Organizing for Innovation in the 21st Century Deborah Dougherty Professor, Management and Global Business Dept, Rutgers Business School Rutgers, The State University of New Jersey doughert@rbsmail.rutgers.edu September 2004 DRAFT Product innovation is a primary means to adapt to changing markets, technologies, and competition. Innovative organizations are more profitable, grow faster, create more jobs, and are more productive than their non-innovative competitors, even in mature industries (Capon, Farley, Lehmann & Hulbert 1992; Baldwin & Da Pont 1993; Christensen & Raynor 2003). The ability to generate streams of new products or services over time is therefore vital to many organizations. Organizational design plays a significant role in this ability, so understanding how to organize for innovation is a central problem in innovation management (Galbraith 1995; Tushman and O’Reilly 1997; Dougherty 2001). The most recent Product Development Management Association survey finds that organizing is the “last frontier” in innovativeness (Adams 2004). Previous PDMA surveys found that innovative companies used more “best practices” (1990), especially strategic systems (1995). The 2003 survey found that the most innovative organizations used practices selectively, molded the ones they did use to fit particular situations, and implemented them more effectively. Moreover, innovative organizations managed the entire organization to support innovation by ensuring that resources flowed smoothly to innovation teams, that structures, processes, and other organizational mechanisms supported innovation, and that long-term investments in supporting technologies were made. The purpose of this paper is to develop four principles of organizational design that synthesize current ideas into a new understanding of organizing for innovation. I focus on the principles for defining jobs, grouping jobs into work units (aka differentiation), integrating the 1 differentiated units, and controlling the whole system over time. These are the principles that are often discussed in textbooks (e.g., Bartol and Martin 2000; Dessler 2003), and by theorists (Blau and Meyer 1987; Mintzberg 1991; Galbraith 1995). The argument developed in this paper is that the essence of organizing for innovation depends on how these four principles are understood and enacted, not on how more manifest structures are deployed (e.g., centralization vs. decentralization of R&D, organizing businesses on products versus markets, use of portfolio strategies and other processes and techniques). These four principles constitute the core structuring upon which systems, processes, procedures, routines, and techniques are built. I propose new approaches for each principle that enable sustained product innovation in general, and suggest contingencies that may require different configurations among the principles. These new principles of innovative organizational design are both empirically and theoretically grounded. Regarding the first, in all my fieldwork on innovation (more than 350 interviews since 1985 with people working on new products or services in 33 different organizations), the organization’s design always had a strong presence – it was a significant part of people’s stories. Even in the ‘early days’ when product innovation was fairly rare in established firms, most people who worked on new products understood that it was necessary to involve different functions, to work with customers, and that good teamwork and project management mattered. Most people ‘got it’ about basic new product development, but were prevented from following these practices by how their company was organized (Dougherty 1990; Dougherty and Heller 1994; Dougherty and Hardy 1996). The organization of work in many of these firms straightforwardly inhibited teamwork, knowledge creation and recombination, and creativity. Many of the people interviewed were very frustrated. But much of the scholarly and managerial attention was focused at the individual level – sending people off to creativity 2 training or encouraging renegade “champions,” or at the project level – facilitating gatekeepers, team meetings, or fixing the “market-technology” interface locally. My insight that ‘it’s the organization, stupid!’ was also grounded in theory. Indeed, the importance of organization design to a firm’s ability to innovate is hardly new, but it is continually rediscovered. Burns and Stalker (1961), Schon (1963; 1967), Kanter (1983), Nord and Tucker (1987), Jelinek and Schoonhoven (1990), and Brown and Eisenhardt (1997) discovered essentially the same thing: people working on successful innovation know what their jobs are, who they work with and how, to whom they report, what the priorities are, and how their activities fit with those of the enterprise as a whole. In other words, the work of innovation is organized, as any complex set of activities would have to be. All these studies also rediscover the fact that organizing for innovation is qualitatively different from organizing for efficiency via the familiar bureaucracy. However, many describe the innovative design in terms of its deviance from the bureaucratic model: as something that occurs in a parallel place (Kanter 1983), or as “quasi-formal” (Jelinek and Schoonhoven 1990), “semi-structured” (Brown and Eisenhardt 1997), or not bureaucracy (e.g., de-centralized, in-formal, de-layered). To develop my argument, I build on four basic premises. The first premise is that the innovative organization needs to be defined in its own terms, not in terms of its deviation from the 20th century bureaucratic model. The second premise is that to do the first premise, theory must be based on the actual practices that are being organized. While many theories break away from the bureaucratic model, they are all over the map because “anything” is being organized. Alternate models include the post modern, post bureaucratic, heterarchical (Hedlund 1986), boundaryless (Devanna and Tichy 1990), hyptertext (Nonaka and Takeichi 1995), ambidextrous (Tushman and O’Reilly 1997), horizontal (Ofstroff 1999), “edge-of-chaos” (Brown and 3 Eisenhardt 1998), transnational (Bartlett and Ghoshal 1989), “front-back” (Galbraith 1995), shamrock (Handy 1989), and adhocracy (Mintzberg and McHugh 1985), to mention but a few. To crystallize insight from all these ideas, I focus only on sustained product or service innovation. “Product innovation” concerns bringing new products and services into use, and encompasses the whole process of conceptualizing, developing, designing, manufacturing, and distributing new products (Schon 1967). Sustained innovation refers to continually generating streams of new products over time. This focus limits the generality of the implications, but affords a thorough treatment of the particular activities underlying this kind of innovation. The third premise is that organizing for innovation can be the ‘default option’ for many organizations, not the bureaucracy, because high volume and low cost are subsets of innovation management that also includes radical, breakthrough, platform, and derivative products. Effectively innovative organizations like Intel, Sony, or 3M all produce products with high volumes at low cost as well as new and breakthrough products. They are capable of these economies of scope because they organize around innovativeness, not around efficiency. Being innovative is not incommensurate with being efficient, although limits can be explored empirically. I argue in the next section that organizing around efficiency with the bureaucratic model is, however, incommensurate with innovativeness. I show that the bureaucracy is not a monolithic institution of modern life, but a particular kind of organizing with particular principles of design that happen to preclude innovation. Innovation is not more costly, more risky, or more difficult to do per se, but it is all three if attempted from within the bureaucracy. The final premise is that the underlying principles of organization design helps to put the “T” and the “O” back into OT (organization theory). Organization design was the “meat and potatoes” of organization theory at the outset of this field (Tushman 2004; Nystrom and Starbuck 4 1981). While still central for managers, organization design but has fallen off the OT table for two reasons. First, organization design has been given over to consultants in part because “design” is understood in a-theoretical terms, as the use of tools and techniques that are based more on fads than on scholarship (e.g., process re-engineering, total quality, stage-gate). Bringing the “T” back into organization design will build in a theoretical foundation for when, why, and how to use these tools. Second, OT has shifted to inter-organizational issues such as the causes and effects of institutional pressures, industry emergence and evolution, or networks. Inter-organizational issues are important, since value creation and even specific innovations are conceived, developed, and launched by a community of organizations, not simply by “the firm.” But the “O” of OT is either a blackbox, or a series of poorly defined “umbrella concepts” such as routines, dynamic capabilities, organizational capacities, cognitions, culture, slack, effectiveness, and identity (Hirsch and Levin 1999). These ideas are used so broadly that each often encompasses the others. Rethinking the essence of organization design relieves these other constructs of the burden of explaining work roles and relationships, and pushes for more thoughtful conceptualization of them. This chapter has three main sections. The first section lays the foundation for new principles of organizing by explaining four paradoxical practices that comprise organization-wide product innovation. These practices reflect the inherent tensions of innovation by capturing both the activities of product innovation and the challenges of organizing them. To justify the need for new principles of organizing, I explain why the bureaucratic principles for job definition, differentiation, integration, and control cannot handle the inherent tensions of product innovation, and thus cannot organize these activities. In the second section, I review current theories of organizing for innovation and derive from them four new principles of organizing that 5 do effectively embrace these tensions, and thus are able to organize innovation. I also suggest how various best practices for innovation and for organizing fit with each new principle. In the final section, I clarify the implications of this new theory of organizing for innovation, and suggest ways to test, refute, or elaborate it. Section I: What Is to Be Organized and Why Bureaucratic Principles Cannot Work It is first necessary to define the practices of innovation that are to be organized. There are many processes, techniques, and procedures for managing innovation teams and projects, building and shaping technology and manufacturing for innovation, assessing user needs and strategic opportunities – indeed the PDMA study mentioned in the first paragraph asked respondents about their use of more than 70 specific techniques. I focus on four higher level practices that cut across functions, levels, and locations but still represent particular challenges of sustained product innovation, as outlined in Table 1 (second column). These are presented as “paradoxical practices” to reflect the fact that innovation embodies inherent, complex tensions that must be balanced in some way rather than ignored or split up. Innovation researchers have always pointed out that product innovation involves true tensions (Pelz and Andrews 1966; Lawrence and Lorsch 