (the U-form and some kind of CM-form23). For technology-based companies, the conventional M-form is a rare governance structure. Rather they tend to follow either a U-form or some kind of CM-form, sometimes even a matrix-like CM form. This paper has sought to provide an analytical approach to understand what determines the mode of managing I&T in technology-based companies that follow either a related diversification strategy or a vertical integration strategy. It has been argued that the specific modes of managing I&T cannot be understood exclusively on the basis of the general strategy–structure profile of the companies. A more adequate understanding has to add a separate analysis of the firm-specific features characterizing the innovation processes (interface focus, and the depth and costs of R&D) and the dynamics of technology diversification and integration. The suggested propositions concerning the interplay between innovation and technology strategy and the management and organization of I&T were illustrated and supported by two case studies. But the case studies also revealed that the determinants of a specific mode of organizing I&T may differ. In Danfoss, the development of the organization of I&T has been primarily determined by the overall development in corporate strategy and structure, and de facto innovation and technology strategies have adapted accordingly. In Grundfos, the development of the organization of I&T has to a larger extent been determined by specific objectives of innovation and technology strategies. This may point to a more general problem widely discussed in recent years, that many diversified incumbents as 23For Williamson, CM-form connoted the corrupted M-form, while Hill and Hoskisson (1987) used the term to imply an M-form with some centralized features.


285

they mature grow less entrepreneurial and less able to keep pace with innovation challenges from newcomers despite their proclaimed radicalism in innovation and technology strategy, and despite their superior access to financial and technological assets to support such strategy (Christensen, 1997). The explanation may relate to the insights from this paper that some growing and diversifying companies gradually see their de facto innovation and technology strategies constrained by the increasingly decentralized and financially controlled governance structures. Further research should explore the significance of organizational incongruities between the strategy and structure of I&T and the more general corporate strategy– structure trajectory, for not only creating incentive and agency problems, but also for creating momentum for organizational renewal and innovation.

Acknowledgements I am grateful for the constructive comments on earlier drafts from Lee Davis, Hans Jørgen Pedersen, Mikael Iversen, Keld Laursen, Morten Overgaard Nielsen and Finn Valentin, and two anonymous referees.

Address for correspondence Department of Industrial Economics and Strategy, Copenhagen Business School, Howitzvej 60, 5.2000 Frederiksberg, Denmark; [email protected].

References Argyres, N. S. (1995), ‘Technology strategy, governance structure and interdivisional coordination,’ Journal of Economic Behaviour and Organization, 28, 337–358. Boutellier, R., O. Gassmann and M. Von Zedtwitz (2000), Managing Global Innovation, 2nd rev. edn. Springer Verlag: Berlin. Burns, T. and G. M. Stalker (1961), The Management of Innovation. Tavistock: London. Cantwell, J. and F. Fai (1999), ‘Firms as the source of innovation and growth: the evolution of technological competence,’ Journal of Evolutionary Economics, 9, 331–366. Chandler, A. D. (1962), Strategy and Structure. Chapters in the History of the Industrial Enterprise. MIT Press: Cambridge MA. Chandler, A. D. (1977), The Visible Hand: The Managerial Revolution in American Business. Belknap Press: Cambridge, MA. Chandler, A. D. (1991), ‘The functions of the HQ unit in the multibusiness firm,’ in R. P. Rumelt, D. E. Schendel and D. J. Teece (eds), Fundamental Issues in Strategy. A Research Agenda. Harvard Business School Press: Boston, MA. Christensen, C. M. (1997), The Innovator’s Dilemma. Harvard Business School Press: Boston, MA. Christensen, J. F. (1995), ‘Asset profiles for technological innovation,’ Research Policy, 24, 727–745.

