hello
Journal of Engineering and Technology Management, 8 ( 1991 ) 67-83 Elsevier Towards a new theory of innovation management: A case study comparing Canon, Inc. and Apple Computer, Inc. Ikujiro Nonaka Institute of Business Research, Hitotsuhashi University, Kunitachi, Tokyo, Japan Martin Kenney Department of Applied Behavioral Sciences, University of California, Davis, CA 95616, USA Abstract This paper argues that innovation can be best understood as an information process which is then concretized as a product that meets demand. Two very different firms, Canon Inc. and Apple Computer Inc., are used as case study illustrations. Innovation does not proceed through logical deduction, but rather is furthered by the use of metaphors and analogies. The bureaucratic and staid structures of the firm can be challenged and broken up to provide the space for innovations to emerge. The leader’s role in the innovating firm is as a catalyst and facilitator, not as an allknowing despot. The importance of innovations is not merely in the new product, but also the “ripple” effects of innovations which can propel the firm into a self-renewal process. Keywords: Innovation management, High-technology, Case study. 1. Introduction Increasingly, corporate competitive success is hinging upon the effective management of innovation. Innovation has been the object of considerable academic study from a variety of perspectives. However, innovations are usu- ally considered as objects. We choose to look at innovation differently. For us, innovation is a process by which new information is created, and it is this information that is embodied in the product. To understand this process we conceptualize human beings not merely as information processors (Galbraith, 1973), but more importantly as information creators. Inherently, innovation is the process by which new information emerges and is concretized in a prod- uct that meets human needs1. The healthy firm is a negative-entropy system 1For one of the most interesting treatments of the process by which the inventor or innovator imagines the new development, see Usher (1954). 0923-4748/91/$03.50 ( 1991—Elsevier Science Publishers B.V. 68 which constantly creates new order and structure in its struggle to survive and grow2. 2. Approaches to innovation We break with the model posited by Simon (1969) and elaborated by Abernathy (1978) of innovation as problem-solving. In fact, most important in the innovation process is the problem creation moment. That is, the positing of the correct problem, which allows the solution to be discovered (information to be created). The key to the creation process is information. Information can be divided into two categories: syntactic and semantic (Machlup and Mansfield, 1983; Nonaka, 1987). Syntactic information can be reduced to a digital form which has no inherent meaning. This is exemplified by the discrete type of information used in computers and which can be manipulated through logical operation. The process of interpreting the results gives this type of information value and meaning3. Semantic information is qualitative; what is important here is the meaning or content of the information4. Semantic information is more holistic and capable of evolving and transforming. It is not created in the traditional deductive model hypothesized as the way Western science operates, but rather through insights which allow the creation of new models5. The tools in this creation process are often metaphors and analogies (i.e., devices to help in rethinking or even discarding old ways of thinking). Throughout this paper “information” refers to semantic rather than syntactic information. Thus, information creation is used synonymously with meaning creation. Product innovation can be considered as a restricted subset of this creation process. However, it would be a mistake to think that the meaning creation process either can be or should be restricted to product development; in a larger sense, the “culture” of the firm is semantic information. Informa- 2 This analogy, however, is not that of the Parsonian structural functionalism. See, for example, Parsons (1951). Rather, we are using chaos models, or what one might term "emergence" models, more in keeping with Prigogine ( 1980 ). See also Prigogine and Stengers (1984). For further discussion, see Nonaka (1988a). 3 Artificial intelligence is the technique being applied to make the more routine aspects of interpretation amenable to machine control. See Daft and Weick (1984) tor a discussion of organizations as interpretation systems. 4 This type of information is in sharp contrast to that manipulated by traditional academic economists which is assumed to be based on so-called "hard" data. The real world of business is more often predicated on hunches, "gut" feelings and inarticulable experience. 5 "There is an increasing literature in the sociology and history of science which argues that the traditional deductive model of science does not describe the way science is actually done. See, for example, Latour and Woolgar (1979), and Rose and Rose (1976). 