1967; Jelinek and Schoonhoven 1990; Weick and Westley 1996). While other tensions are also involved, research suggests that these four relate to innovation success (Kohler 2003). As argued in Dougherty (1996), thinking of the organizing challenges of innovation as inherent tensions emphasizes the dialectical nature of ongoing innovation, where the organization and new products mutually constitute each other. The first column in Table 1 connects each paradoxical practice with a particular principle of design. There is no strict one-to-one correspondence between practices and principles, but I organize the discussion row by row in Table 1 for ease of exposition. The third and fourth 6 columns explain that the familiar bureaucratic version of these principles cannot organize the paradoxical practices of innovation, which is why a new set of principles is necessary. Enabling Motivation and Commitment to Balance Freedom Vs. Responsibility The first paradoxical practice is about motivating employees to engage in, and be committed to accomplishing, the very complicated, confusing, and ambiguous work of sustained innovation. People must work in multi-functional teams because people from different specialties have unique insights (Dougherty 1992), but orchestrating teamwork is always difficult. In addition, the teams inevitably encounter unanticipated problems as they design and develop a new product, which means that they often must rethink the very problem they are solving – what will the product do for whom and how – as they proceed. They continually “set” their problem, defined by Schon (1983) as the artful competence of finding the problem by defining the decisions to be made, the ends to be achieved, and the means that may be chosen. Innovators also actively solve that problem, but must proceed in parallel, with horizontal flows of activities that continually interrelate: people push forward issues in their own area (e.g., prioritizing customer needs, planning advertising for the launch, developing the production process) in sync with others’ activities, but they also must shift their attention quickly to problems that affect all functions (e.g., a part cannot be made reliably after all, which affects the price, or the launch time, or the target market –Yang and Dougherty 1993). To work collectively in parallel on such ambiguous tasks, innovators must anticipate problems in other functions and appreciate others’ constraints as they work (Clark and Fujimoto 1991). They need “T”-shaped skills, or both a deep understanding of a specialty and an intimate understanding of the potential systemic impacts of their specialty (Iansiti 1993). People must shape their specialized knowledge to fit the problem at hand rather than insist that the problem 7 appear in a certain way (Leonard 1998). These are very sophisticated skills that can be enacted only if others also enact them, which means that developing these skills is not simply a matter of hiring the right people! Motivating employees to do the work of innovation invokes the tension between freedom and responsibility. People are expected to be free to do what is necessary, and to be creative and proactive, yet also to be responsible about the use of firm resources, and to coordinate their actions with others carefully. People must rely on their peers to achieve what they personally are held accountable for, yet they have little control over their peers. The design principle most associated with the tension between freedom and responsibility is defining “jobs” in the first place, which clarifies people’s roles and responsibilities. The question is, how to construct sensible but motivating jobs and also invoke responsible, principled action? Bureaucratic rules produce autonomous, precisely defined jobs (Weber 1946; Blau and Meyer 1987). One rule is that “the office” comprises clearly defined set of solutions that “the officer” applies. Problem setting occurs only occasionally and only by officers, and work is split into autonomous specialties. The second rule is hierarchy: the officer has control over what people reporting to him or her do and how they do it. Managers think and employees do as they are told, which focuses attention on vertical information processing rather than on parallel or horizontal flows of activities. The third rule is individual accountability, which means that jobs must be precisely defined ahead of time. Work in a bureaucracy is understood as static slices of specialized labor, whose pre-defined activities can be executed without ambiguity (Schon 1983; Brown & Duguid 1991; Barley 1996). Bureaucratic jobs eliminate the tension between freedom and responsibility by specifying responsibility and removing freedom from all below the officers. 8 The challenge is to re-conceive work at all levels, so that people can take on a fuzzier set of responsibilities without feeling overwhelmed, and that managers can “manage” the work without being “officers” and thus micromanaging, or making all the choices and decisions. Redefining work touches on many complex issues in human relations (e.g., justice, creativity, identity), but this challenge can be sorted out in part by the next three design principles. Creative Problem Solving to Balance Old Vs. New The second paradoxical practice is creative problem solving. Conceptualizing, designing, developing, manufacturing, and launching a new product requires that innovators solve a variety of complex problems, many of which arise unexpectedly as new alternatives are tried out or as unfamiliar customers are addressed. A large organization will have scores of new product teams creatively solving unique problems while drawing on common R&D, manufacturing, distribution, or selling capabilities, for example, so creative problem solving goes on across the organization, at all levels. Functional managers oversee how well their people contribute to the new product teams and how carefully the teams take functional constraints into account (e.g., limited plant capacity); business managers incorporate new products in their business models and portfolio plans; and corporate managers make the necessary investments in capabilities to support all the teams, functions, and businesses. Together, these people combine resources in new ways, fit new products into existing families and businesses, and transform businesses to adapt to shifting market opportunities (Helfat and Raubinichek 2000). All this creative problem-solving invokes the tension between old and new, for three reasons. First, organizations cannot abandon existing assets as they create new products. Even high technology firms must leverage the significant investments in production facilities and the knowledge that is embedded in them (Jelinek and Schoonhoven 1990). A very new firm might 9 use only new assets for their first new product, but their second product necessarily invokes the tension of old versus new, since the two products now must share resources. Second, path dependencies exist since knowledge builds up over time based on choices that are made and on opportunities that are selected – history matters (Dosi 1982). Third, there is a large empirical record of rigidity, local search, and becoming stuck in old ways even in high technology firms (Kanter 1983; Henderson and Clark 1990). Any theory must explain how organizations can overcome pressures to favor old over new. The organizational design principle most associated with the creating problem solving practice is differentiation, or the grouping of work and workers around related activities. Differentiation is essential in complex organizations because it simplifies work in three key ways. First, differentiation helps to define what people are to do on a regular basis. People can come to the “office,” the “lab,” or the sales route every day with a good idea already in their minds about what to do with whom and how. Without differentiation, people would have to be assigned to specific jobs everyday, like day laborers. Second, the specialization inherent in differentiation enables continued learning and development since what to know is bounded into a coherent system. Third, differentiation enables self-designed work. As Galbraith (1995) suggests, if work is differentiated appropriately, people will interact, communicate, and make sense of everyday work fairly automatically and naturally with others in their unit, so work is self-designing within those boundaries. The bureaucracy differentiates work by functional specialty that decomposes the value creation process into separate parts, or steps. Differentiation by function cannot organize the practice of creative problem solving because it precludes the collective definition of problems in the first place. The problem that people are working on is already set or defined, so there is no 10 “new” to weave together with “old.” Only people at the very top understand the problem in its entirety. As well, functions are solutions, not problems, so organizing by function further inhibits problem setting, which precludes creative problem solving. People can create new solutions within their functions, but that increases conflicts because there is no way to assure that one function’s new ideas fit with another’s. Expanding bureaucratic differentiation to include product line, geographical location, national locality, market place, and business increases conflict further, since many more lines of authority are involved in each solution, and all these lines have to be checked and perhaps rewired every time something changes. A fundamentally new principle for differentiating work is necessary for innovation. The question is, what are those related activities that would provide the simplicity and familiarity, yet also enable people to creatively solve all the problems of innovation? What core sets of related activities can connect so many old ways, processes, and so forth with so many new ones? Market-Technology Linking to Balance Inside with Outside The third paradoxical practice of innovation is linking technologies with markets, which invokes the tension of inside versus outside in the broad sense, because usually we think of technology as inside and the market as outside. A new product is a package of features and benefits, each of which must be articulated and designed. To create value, this package must address core customer needs and wants in a way that leverages the organization’s unique technology capability (Burgelman 1984). Market-technology linking connects insides and outsides across multiple boundaries. First, market-technology linking is multi-functional because all functions have unique insights for a given product design (Dougherty 1992). Engineers, marketers, and manufacturing all must reach beyond their boundaries to absorb others’ insights, while also making their own specialized insights sensible to others. Second, products rarely 11 stand alone. The necessary integrity of each product (cf Clark and Fujimoto 1991) must be balanced with how well it fits into product families and platforms. And the integrity of product families and platforms must be balanced with firm-wide resources that must be collectively leveraged, and perhaps with a common brand or image that presents a common face to “the marketplace.” Third, inside functions and businesses are linked outside in a variety of alliances, partnerships, and associations (R&D, distribution, supply chain). In most industries value is created and delivered by a network of players who provide complementary assets, components, basic R&D, and so on (Chesbrough 2003; Miller and Floricel 2003). The design principle most associated with the tension of inside versus