286


Christensen, J. F. (2000), ‘Building innovative assets and dynamic coherence in multi-technology companies,’ in N. J. Foss and P. L. Robertson (eds), Resources, Technology and Strategy. Routledge: London. Christensen, J. F., M. O. Nielsen and M. Iversen (2002), Grundfos—Strategy, Structure and Management of Technology, Department of Industrial Economics and Strategy, Copenhagen Business School: Copenhagen. Cohen, W. M. and D. A. Levinthal (1990), ‘Absorptive capacity: a new perspective on learning and innovation,’ Administrative Science Quarterly, 35, 128–152. Coombs, R. and A. Richards (1993), ‘Strategic control of technology in diversified companies with decentralized R&D,’ Technology Analysis an Strategic Management, 5, 385–396. Cooper, R. (1988), ‘The new product process: a decision guide for management,’ Journal of Marketing Management, 3, 238–255. Dierickx, I. and K. Cool (1989), ‘Asset stock accumulation and the sustainability of competitive advantage,’ Management Science, 35, 1504–1511. Gambardella, A. and S. Torrisi (1998), ‘Does technological convergence imply convergence in markets? Evidence from the electronics industry,’ Research Policy, 27, 445–463. Gerybadze, A. and G. Reger (1999), ‘Globalization of R&D: recent changes in the management of innovation in transnational corporations,’ Research Policy, 28, 251–274. Goold, M. and A. Campbell (1987), Strategies and Styles. The Role of the Centre in Managing Diversified Corporations. Basil Blackwell: Oxford. Goold, M., A. Campbell and M. Alexander (1994), Corporate-level Strategy: Creating Value in the Multibusiness Company. John Wiley: New York. Granstrand, O., C. Oskarsson, N. Sjöberg, N. and S. Sjölander (1990), ‘Business strategies for new technologies,’ in E. Deiaco and G. Vickery (eds), Technology and Investment. Critical issues for the 1990s. Pinter: London. Granstrand, O., P. Patel, and K. Pavitt (1997), ‘Multi-technology corporations: why they have ‘distributed’ rather than ‘distinctive core’ competencies,’ California Management Review, 39, 8–25. Hill, C. W. (1988), ‘Internal capital market controls and financial performance in multidivisional firms,’ The Journal of Industrial Economics, XXXVII, 67–83. Hill, C.W. and R. E. Hoskisson (1987), ‘Strategy and structure in the multiproduct firm,’ Academy of Management Review, 12, 331–341. Hoskisson, R. E. (1987), ‘Multidivisional structure and performance: the contingency of diversification strategy,’ Academy of Management Journal, 30, 625–644. Hoskisson, R. E. and M. A. Hitt (1988), ‘Strategic control systems and relative R&D investments in large multiproduct firms,’ Strategic Management Journal, 9, 605–621. Iansiti, M. (1997), Technology Integration: Making Critical Choices in a Dynamic World. Harvard Business School Press: Boston, MA. Iversen, M. and J. F. Christensen (1999), Danfoss – Corporate Strategy, Economic Organization and Management of Technology. Department of Industrial Economics and Strategy, Copenhagen Business School: Copenhagen.


287

Klein, P. (2001), ‘Are internal capital markets good for innovation?,’ working paper, Department of Economics, University of Georgia. Markides, C. C. and P. J. Williamson (1996), ‘Corporate diversification and organizational structure: a resource-based view,’ Academy of Management Journal, 39, 340–367. Miyazaki, K. (1995), Building Competences in the Firm. Lessons from Japanese and European Optoelectronics. Macmillan: London. Nonaka, I. and H. Takeuchi (1995), The Knowledge Creating Company. Oxford University Press: Oxford. Oskarsson, C. (ed.) (1993), Technology Diversification—The Phenomenon, its Causes and Effects. Chalmers University of Technology, Department of Industrial Management and Economics: Göteborg. Pavitt, K. (1998), ‘Technologies, products and organization in the innovating firm: what Adam Smith tells us and Joseph Schumpeter doesn’t,’ Industrial and Corporate Change, 7, 433–449. Pavitt, K., M. Robson and J. Townsend (1989), ‘Technological accumulation, diversification and organization in UK companies 1945–1983,’ Management Science, 35, 81–99. Pfeffer, J. and G. Salancik (1978), The External Control of Organizations: A Resource Dependence Perspective. Harper and Row: New York. Porter, M. (1985), Competitive Advantage: Creating and Sustaining Superior Performance. Free Press: New York. Riahi-Belkaoui, A. (1997), ‘Multidivisional structure and productivity: the contingency of diversification strategy,’ Journal of Business Finance and Accounting, 24, 615–627. Rothwell, R., C. Freeman, A. Horlsey, V. T. P. Jervis, A. B. Robertson and J. Townsend (1974), ‘SAPPHO updated—project SAPPHO phase II,’ Research Policy, 3, 258–291. Sanchez, R. (1996), ‘Strategic product creation: managing new interactions of technology, markets, and organizations,’ European Management Journal, 14, 121–138. Teece, D.J. (1982), ‘Towards an economic theory of the multiproduct firm,’ Journal of Economic Behaviour and Organization, 3, 39–63. Tidd, J., J. Bessant and K. Pavitt (1997), Managing Innovation. Integrating Technological, Market and Organizational Change. John Wiley: Chichester and New York. Tushman, M. L. and P. Anderson (1986), ‘Technological discontinuities and organizational environments,’ Administrative Science Quarterly, 31, 439–465. Wang, Q. (1997), ‘R&D/marketing interface in a firm’s capability-building process: evidence from pharmaceutical firms,’ International Journal of Innovation Management, 1, 23–52. Williamson, O. E. (1975), Markets and Hierarchies: Analysis and Anti-trust Implications. The Free Press: New York. Von Hippel, E. (1988), The Sources of Innovation. Oxford University Press: New York. Von Hippel, E., S. Thomke, and M. Sonnack (1999), ‘Creating breakthroughs at 3M,’ Harvard Business Review, September–October, 47–57.

288


Appendix: methodology of case studies The case studies on Danfoss and Grundfos are based on numerous interviews with managers in the two companies, internal documents, annual and other public reports from the companies, and articles from newspapers and the business press. Interviews and data collection have taken place regularly since 1994 in the Grundfos case, and since 1996 in the Danfoss case. The case studies have been conducted and written jointly by Jens Frøslev Christensen and Mikael Iversen (the Danfoss case) and Morten Overgaard Nielsen (the Grundfos case). The studies have been published separately as case studies (see Iversen and Christensen, 1999; Christensen and Nielsen, 2002).