69 tion creation can and must go on at every level and in every part of the firm6. For example, the development of new process innovations within quality control circles is an aspect of information creation. To remain competitive any firm must constantly be creating new strategies, new products, and new ways of manufacturing, distributing and selling. Constant reexamination, reconceptualization and reorganization are necessary—and this entails and requires the constant creation of new meanings. The newly created information must then be diffused throughout the firm setting off further innovations. This diffusion within the firm is important because it allows the firm to reap more of the benefits of its newly created information. As Schumpeter observed, new innovations trigger other innovations in an effect that resembles the dropping of a stone in a still pond. If the company can internalize some of the resulting “ripples”, then it can secure greater economic benefits. It should also be noted that often the “ripple” effects that emerge cannot be predicted, but result in major economic advantage. Thus, the innovation process becomes a moment in the constant evolution and transformation of the firm. Though different from that of Schoonhoven and Jelinek (1990, p. 107), we provide a model for conceptualizing the “output” of the “quasi-formal structure” of the firm. It is from this quasi-structure of teams, task forces and/or committees that new meaning (semantic information) emerges. The formal structure is devoted to transmitting syntactic information, necessary for operating the firm, but unable to create the new meanings necessary for growth and development. Hence, we find ourselves in agreement with Schoonhoven and Jelinek, but believe that the “ad hocracy” proposed by Burns and Stalker (1961) provides valuable insight into the operation of the quasi-formal structure. In this paper we compare the product development process in two firms that have reputations for innovation. We will show how the information creation process operated and how that assisted in the self-renewal of the firms. The Japanese case study is the development of the Mini Copier by Canon. The U.S. case study is the development of the Macintosh Computer by Apple Computers. For each company these were crucial product developments. The discussion following the presentation of the cases will reflect upon the dynamics of the information creation process in both companies and the lessons for management. Further, we will discuss the advantages and disadvantages of the Canon and Apple methods of organizing and encouraging the creation process. 6In many American firms information creation is limited to the top executives and perhaps their strategic management group. They then transmit the new information down through middle-level managers who are seen as information channels and not information creators. This perspective of middle-level managers is critiqued in Nonaka (1988b). For another critique of this style of management, see Peters (1988). 70 3. Case study 1: The Canon Mini Copier—New product development and information creation7 Canon was founded in 1933 with the aim of developing and manufacturing a 35 mm camera. After World War II the company once again began to manufacture cameras. Throughout the 1950s the company grew rapidly, developing a series of new single reflex cameras. By the early 1960s, however, it became evident that Canon would have to diversify into new areas to maintain its growth8. Thus, in the 1960s Canon began to diversify into office machinery through the development of electronic calculators and copying machines. In the mid-1970s, though, Canon ran into severe difficulties as demand for existing products declined precipitously and Canon was forced to retreat from the low-priced calculator market. In 1975 Canon had to suspend its dividend payments—a radical step of a Japanese firm9. In the words of the Canon Handbook (1987): “Canon was experiencing a managerial crisis. Top management lacked coherence and direction; the company's inflexible corporate structure made it impossible to deal effectively with divers)fication, and insufficient attention had been paid to the rationalization of production. Or, as the new President, Ryusaburo Kaku has put it, ‘Canon was like a ship that constantly changes course and goes nowhere’.” After a wrenching organizational restructuring Canon once again prospered, with an average annual growth rate of over 20% from 1975 to 1985. Canon was well-equipped to undertake difficult product development projects as it had accumulated a large and diverse technical staff of over 3,000 engineers, consisting of mechanical engineers (30%), electronics engineers (30%), physicists (17%), chemists (10%), and computer-related and other fields (13%). The electronics engineers had been hired in earlier relatively unsuccessful efforts to enter the synchro-reader and electronic calculator fields. The accumulation and interaction of personnel with these diverse technical capabilities provided an environment that induced the creative tension which is necessary to lead to synthesis and new information creation10. To further 7 The bulk of this section is based on intensive interviews by Ikujiro Nonaka with Canon Inc. personnel. Some of the information is available in the Canon Handbook (1987). For another discussion of the Canon case study, see Nonaka and Yamanouchi (1989). 8 An evidence of its successful technology development is the fact that in l988 Canon fiied more patents in thc U.S. than any other company. Japanese or American. 9 On the seriousness of such a move, see Aoki (1989 ). 10 Canon does not separate technical personnel into departments according to discipline so there is much greater mixing of research personnel (for the importance of this method of organizing the innovation process, see Florida and Kenney (1990). This heightens the possibility of new concepts and ideas emerging serendipitously. For further discussion of Japanese strength in hybrid technologies, see, for example, Kodama (1988). 71 accelerate information creation Canon has had a policy of hiring mid-career personnel from other firms to create “counter-cultures” or diversity within Canon to increase the potential for new information creation. In Fall 1970 Canon began internal development of a plain paper copier (PPC ) technology. In the early days divers)fication into PPCs was opposed by many in the company, some going so far as to advocate discontinuation of the entire effort, recommending instead that Canon concentrate on its camera industry. But