outside is integration, defined by Lawrence and Lorsch (1967:11) as “… the quality of the state of collaboration that exists among departments that are required to achieve unity of effort by the demands of the environment.” This “state” of collaboration is far more complex in 21st century innovative organizations, since the organization is not only integrating functional departments, but also product families, platforms, opportunities, alliance partners, newly acquired units, geographical foci – many more “insides” need to integrate with many “outsides.” While there are numerous specific techniques for integration (e.g., standards, direct supervision, liaison people, common values), there are also two general aspects: mutual adjustment (Thompson 1967; Mintzberg 1991) and mutual adaptation (Leonard 1998). Mutual adjustment refers to the continuous adjustment of actions by one specialist to the actions of others on a team, as people literally figure out what to do and how to do it as they do the work. To work in parallel, people must interrelate heedfully – carefully, critically, willfully, and purposefully rather than habitually. According to Weick and Roberts (1993) if interrelations are heedful, people are more mindful of the big picture to which they contribute, can comprehend 12 unexpected events, and can share ideas and experiences more quickly. To interrelate heedfully, people must envisage a system of joint action, construct their own contribution to fit that collective representation, and interrelate their contribution to the system that is envisaged. Dougherty and Takacs (2004) argue that a very new kind of boundary is necessary if scores of new product teams that exist at any one time in a large organization are to work interdependently, or heedfully. Rather than the imaginary line that signifies a limit, they propose a new metaphor based on “team play.” People involved in team play easily engage in heedful yet improvised action, provided there is a clear game with a few simple rules. They propose that “team play” is a metaphor for ongoing mutual adjustment by heedful interrelating, because play enables ‘adaptive variability’ (Sutton-Smith 2001), other oriented roles, and a sharp focus on the means, or the processes and manner in which activities are carried out. Team play, however, requires a very different strategic context to enable multiple teams to work together at once: senior managers must actively define the business, customers, and goals to create a “game” in the first place, and they must develop capabilities and resources to support all the team play. People cannot readily join projects over time, and scores of teams cannot play at once, without resources, facilities, shared rules, and of course a clear game (albeit one that emerges over time). Mutual adaptation refers to the reinvention of one system such as manufacturing to conform to a new strategy, and the simultaneous reinvention of the organization to conform to the new system. Mutual adaptation involves moving back “up” the design hierarchy (Clark 1985), to rethink prior choices among alternatives, because new technologies now make previously rejected options feasible, or changes in customer needs now highlight different kinds of performance. Mutual adaptation thus involves revisiting prior decision points, reopening issues that had been resolved, and unfreezing organizational routines (Leonard 1998). 13 Bureaucratic integration is based on pre-set standards that standardize action rather than facilitate the interweaving of improvised plays. Bureaucratic integration therefore presumes preprogrammed coordination (sequential and pooled rather than reciprocal), and focuses primarily on the inside (efficiency). Bureaucracies accommodate mutual adjustment as a very costly addon (e.g., in a venture unit or special department), not because mutual adjustment is inherently costly but because it does not fit. Thompson (1967), for example, argued that mutual adjustment should be used sparingly, and only in task forces that are bracketed off from the rest of the organization. Bureaucratic organizing cannot accommodate mutual adaptation at all, since managerial attention is focused on small cycles of change and cannot detect when large spirals of change are necessary. Organizing for innovation requires a fundamentally new principle for integration that enables mutual adjustment and mutual adaptation across so many insides and outsides, or so many market-technology linkages. Monitoring and Evaluation to Balance Emergence Vs. Determination The fourth paradoxical practice is monitoring the work, which invokes the tension of emergence versus determination. Innovations always involve surprises since customers, needs, and technologies are often new and untried, and actually emerge through trial and error. However, innovation also makes it easy to squander huge amounts of resources, so resources must be channeled, processes must be monitored, and activities occurring at different levels must be kept in sync somehow. The fact that many new products, even relatively incremental efforts, still fail is testimony to the need to monitor and evaluate the activity. The design principle most associated with the tension of emergence versus determination is control, or assuring that the right activities are carried out at the right time with the right resources. Indeed, many “best 14 practices” for innovation are formal approaches for establishing priorities, milestones, and objectives, which highlights the utter necessity of control in innovation work (Beck et al., 1994). Controlling innovation involves two thrusts: focusing on the processes or activities through which people accomplish their work to keep means and ends connected, and enabling situated judgement (Dougherty 1996) on the part of innovators. Each complements the other, so both are necessary. First, the ends or outcomes of innovation take a long time to be achieved and may change in any case, so innovation cannot be controlled by measuring achievement of objectives alone, especially not in the short term. Managers must also control the processes of innovation by monitoring such things as cycle time, yields, percentage of usable components, platform efficiency, share of industry value, and market shares and changes. These process measures help to both surface problems and prevent capability traps that can arise if people shift attention from process improvements to defects. Jelinek and Schoonhoven (1990) describe weekly operational reviews that are focused on providing individuals with useful information about their status, burn rates, and so on. Clark and Fujimoto (1991) outline an alternate paradigm for prototyping and plant ramp-up that helps to surface potential design and manufacturing problems rather than to show proof of concept. The purpose of the process controls is to enable people to surface (or set) and to solve problems, and to revise processes themselves as necessary. Second, innovation is inherently unpredictable so the organization must rely on people to make the right call as they monitor emergent flows of activities. The control system must enable both situated judgement, and the continued situating of the process controls outline above, to keep them grounded in everyday reality, and continually refreshed. Benner (2003) refers to a similar term, the “clinical judgment” that practitioners such as nurses, teachers, architects, 15 technicians, and other knowledge workers must rely on to do their work. As Dougherty (1996) explained, “situated” refers to being engaged in the details of the innovation and its emergence, because these are complex, often tacit issues that must be “visceralized” to be understood, to use a phrase developed by Schon (1985). “Judgement” refers to the capability to use insights and heuristics developed from experience, and to “appreciation,” which Vickers (1965) uses to refer to sizing up unstructured situations and judging the significance of various facts – much like the reflection-in-action that Schon (1983) argues is central to work as a professional. Situated judgement implies occupational control (Barley 1996), which is based on the authority of expertise rather than of the position. The bureaucratic principle for control specifically excludes processes, situatedness, and judgement (except by officers), and therefore cannot organize innovation. First, bureaucratic control rests on premises of rationality that separate means from ends (Schon 1983; Benner 2003) and concentrates on the accomplishment of ends. Processes are monitored to detect deviance or variance from prescribed standards, not to surface problems with the process. Once they have been scientifically measured, processes are given. Standards are generalized to fit across contexts, so situational issues are noise, not important data. And judgement by workers is illegitimate: knowledge workers must use judgement to do their work but that fact is not recognized by the managerial system (Brown and Duguid 1991, Barley 1996, Benner 2004). In the next section, I develop an alternate approach for each of these design principles that can embody the tensions of innovation, by contrasting some current thinking on these aspects of innovation. Table 2 outlines the overall argument, and table 3 provides details for each section. 16 Section II: New Principles for Organizing the Practices of Innovation Commitment, freedom and responsibility The first challenge outlined above is to develop a new principle for defining work, so that jobs encompass: 1) problem setting and problem solving rather than the application of predetermined solutions, 2) working in parallel with other experts rather than apart from them, and 3) being mindful of the systemic impact of one’s activities rather than ignoring it. Innovation researchers do not directly address the principle of defining work, but they do propose that variety of special roles be added to the organization to enable innovation. For example, the “gatekeeper” (Allen 1977) connects a research team with external sources of knowledge and filters out unnecessary noise; “ambassadors” obtain resources and legitimacy, “scouts” and “coordinators” manage dependence on other functions, and “guards” control outward information flow (Ancona and Caldwell 1990); “hunters” and “gatherers” search for new opportunities (Delbecq 1989); the “product champion” forces new activities through the organization (Schon 1963; Burgelman 1984); senior sponsors enable champions (Maidique 1980; Day 1994); and the “heavy weight” project leader possesses the necessary clout to complete a project successfully (Clark and Fujimoto 1991). See Dougherty (1992, 2001) for exceptions. Knowledge work researchers, however, do emphasize an alternate understanding of work, that of practice (Schon 1983; Brown and Duguid 1991; Barley 1996). Since creating new products is about creating, combining, and recombining knowledge about markets and technologies, I propose that everyone’s jobs, from front line workers to senior managers, be based on the principle of work as practice. All jobs are understood as a professional practice or occupation, which involves hands-on, active participation in innovating. Defining work as practice means that jobs embody the means and the ends of work, the practical wisdom people 17 rely on, and the “… rich, socially embedded clinical know-how that encompasses perceptual skills, transitional understandings across time, and understanding of the particular in relation to the general” (Benner 2003: 5). “Practice” highlights the spontaneous, inventive, and improvisational ways through which people get things done (Schon 1983; Brown and Duguid 2001). According to Brown and Duguid (2001), practice concerns “knowing how,” defined as the ability to put know-what into practice. Work as practice is based on two very different rules than is work as a bureaucratic functionary. First, instead of “the office,” the rule for practice is that each person takes responsibility for the whole task and for one’s contribution to that task. In professional communities like medicine, equipment repair, and technicians (Brown and Duguid 1991; Barley 1996), people understand themselves to be responsible for the practice and for contributing their own expertise effectively to it, which helps them understand their work in terms of its contribution to the larger whole. Second, instead of the rule of top-down authority where senior people think and lower levels do, practice requires that everyone be a knowledgeable expert. The rule for work as practice, therefore, is that knowledge is central, and that people actively create new knowledge as they practice innovation. Dougherty, Bernard, and Dunne (2004) suggest that as part of the rule that knowledge is central, people are skilled at articulating their own knowledge in a way that is sensible to people with different expertise, which keeps all the knowledge flowing. While it is different than being a cog in a big machine, working as a professional practitioner is not entirely unfamiliar. For example, scholars have always emphasized the importance of extensive hands-on involvement on the part of senior managers in innovation (Quinn 1980; Jelinek and Schoonhoven 1990). Maidique and Hayes (1984) argue that senior 18 management involvement in innovation provides them with an in-depth appreciation for how the technology works and how their organization works, so they can better understand strategic possibilities. Hands-on involvement like this reflects work as practice. Scientists, engineers, market researchers, technicians, and even shop floor workers work as practitioners rather than as bureaucrats. However, I suspect that the design principle of work as practice cannot be implemented without transforming the other three principles of organizational design as well, since differentiation helps to define work, integration lays out pathways for collective action, and control keeps attention focused on the practices and channels resources to the right places. Differentiation by Communities of Practice Focused on Innovation Problems The second organizing challenge is how to differentiate the work of innovation in a way that makes the necessary creative problem solving sensible, doable, and reasonably simple. Innovation researchers focus considerable attention on defining work boundaries, and on differentiating activities into ambidextrous units that focus on different kinds of innovation (e.g., era of ferment, incremental, discontinuous – see Tushman and O’Reilly 1997), or into various product families (Brown and Eisenhardt 1998) and shifting charters among units (Helfat and Rabinichek 2000), or setting up special “hubs” for radical innovation (Liefer et al 2003). These theorists recognize the need to balance the tension between old and new in creative problem solving for innovation, acknowledge (at least implicitly) the centrality of differentiation in this balancing, and emphasize horizontal organizing around activities that flow over time. However, the theorists differ over whether or not “old” practices, markets, technologies, and perspectives dominate new ones, whether that imbalance is inherent in organizing, and how the balancing is done. This difference arises from the kinds of industries and scopes of activities that the theorists focus on: the corporate wide strategic and business challenges of innovation in 19 established organizations that become stuck in existing markets, versus the R&D and project challenges of innovation in high technology firms that can emerge effectively over time. I combine both perspectives to suggest that all four sets of innovation challenges – working on the strategic, business, R&D and other capabilities, and project-level problems of innovation – are the core sets of related activities around which work is differentiated for all innovative organizations (Dougherty 1996; 2001). I propose that the innovation practitioners be differentiated into four communities of practice that focus on setting and solving each of these kinds of innovation problems. A community is not a tightly knit group, but an identifiable, recognizable group of people who do that set of work. They know who they are and others know who they are, which means that none of the problems are addressed on an ad hoc basis. Each problem area is a horizontal slice of activities that unfold in time, thus embracing old and new as they flow together. Each of these problems is also a simple set of related activities that can be separated from the others, provided the other problems are being addressed competently. Work is self-designing within each community of practice, where people set and solve specific problems autonomously within strategic guidelines, choose among alternatives, access resources, and apply their expertise as they see fit. For example, new product projects and R&D capabilities communities both work on technology. But the latter develops basic physics for a certain use of optics, for example, while the former builds a piece of equipment that applies those basic principles of physics in a way that can be produced at a certain cost. The business unit matches bundles of products to specific markets to make a profit, while the strategic managers look for new opportunities. These are very different problems that should be managed in their own way. This principle of differentiation makes these four activities the most central ones for the innovative organization. 20 Tushman and O’Reilly (1997) and Christensen (1997) emphasize the strategic and business problems of innovation. The strategic problem involves looking where the marketplace might be going and for emerging customers and technologies, articulating the corporate direction for value creation and making it sensible to people, and making long-term investments in competencies. The business problem involves making a profit by bundling organizational resources to fit with particular market, technology, and competitive trends over the life cycle of a business, and tracking changes in marketplaces. When organizations fail to address these two problems effectively, they become trapped by “the tyranny of success,” according to Tushman and O’Reilly (1997) and fixate on current products or customers (Christensen 1997). Units tend to develop a tight configuration of structure, strategy, culture that limits activities. As well, business units cannot evolve on their own from one configuration to another, because new technologies are discontinuous. A big gap exists between one technology and the next, like a step function, and requires a big “leap” to cross it. These authors argue that senior managers must make this leap by creating new business units and differentiating them from old ones. Other theorists focus on projects and R&D management, and tend to emphasize continued incorporation of new activities. Leonard (1998), for example, suggests that projects drive daily business as teams dissolve and regroup according to the dictates of the business needs. At the project level, people define product concepts within strategic guidelines, determine customer needs in priority order, and jointly solve problems in the design, manufacture, and marketing of the new product. All these projects give the organization considerable fluidity, because managers can shift business charters from one unit to another, and people can move along product lines (Eisenhardt and Galtung 1996, Helfat and Raubinichek 2000, Helfat and Eisenhardt 2001). As well, these theorists argue that technology discontinuities are not so 21 drastic when managers concentrate on the activities that create the underlying technical knowledge and make small checks on position, direction, and proximity to competitors (Leonard 1998). If technology, manufacturing, IT, and marketing are managed as capabilities, then practitioners can develop deep expertise in these knowledge bases for long-term value creation, and actively keep these capabilities connected to projects and businesses (Dougherty 2001). Differentiating the practice of innovation into four communities of practice organized around distinct problems of innovation management captures these alternate views. Table 3 outlines each community’s practice. This principle of differentiation makes each of the four problems in innovation management salient, and centers the organizing on these core processes of work – highlighting the practices of doing rather than only outcomes. My theory is that all four communities of practice are necessary in all organizations, because it is only by actively working on these very different problems do people maintain and enact the requisite hands-on, situated knowledge of each of these issues. However, the horizontal communities and their flows of action must now be integrated, since each is dependent on the other. I turn next to a new principle for integration that would allow people to work with a common understanding of the same problem, so they can readily interact, make sense of unexpected issues, and absorb market and technology knowledge across the organization. Integration by Strategic Articulation The third principle of organizing for innovation must integrate the differentiated communities of practice to link up markets and technologies and to balance the tension between inside versus outside. The horizontal communities of practice facilitate market-technology linking within each, since each encompasses users issues with technological ones. As well, each community is a network that extends beyond the organization’s boundaries. For example, 22 managing the R&D capability includes participation on standards boards and various partnerships with others to develop technology, while managing the business embraces supply chains and customer partnerships. Mutual adjustment within each community is important but can be managed straightforwardly with various techniques (committees, task forces, product family and portfolio management, regular meetings). The major strategic challenge for integration that remains is mutual adaptation among the four horizontal communities of practice, to keep them in synchronization. Innovation theorists agree that the organization’s strategy is the primary source of integration, but disagree over how tightly integrated the insides are with their outsides, and whether strategy is mostly imposed or mostly enacted. Tushman and O’Reilly (1997) highlight limited strategic choices for businesses, arguing that each business must take a particular position along the technology trajectory, and that senior managers keep the organization connected with its market by making experiments and selecting among them. Strategic meaning is largely imposed. Others argue that strategy will not be meaningful unless innovators continuously enact it, so managers cannot impose the strategy. Brown and Eisenhardt (1997) suggest that innovative organizations operate at the edge of chaos, and continuously balance present and future (and inside with outside) with such techniques as patching (mapping modular activities onto marketplaces), probing (experimental products, future oriented alliances), and pacing (setting rhythms for change). Leonard (1998) argues that a focus on technology draws attention outward to connections with universities, alliance partners, sourcing mechanisms such as licensing, and acquisitions, which keeps people attuned to the outside rather than primarily inward. The more fluid strategies are not imposed but rather directed and shaped by managers. 23 Organizational sensemaking captures the insights of both these alternate views. I propose that the communities of practice are integrated by ongoing sensemaking of the strategy across communities of practice. Organizational sensemaking balances generically sensible understandings, such as an articulated strategy, with intersubjective understandings that are enacted in the situation (Weick 1995). Sensemaking emerges from continuous processes of renegotiating and reconciling understandings. Integration by strategic sensemaking juxtaposes the articulated strategy of one community of practice with the unarticulated and thus intersubjective strategic insights of another, which refreshes the meaning of the articulated strategy by drawing out insights from the emerging insights. To understand how this sensemaking works, first, research shows that that each of the four communities requires a strategic framework to do its work: a strategic path, a competitive strategy, a technology strategy, and a product concept (Khurana and Rosenthal 1997; Wheelwright and Clark 1991; Cooper 1998). These strategies need to be articulated, which refers to lifting them out of the tacit, private, complex, activity to make them explicit, simpler, ordered and relevant to the situation (Obsfeld 2004). Articulation is vital since without it complex frames like innovation strategies tend to fall into the background and become taken for granted (Suchman 1996). Moreover, the articulated strategy integrates markets and technologies for its level, and connects a different scope of inside with outside. As outlined in Table 3, the strategic community integrates the various businesses with capabilities into a strategic path that defines what the company as a whole will do for what kinds of customers. The business community integrates products with business opportunities and with the life cycle of a business as it emerges over time to generate profit. The capabilities communities integrate businesses and projects into functionalities and performance levels that 24 the organization can provide. And the product development community integrates specific technology designs with specific customer needs into product concepts. In any kind of complex work, however, the press of continual challenges draws attention to the minutia of everyday life, so these articulated strategies must be continually articulated and re-articulated to keep them salient, and to refresh and update them as new experiences become relevant. And the articulation of the strategy in one community enables the re-articulation of strategies in the others. For example, in the product development community, teams create their product concepts by drawing on strategic thrusts and available resources from businesses and capabilities. The product concept articulates a particular value to be produced for customers, and provides a blue print for how the team will create that value. However, the collection of product concepts developed by the teams across the organization embodies new insights into market and technology trends, and can inform the ongoing directions of businesses and capabilities. As well, people also continually enact the strategy as they do their work, so the strategy takes on new meanings. People make sense of new or tacit knowledge (such as “sticky” customer needs, new strategic paths) by interacting face-to-face to communicate interpretations of these, to sort out possible attributes of them, and to explore different takes on them. The organization can build up new manufacturing and technology capabilities based on what works and does not work. The theory suggests that each community of practice needs strategic sensemaking, and that the articulation of each depends on interactions with activities in other communities. A business strategy will become stuck if there are no product trials to explore options and no technological alternatives to think about. Technology development will become limited or esoteric if the businesses become static. The theory also says that the firm’s strategy has to be clear and vivid, which may limit in the scope of innovation activities that any organization can undertake. 25 Part of integration is acceptance of the apparent looseness within and among the four domains of innovative action. Integration is enacted in the practice. The looseness is not a loosening of standards, but rather a recognition that a common space of work, based on a deeper understanding of shared appreciation of the joint activity that people are collectively engaged in, is necessary. In the bureaucratic model, such looseness implies out of control, but now, looseness is necessary for survival, and so is necessary for control. It is thus necessary to redefine control as well, to focus on controlling the terrain of the looseness. Control by Generative Rules and Resources Finally, I argued above that the principle of control for the work of innovation must combine a focus on the processes of value creation with situated judgement. Most theorists agree that the kind of control that is needed for innovation is cultural, not mechanical, because people must continually make sense of what they need to do and how they need to do it when they do innovation work. However, theorists differ on how culture affects everyday actions – is culture a set of meanings, or a medium for enacting meanings? Some suggest that culture itself conveys meanings that people must share, and that senior managers impose these meanings by promoting particular norms and values. According to Tushman and O’Reilly (1997:102): “Culture is a system of shared values and norms that define appropriate attitudes and behaviors for its members.” A problem with this perspective is that the list of possible norms and values can be quite large (e.g., change-oriented, focused on customers, entrepreneurial, risk-taking, valuing failure, proactive). Others draw on complexity theory at least metaphorically, and propose a few simple rules such as sharply defined priorities, extensive communication, especially across projects, where people stay abreast of what everyone else is 26 doing with weekly formal meetings, and time-pacing (Brown and Eisenhardt 1998). A problem with this perspective is what are those few simple rules? I propose to resolve this theoretical conflict by bringing good old, instrumental process controls back in. However these process controls focus on human rather than physical processes. I also propose to drop control by “culture” and build instead on structuration theory (Giddens 1982): that a few particular rules and resources structure everyday work by providing the common ground upon which people can work together. First, all three new principles are based on processes that emerge over time. These processes include objectives to be achieved, but emphasize how those objectives come to be, how they emerge continually over time, and how they are interpreted. Rather than abstracted, objectified, and simplified physical flows that can be represented mathematically (e.g,. by OR), the processes to be controlled are human processes that are socially enacted, interpreted, negotiated and renegotiated over time. The process controls therefore must foster these human activities of interpretation and sensemaking. Control by controlling the human processes keeps these practices of work out in the open, to be evaluated, monitored, and reconceived as necessary. These process controls therefore keep the practices themselves rich, sensible, and visceral, therefore enabling “reflection-in-action,” or situated judgement. I have already noted that many of the best practices, procedures, processes, tools, and techniques for managing innovation are process controls, in the sense that they map out possible pathways and frame shorter-term activities. I theorize that these best practices will not work unless the organization is based on control by keeping the human practices as processes out in the open and revising them as necessary. If this design principle is incorporated, then managers have any number of systems, milestones, and devices to pace the flows in each community of practice over time (e.g., 27 aggregate product plans combines product concepts with technology and manufacturing strategies; measuring cycle times, yield rates, percent of which sets of experts are oversubscribed or not, capacity utilization, product quality, capacity for variety, customer satisfaction, delivery times, market share, or profit margins may help keep the projects, businesses and strategies in sync; measuring effectiveness of alliances for bringing new products to market, for developing new technologies that have certain potential, for absorbing and leveraging acquisitions (e.g., how many people were retained), and how well platforms deliver over time keeps organizational capabilities in sync with businesses – see Leonard (1998); Meyer and de Tore (2001); Wheelwright and Clark (1991); Cooper (1998). Paralysis by analysis (Langley 1989) is still possible, perhaps more so now because processes are much more sensible, and because scholars and consultants alike seem to really like all these tools, techniques, tactics, and “best practices” (e.g., everything from concurrent engineering to phase reviews, to QFD, to aggregate product planning… the list is almost endless). The organization therefore must also foster situated judgement, as argued above, so that people can make selective and judicious use of the many “best practices.” Situated judgement perhaps can be fostered or even instituted by a certain kind of culture. An innovative organization may well have a certain universal culture in the sense that they all have certain values or norms – this is an empirical question. Rather than culture as broadly defined, it is more to the point to focus on what people do, on enabling certain actions, and on fostering a common experience people can draw on to make sense of their work. Managing what people think is difficult even in closed or total institutions (Goffman 1959), and the innovative organization is inherently open. Managers cannot make people think certain things easily, but they can enable and foster particular activities. 28 The process controls outlined above would be enabled if managers fostered certain kinds of social rules and resources. Structuration theory explains the duality of emergence and determination with the insight that they are part of each other. Emergence needs something that is already determined, since people cannot improvise easily or meaningfully without some preexisting rules. But determination needs emergence, because social structures do not exist on their own – they must be enacted. And since they are always enacted in particular contexts, their reality is always situated. In structuration, “structure” refers to the set of rules and resources that are instantiated in recurrent practice. The innovative organization, I propose, is controlled by three particular rules and three particular kinds of resources. The first two rules have already been discussed as the basis for work as practice: take responsibility for the whole task rather than just your part, and value knowledge for its own sake, so you make what you know accessible to others. The third rule concerns how people approach the problem of value creation that they are working on – as exploring potential (Dougherty, Munir, and Subrahanian 2004). The opposite is to move quickly to a solution or to eliminate problems – which is the rule in most non-innovative organizations I have studied. This third rule frames and helps define the problems that the four communities of practice set and solve, and helps control the nature of these problems. If people define all problems in terms of what we can do, and what the customers will do or can do, then searching for alternatives and options and exploring new opportunities becomes normal and natural (Dougherty, Barnard, Dunne 2004). With this rule for defining problems as exploring potential, people look for longer-term possibilities and actively search for opportunities rather than fixate on the current product. A given product might not embody new ideas, but the ideas would be considered so that 29 possibilities for the next iteration or the next generation are on the table (Garud and Kuraswammy 1995). This rule for work enables people to consider a variety of possibilities at once in their decision making for strategic directions, building manufacturing capability, or designing a product, and also to quickly weed out alternatives (cf. Iansiti 1993). Dealing with the inevitable challenges, glitches, and slips of innovation by looking for alternatives and exploring potential in fact makes the work easier to do. The rule fosters stepping away from the immediate, so people can consider a wide variety of options. If everyone follows this rule, there would be a variety of viable options ready at hand. Resources are central elements of structuring as well, because they provide the capabilities of achieving outcomes (Giddens 1982). Resources are the medium through which power is exercised, and through which “structures of domination” are reproduced. “Structures of domination” refer to who dominates and how, and brings to the fore the fact that large organizations can be dominated by a managerial elite through a “command and control” structure (Perrow 1986). Innovation scholars have always emphasized the importance of power. For example, Clark & Fujimoto (1991) find that “heavy weight” project managers have enough clout to push new products through the system. Dougherty & Hardy (1996) find that successful innovators had the power to make their projects meaningful, while Kanter (1988) points to the need to “empower” innovators. However, most studies take the existing structure of domination as given, rather than question it. I propose that all innovators have control over three kinds of resources that are necessary to achieving their outcomes, which are to set and solve their problems: access to other people’s time and attention, control over how one’s own work will contribute to the entire task (e.g., participatory control over the product concept or technology direction), and control over the 30 specific choices that are made to solve situated problems. These are the most central resources in an organization, so everyday control over them assures that the necessary resources are available to innovators. These resources go hand-in-hand with the three rules. If everyone takes responsibility for the entire task, then each person working on that task has access to the time and attention of other people. Each knows that the others will do their part. If everyone makes their knowledge accessible to others, everyone has the resource of being in the know, and of control over their expertise, how their knowledge will be used. If everyone is exploring potential and pushing for options, then whenever anyone encounters a problem some options are already available. Together, these resources comprise a qualitatively different structure of domination. Managers control the processes and the procedures that enable these rules and resources. Plus they control over their own processes. With these basic rules and resources for control, entrepreneurial, attentive, energized people are the outcome – no magic culture or mystical capabilities are needed since these arise from the structuring of everyday work. Section III: Synthesis, Contingencies, and Empirical Questions I have proposed four new principles of organization design that together define the innovative organization in its own terms rather than as a deviation from the bureaucracy. These are simple principles of design, but also very different from the familiar bureaucratic principles, in fact almost inside out. Instead of generalizing, they situate work in actual, hands-on activities that are carried out by everyone, not just by laborers. Instead of breaking work down by established technique, they weave work together around knowing and learning. Instead of static efficiency, they highlight evolving processes of work. Instead of standardizing action, they activate emergent standards that are negotiated and reconciled as the work proceeds. These principles organize the four paradoxical practices of innovation, and they are grounded 31 specifically in doing this kind of work. These are high-level principles through which people collectively generate and regenerate the work of innovation. The principles produce the context for innovation, and can be complemented by specific structuring techniques to further shape that context for particular industry, strategic, and cultural conditions. Established organizations with a history as a bureaucracy will have considerable difficulty transforming, unless they can implement all four principles, because the familiar bureaucratic principles are waiting in the wings. Burns and Stalker (1961) found that many organizations did not become more innovative because people who began to work in more organic ways experienced “chronic anxiety” about what they were to do with whom and how. Chronic anxiety was not induced by the organic form, but by the failure to define people’s new roles and relationships – the new organizing scheme was not articulated. So people slipped back into their defined, comfortable bureaucratic work roles. Given the continued prevalence of “the bureaucracy” as an institution of modern life, organizations that need only occasional new products may be better off managing innovation the old fashioned way – as an exception to regular work that is located outside the normal organization. For organizations that must continually update, rethink, and occasionally transform products and product categories, I propose that all four new principles are necessary. My theory is that these principles are universal, in the sense that all innovative organizations need them. However, the principles are underlying rules that can be enacted in many different ways, depending on how each one is carried out, on the balancing among them, and on the variety of best practices and other structuring devices that are used. I review each principle and the contingencies and implications it raises. Not all implications are positive, in the sense of creating a glorious, free work life for all. 32 Work as practice highlights the knowing competence of workers, and raises two types of issues for further research. One, instituting work as practice will be difficult without a legitimate frame or schema for “the practice.” Professions and occupations have a frame that provides members with a sense of identity, but work in organizations does not (except for particular occupations). Focusing on services, I have proposed that organizational practices must be defined vividly by articulating the kind of value the practitioners are creating for customers (Dougherty (2004). I also proposed that activities of designing and using the service be built into every job, so that practitioners had an active, experiential sense of the whole problem setting and solving process. That is, the work must be hands-on, situated, and real rather than conceptual, generalized, and abstracted from the context – aka bureaucratic. The second issue is that work as practice opens the organization to partnering with external parties, as well as outsourcing. If all partners were oriented to practice, coordinating joint work and sharing knowledge should be fairly easy. But practice also makes it easy to outsource professional work, and to rely on temporary professionals rather than employees (Barley and Kunda 2004). Differentiating work into four communities organized around distinct problems of innovation reinforces the principle of practice. This principle of differentiation also reduces, if not eliminates, two of the factors that Morison (1966), in his “Gunfire at Sea” example, suggested prevented innovation – severely limited identification with just a part rather than the whole, and a fixation on a specific product rather than on the process. I would expect to find all four communities in all innovative organizations, but how each is organized would vary considerably. In a small organization, the strategic community of practice may be a board of directors or a committee of managers who come together regularly to work on this set of problems, while in a large multinational, this community may be comprised of all business and 33 functional heads who come together quarterly. The relative balance within and among the communities would vary by industry as well. Technology dominates in some industries, while businesses dominate in others. When the “shelf life” of new products or technologies is short, projects and technologies dominate, as Brown and Eisenhardt (1997) suggest. Turning vertical functions into horizontal flows of capabilities is a noteworthy part of this principle. The specialized labor that is the hallmark of the bureaucracy is still here, but managed as capabilities to support continuous streams of new products, businesses, and opportunities. For example, if R&D is organized as a horizontal flow of capability rather than a vertical function, specialized centers of excellence are much less likely to become fixated on sub-optimal goals. Whether or not the capabilities have authority over all practitioners in their specialty is contingent upon the industry and dynamics of knowledge. In some industries it may be necessary for R&D to supervise all scientists, engineers, and/or technicians to assure that people stay up to date, while in others a career process that rotates experts among projects over time may be all that is necessary. But the theory is that the managers of the capabilities are held responsible for how well their people contribute to the projects and businesses. Imbalances among communities can also arise, which if severe can lead to the failure to attend to the four key problems, which returns the organizing system back to a bureaucracy. If the businesses dominate then short term ROI is the focus, and adaptation to new business models would be difficult (as Christensen predicts). One biotech company I know seems to ignore strategic problems – or perhaps the striking complexity of this industry precludes strategic problem setting and solving. The upshot, however, is that they appear to be drifting from one relatively short-term income source to another. A variety of empirical questions can be raised to explore the differences among existing theories that were outlined in the second section above. 34 For example, the strict separation of new from old businesses proposed by Tushman and O’Reilly (1997) may be necessary when established organizations must begin to change, but less necessary once the transformation has occurred. The principle of integration by strategic articulation across communities of practice emphasizes a much more fluid mutual adaptation among the four communities of practice. However, this form of integration works only if the strategies of each community are regularly articulated and juxtaposed – something that apparently is very hard to do, based on the fact that it does not occur often (cf Tushman and O’Reilly 1997). This principle of integration may also require a more focused kind of strategy, a clear, crisp understanding of the value the corporation as a whole creates, and a clear connection among the businesses. Such limitations are empirical questions, and management teams may possess unique competencies for strategic sensemaking. As well, in some industries, the possibilities for innovation may be constrained by regulations, by the capital intensity (e.g. oil and gas), or by preferences of customers. Such constraints would limit the spiraling of adaptation, and may cause the organizing to revert, once again, to a static bureaucracy. One can also ask how tightly or loosely synchronized the communities of problem setting and solving need to be, how various approaches to time-pacing can help, and how more manifest integrating tactics and techniques fit in. Finally, a number of people would be participants in two or more communities (e.g., business level technology directors link technologies to their business and link their concerns to the technology capability community). How these linkages work to reinforce the principle of integration requires study. Finally, the new principle of control transforms the structure of domination. The process controls include many of the best practices for innovation, and serve to channel attention to the processes themselves, not only to abstracted outcomes. This new principle of control supports 35 reflection-in-action, and helps define the problems that all the practitioners are working on. These necessary controls will not work, however, unless the practitioners have control over the everyday resources of time and attention, access to knowledge, and the ability to make their own choices from among the options and alternatives that have been generated. The new principle of control brings us full circle, because it requires the new understanding of work roles and relationships – work as practice – that focuses people’s attention on how they can contribute to the overall task, and how they can make their special knowledge accessible to others. It requires work differentiation into coherent yet sensible practices of problem setting and solving. And it requires strategic articulation, since that shapes the process controls. Contingencies involve which rules and resources are more or less dominant in different industries, how they are enacted, and the number and types of procedures that can be used effectively, based on the underlying rules and resources that make them usable. In conclusion, organizing for innovation is not mystical, weird, or visionary. It requires reasonably straightforward approaches to defining jobs, grouping those jobs into sensible units, coordinating the units, and controlling it all over time. These are ordinary, everyday processes of managing and organizing. However, rather than ignore the work to be done, I focused explicitly on the actual work of innovation, which leads me to develop new principles of organizing. The challenge of organizing for innovation in the 21st century is to develop experience with these new principles, so that this new kind of organizing becomes obvious, simple, and sensible, just like the bureaucracy became for the 20th century. 36 REFERENCES (incomplete; listed articles are not yet properly cited in text, and some citations are missing) Adams, M (2004) Comparative Performance Assessment Study Findings, presentation at the Comparative Performance Assessment Conference, PDMA Foundation, March 2004 Allen, T. 1977 Managing the Flow of Technology. Cambridge, MA: MIT Press. Amabile, T., Conti, R., Coon, H., Lazenby, J. and Herron, M. (1996) Assessing the Work Environment for Creativity, Academy of Management Journal, 39:1154-1184. Ancona, D. and D. Caldwell (1990) Beyond boundary spanninig: Managing external dependence in product development teams, Journal of High Technology Management Reserch, 1,2:119-135. Bacon, G., S. Beckman, D. Mowery, and E. Wilson (1994) "Managing Product Definition in HighTechnology Industries: A Pilot Study," California Management Review, 36:34-56. Baldwin, J. & M. Dapont (1993) Innovation in Canadian Enterprises: Survey of Innovation and Advanced Technology, Statistics Canada, 88, 513-XPB. Barley, S. (1986) Technology as an Occasion for Structuring, Administrative Science Quarterly 31,78-109. Barley, S. 1996 “Technicians in the workplace: Ethnographic evidence for bringing work into organization studies.” Administrative Science Quarterly, 41: 404-441. Barley, S. and G. Kunda 2004 Gurus, Hired Guns, and Warm Bodies, Princeton University Press Bartlett, C. and S. Ghosal 1989 Managing Across Borders: The Transnational Solution, Harvard Business School press. Benner, P 2003, working draft Situated Action and Practice, School of Nursing, UC San Francisco Benner P. 2004 working draft Articulating knowledge, skills, and meanings embedded in nursing practice: Confronting over-simplification, leveling, and the limits of formalism in formal classification systems, School of nursing, UC San Francisco Blau, P. and M. Meyer (1987) Bureaucracy in Modern Society (3rd edition); NY: Random House. Bobrow, E. (1997) The Complete Idiot’s Guide to New Product Development, NY: Alpha Books. Brown, J. S. and Duguid, P. 1991 “Organizational learning and communities of practice: Toward a unified view of working, learning, and innovation.” Organization Science, 2: 40-57. Brown, J. S. and Duguid, P. 2000 “Balancing act: How to capture knowledge without killing it.” Harvard Business Review, may-June: 73-80. 37 Brown, S. and K. Eisenhardt (1997) The Art of continuous change: Linking complexity theory and time-paced evolution in relentlessly Shifting organizations, Administrative Science Quarterly, 42:135. Brown, S. and K. Eisenhardt 1998 Competing on the Edge, Harvard Business School Press Burns, T. and G. M. Stalker (1994: originally 1961) The Management of Innovation. Cambridge: Oxford University Press Capon, N, J. Farley, D. Lehmann, & J. Hulbert (1992) profiles of Product Innovators Among Large US Manufacturers. Management Science, 38(2), 157-169. Carroll, J. (1998) “Organizational Learning Activities in High-Hazard Industries: The Logics Underlying Self-Analysis,” Journal of Management Studies,. Chesbrough, H. 2003 Open Research, Sloan Management Review Clark, K. 1985 “The Interaction of design hierarchies and market concepts in technological evolution.” Research Policy, 14: 235-251. Clark, K. and T. Fujimoto (1991) Product Development Performance, Boston: Harvard Business School Press. Christensen, C. 1997 The Innovator’s Dilemma. Boston: Harvard Business School Press. Cooper, R. (1994) “Third Generation New Product Processes,” Journal of Product Innovation Management, 11:3-14. Cooper, R. (1998) Product Leadership: Creating and Launching Superior New Products, Reading MA: Perseus Books Day, D. 1994 Raising Radicals: Dfiferent processes for Championing innovative corporate ventures Organization Science, 5,2: 148-172. Day, G. (1990) Market Driven Strategy, New York, Free Press. Dessler, G. 2002 A Framework for Management, Prentice Hall Devanna, M and N. Tichy 1990 Creating the competitive organization of the 21st century: The boundaryless corporation, Human Resource Management, 455-471. Dosi, G. 1982 “Technological paradigms and technological trajectories.” Research Policy, 11: 147162. 38 Dougherty, D. 1990 Understanding new markets for new products, Strategic Management Journal, 11:59-79. Dougherty, D. 1992a “A practice-centered model of organizational renewal through product innovation.” Strategic Management Journal, 13: 77-92 Dougherty, D. 1992b “Interpretive barriers to successful product innovation in large firms.” Organization Science, 3: 179-203. Dougherty, D. (1996) Organizing for Innovation, in S. Clegg, C. Hardy, and W. Nord, eds., Handbook of Organization Studies, London: Sage, (1996), pp. 424-439. Dougherty, D. (1999) Organizational Capacities for Sustained Product Innovation, Advances In Managerial Cognition Research, J. Porac and R. Garud, volume editors, Greenwich CT: JAI Press, vol. 6, pp. 79-114. Dougherty, D. 2001 “Re-imagining the differentiation and integration of work for sustained product innovation.” Organization Science. 12, 5, 612-631. Dougherty, D. 2004 Organizing Practice in Services to Capture Knowledge for Innovation, Strategic Organization, 2:1,35-64. Dougherty, D. and T. Heller (1994) “The Illegitimacy of Successful New Products in Large Firms” Dougherty, D. and Hardy, C., 1996. Sustained product innovation in large, mature organizations: Overcoming innovation-to-organization problems. Academy of Management Journal, 39 (5):11201153. Dougherty, D., L. Borrelli, K. Munir, and A. O'Sullivan (2000), Systems of Organizational Sensemaking for Sustained Product Innovation, Journal of Engineering and Technology Management, 17:321-355 Dougherty, D., K. Munir & M. Subramaniam (2004) Technology Flows in Practice: Managing the Human Side of Organizational Technology, Proceedings of the Academy of Management meeting, August 2002. Dougherty, D. and C. Heedful Interrelating in Innovative Organizations: Team Play as the Boundary for Work and Strategy, with C. Helen Takacs, Long Range Planning, special issue on boundaries and innovation, forthcoming. Doughetrty, ,D. H. Barnard and D. Dunne The Rules and Resources that Generate the Dynamic Capability for Sustained product Innovation, forthcoming in Qualitative Organizational Research Fiol, M. (1991) “Managing Culture as a Competitive Resource: An Identity-Based View of Sustainable Competitive Advantage,” Journal of Management, 17, 191-211. 39 Floricel, S. and R. Miller (2003) An Exploratory Comparison of the Management of Innovation in the New and Old Economies, R&D Management. Franko, L. (1989). Global Corporate Competition: Who’s Winning, Who’s Losing, and the R&D Factor as One Reason Why, Strategic Management Journal, 10,449-474. Galbraith, J. (1995) Designing Organizations; San Francisco: Jossey-Bass. Galtunic, C. and K. Eisenhardt 1996 The evolution if intracorporate domains: divisional charter losses in high-technology, multi-divisional corporations, Organization Science 7,3:255-282. Garud, R. and A. Kumaraswamy 1995 Technological and organizational designs for realizing economies of substitution, Strategic Management Journal, 16:93-111. Greenwood, R. and C. R. Hinings (1988) “Organizational Design Types, Tracks, and the Dynamics of Strategic Change, “ Organization Studies, 9 (3), 293-316. Giddens, A. (1979) Central Problems in Social Theory: Action, Structure, and Contradiction in Social Analysis. Berkeley, CA; University of California Press. Giddens, A. 1982 Profiles and Critiques in Social Theory University of California Press. Goffman, E. 1959. The Presentation of Self in Everyday Life. Doubleday: Garden City, New York. Handy, C. 1989 The Age of Unreason Harvard Business School press. HedlundG. 1986 The hypermodern mnc – A heterarchy? Human Resource Management, 25,1:9-35 Helfat, c. and K. Eisenhardt (2001) Inter-temporal economies of scope, organizational modularity, and the dynamics of diversification, working paper, Dartmouth College, 100 Tuck Hall Hanover NH. Helfat, C. and R. Raubitschek (2000) Product sequencing: Co-evolution of knowledge, capabilities, and products, Strategic Management Journal, 21, 10-11; pp. 961-980. Henderson, R. and Clark, K. 1990 “Architectural Innovation: The reconfiguration of existing product technologies and the failure of established firms.” Administrative Science Quarterly, 35:930. Hirsch, P. and D. Levin 1999 Umbrella advocates versus the validity police, Organization Science. Iansiti, M. (1993) Real world R&D: Jumping the product generation gap,” Harvard Business Review, May-June, 138-147. Iansiti, M. 1998 Technology Integration. Boston, MA: Harvard Business School Press. 40 Jelinek, M. and C. Schoonhoven (1990) The Innovation Marathon: Lessons From High Technology Firms, Oxford: Basil Blackwell. Kanter, R 1983 The Change Masters, Simon and Schuster Kanter, R. (1988) When a Thousand Flowers Bloom, Research in Organization Behavior, Greenwich, CT: JAI Press, pp 168-211. Khurana, A. and S. Rosenthal 1997 Integrating the fuzzy front end of new product development, Sloan Management Review, Kohler, T 2003, Logistics Innovation Management: Evidnece from Two Longitudinal Studies (DBA project, Cranfield U) Kogut, B. 2000 The network as knowledge: Generative rules and the emergence of structure. Strategic Management Journal, 21: 405-425. Langley, a. (1989) In search of rationality: The purposes behind the use of formal analysis in organizations, Administrative Science Quarterly, 34: 598-631. Lave, J. and Wenger, S. 1991 Situated Learning. Cambridge, UK: Cambridge University Press. Lawrence, P. and J. Lorsch (1967) Organization and Environment, Boston: Harvard School of Business Administration Press. Leonard, D. (1998) Well-Springs of Knowledge: Building and Sustaining the Sources of Innovation 2nd ed, Boston: Harvard Business School Press. Lynn, G., J. Morone, and A. Paulson “Marketing and Discontinuous Innovation: The Probe and Learn Process,” California Management Review, 38, 3, 8-37. Maidique, M 1980 Entrepreneurs, Champions, and technological Innovation, Sloan Management Review, 59-76. Maidique, M. and R. Hayes 1984 The art of high technology management, Sloan Management Review, 24:18-31. Meyer, M. and A. Detore (2001) Creating a Platform-Based Approach for Developing New Services, Journal of Product Innovation Management, 18,3, 188-204. Mintzberg, H. (1991) “The Structuring of Organizations,” in H. Mintzberg and J. Quinn, eds The Strategy Process, Englewood Cliffs, NJ: Prentice Hall; 330-350. Moorman, C. and A. Miner (1998) “Organizational Improvisation and Organizational Memory” The Academy of Management Review, 23,4,698-723. 41 Morison, E. 1966 Men, Machines, and Modern Times, MIT Press. Nonaka, I and H. Takeuchi 1995 The Knowledge Creating Company, Oxford U Press. Nord, W. and S. Tucker (1987) Implementing Routine and Radical Innovations, Lexington, MA: Lexington Books. Nystrom, P. and W. Starbuck 1981 Handbook of Organization Design, Oxford U Press. Obsfeld, D. 2004 Engineering Knowledge: Innovation as knowledge creation and social movement, working paper UC Irvine Orlikowski, W. (1992) The duality of technology: Rethinking the concept of technology in organizations. Organization Science, 3: 398-427. Orlikowski, W. (2002) Knowing in Practice: Enacting a Collective Capability in Distributed Organizing, Organization Science, 13, 3: 249-273. Orr, J. 1990 “Sharing knowledge, celebrating identity: Community memory in a service culture.” In Collective Remembering, ed. D. Middleton and D. Edwards, pp. 169-189. Newbury Park, CA: Sage. Ostroff, F. 1999 The Horizontal Organization, Oxford U Press. PDMA Handbook, Product Development Management Association Pelz, D. and F. Andrews 1966 Scientists in Organizations, Wiley. Perrow, C. (1986) Complex Organizations: A Critical Essay (3rd ed) New York: Random House Rosenberg, N. (1982) Inside the Black Box: Technology and Economics, Cambridge: Cambridge University Press Schon, D. (1963) Champions for Radical New Inventions, Harvard Business Review, 41, 2, 7786. Schon, D. 1967 Technology and Change, Oxford U Press. Schon, D. 1983 The Reflective Practitioner: How Professionals Think in Action. New York: Basic Books Schon D. 1985; Keynote Address: Marrying science, artistry, the humanities, and professional practice, Cornell Univ. Suchman, L. (1996) Supporting articulation work, in R. Kling, ed Computerization and Controversy: Value Conflicts and Social Choices, 2nd ed, San Diego: Academic Press, 407-423. 42 Thompson, J. (1967) Organizations in Action, NY: McGraw Hill. Tsoukas, H. 1996“The firm as a distributed knowledge system: A constructionist approach.” Strategic Management Journal, 17: 11-25. Tushman, M (2004) presentation given at U of Maryland, January Tushman, M. and C. O’Reilly (1997) Winning Through Innovation, Boston: Harvard Business School Press. Tyre, M. and E. von Hippel (1997) “The Situated Nature of Adaptive Learning in Organizations,” Organization Science, 8: 71-84. Vickers, G. (1965) The Art of Judgment. New York: Basic Books von Hippel, E. (1994) Sticky information and the locus of problem solving: Implications for innovation. Management Science, 40:429-39. Weber, M. (1946) From Max Weber: Essays in Sociology, edited by H. H. Gerth, C. W. Mills, NY: Oxford University Press. Weick, K. 1995 Sensemaking in Organizations. Thousand Oaks, CA: Sage. Weick, K. and K. Roberts 1993 “Collective mind in organizations: Heedful interrelating on flight decks.” Administrative Science Quarterly, 38,3:357-381. Weick, K. and F. Westley 1996 chapter in Handbook of Organization Studies, Clegg, Hardy, Nord, eds. Westley, F. (1990) “Middle Managers and Strategy: Micro-dynamics of Inclusion,” Strategic Management Journal, 11,337-351. Yang, E. and D. Dougherty, Product Innovation Management: More than Just Making a New Product, Creativity and Innovation Management, (1993) 2, 137-155 43 Table 1: What Is to be Organized for Innovation, and Why the Bureaucracy Cannot do It Generic Principles of Organizational Design Paradoxical Practices of Work of Innovation, Organization-wide Bureaucratic Version of Principles Why Bureaucratic Principles Cannot Organize Innovation Define jobs, work rights, obligations, roles, relationships, dignity Balancing freedom vs. responsibility to enable commitment; Ambiguous, problem setting, in parallel, be mindful of whole Balancing New vs. Old for Creative ProblemSolving Many interdependent problems, actions, products, etc. leveraged or woven in Balancing Inside Vs. Outside, for MarketTechnology Linking mutual adjustment and mutual adaptation Balancing Emergence vs. Determination for Monitoring and Evaluation Processes, means, not ends; situated judgement By precisely specified duties, clear and individual accountability, contract By function or solution, add product, market, region; chops of flow of activity into separate parts By hierarchy, snap together decomposed units Work cannot be precisely defined; induces chronic anxiety. Big Q: how to define work so people are energized, reliable, competent? Eliminates problem (can’t be seen), induces dynamic conservatism (fixation on old): Big Q: how to define related sets of activities that makes work “simple” but whole? Prevents continual markettechnology linking, mutual adaptation; drives inward focus. Big Q: how to enable MA and MA? Ignores emergence, so innovation out of control; focus on ends and ignores means, removes situatedness, judgement. Big Q: what to control? Group Jobs (differentiation) Simplifies, gives familiarity, selfdesigning within boundaries Integrate Separate Groupings Quality of state of collaboration… Control so right things are done a right time 44 By A Priori Standards, direct supervision Table 2: New Principles for Organizing Innovation Design Define jobs, work Proposed Principle of Organization Design for Innovative Organizing Define “work” as practice, like professional practice Contingencies By Core Problems of Innovation, with Communities of Practice Around Each: Strategy, Business, Capability (technology, manuf, IT), Project: Highlights problems and horizontal flows of work: Relative emphasis, which dominates, why, when, or gets stuck; R&D as capability solves decent/cent problems; context may intervene, prevent define work so people are energized, reliable, competent Differentiate define related sets of activities that makes work “simple” but whole? Integrate enable MA and MA Control Proceses, not just ends; judgement All jobs are practice? Need a frame for practice (I propose value creation); Enables alliancing and outsourcing of prof. jobs Requires clear articulation and strategic abilities (rare); regulations etc. may reduce reach and sight; ask how well connected externally, Most best practices are process Human Process Controls, Rules and controls, so select and apply to Resources: explore potential, knowledge context; relative emphasis of matters, take resp for whole cycle; resources are access to: options and control rules, resources? of choices; knowledge; and time, attention. Strategic Articulation across four communities (each with own strategy); Lines of sight into future 45 Table 3: Details for Differentiation, Integration, and Control for Each Community of Practice The core problems of innovation that communities of practice differentiate by Integration by Strategic Articulation Strategic community and problem to be Set and solved: problem of strategic articulation, investments, building org. systems; make investments; Integrate businesses, capabilities into Strategic Path (link corp vision with business models, capabilities): juxtapose bus, tech enactments vs. Strategy to keep strategic directions refreshed; select kinds of opportunities; track evolution of performance and functionality network with other CEOs, competitors, governments Business management problem: making money over time by matching bundles of resources with market opportunities, filling in product for life cycle of business Integrate products with business opportunities (link business models with product ideas and portfolios): juxtapose Products vs. strategic standards (attributes, value that are basis of competition): select new products; manage product portfolios; probe new opportunities Capabilities problem – includes technology, manufacturing, marketing, supply chain: developing deep expertise and capabilities for long term value creation Integrate technology with products and businesses (link tech strategy with projects and business models); Select kinds of technologies needed to produce functionality, performance; select new functionality for new opportunities New product development problem: defining and fleshing out product concepts, pulling new product together from firm resources to fit specific market needs Integrate market and technology possibilities in new products. Link product concepts with actual needs and tech options); select, apply technologies to achieve particular functionality of project; interpret specific user needs, 46 Control by Processes, Everyday Rules (explore potential; honor knowledge; take resp for whole project) and Resources (time, authority, options) Define general value to be created by firm, where it is going. Control resources for exploration to communities; provide common frameworks, practices, and procedures; create selection rules for setting and solving problems Process controls: managerial and reward systems, control and movement of charters, strategic planning, probing and learning, six signma Define business model for achieving value over time; provide business expertise to others; pace evolution within businesses via platforms and portfolios Processes: patching, probing and learning, market share growth, new product contributions, customer satisfaction Define, create knowledge sets and expertise for innovation, keep it viable and emerging over time; Control technology trajectories and monitor alternates; oversee how well people contribute to projects, businesses. Processes: concurrent engineering, QFD, TQ, technology mapping, market value measures Provide expertise in project management, integrate across projects, move new insights around projects, Control how criteria for performance are understood, implemented; Processes: phase reviews, project mgt, VOC,