Managing invention and innovation: technological innovation can alter the competitive status of...

When the Industrial Research Institute was founded in 1938, industrial research in the United States had experienced 20 years of dramatic growth, despite the shock of the Depression, and was poised on the brink of World War II expansion that gave it the form and scope we see today. MIT historian

Howard Bartlett reported that from 1921 to 1938 the number of U.S. companies with research staffs of more than 50 persons grew from 15 to 120 (1).

Despite continued rapid increases in industrial R&D involvement and resource commitment over the following 25 years, in 1962, when we founded the MIT Sloan School of Management's Research Program on the Management of Science and Technology, we encountered an academic tradition that for the most part had paid little attention to the organization and management of large-scale technology-based programs. Indeed it was for the purpose of bringing academic research-based insights to bear on such technological enterprises that James Webb, the visionary administrator of the National Aeronautics and Space Administration, urged us with exhortation and funds to begin our research program.

Prior to our start, academics had concentrated largely on two themes: historical romanticism about the lives and activities of great "creative inventors," like Edison and Bell, and psychological research into the "creativity process." While those writings made interesting reading, in my judgment neither track contributed much useable knowledge for managers of technical organizations. Indeed with such rare exceptions as Jewkes et al. (2), the few university researchers who were focusing at that early time upon issues of R&D management were not paying much attention to organizational variables or to innovation as a multi-stage, multi-person, complex process. Perhaps not surprisingly, industry in the early 1960s appeared rather unenthusiastic about social science attempts to probe the underpinnings of effective research, development and technology-based innovation. In contrast, I sense broad acceptance in the 1980s of the results of many academic studies of RD&E, with the Industrial Research Institute especially noteworthy in its efforts to advance collaboration in the field of management of technological innovation.

Invention and Innovation

Roundtable discussions at the 1970 annual IRI spring meeting provide a useful starting point for this review--a set of definitions of the invention and innovation process:

Innovation is composed of two parts': (1) the generation of an idea or invention, and (2) the conversion of that invention into a business or other useful application ... Using the generally accepted (broad) definition of innovation--all of the stages from the technical invention to final commercialization--the technical contribution does not have a dominant position (3).

This leads me to a simple definition of my own, but nonetheless one that I feel is critical to emphasize: Innovation = Invention + Exploitation.

The invention process covers all efforts aimed at creating new ideas and getting them to work. The exploitation process includes all stages of commercial development, application and transfer, including the focusing of ideas or inventions toward specific objectives, evaluating those objectives, downstream transfer of research and/or development results, and the eventual broad-based utilization, dissemination and diffusion of the technology-based outcomes.

The overall management of technological innovation thus includes the organization and direction of human and capital resources toward effectively: (1) creating new knowledge; (2) generating technical ideas aimed at new and enhanced products, manufacturing processes and services; (3) developing those ideas into working prototypes; and (4) transferring them into manufacturing, distribution and use.

Technologically innovative outcomes come in many forms: incremental or radical in degree; modifications of existing entities or entirely new entities; embodied in products, processes or services; oriented toward consumer, industrial or governmental use; based on various single or multiple technologies. Whereas invention is marked by discovery or a state of new existence, usually at the lab or bench, innovation is marked by first use, in manufacturing or in a market.

Most organized scientific and engineering activity, certainly within the corporation, is beyond the idea generating stage and produces not radical breakthroughs but rather a broad base of incremental technological advance, sometimes leading cumulatively over time to major technical change. Academic research in the area of technology management has focused primarily on incremental product innovations oriented toward industrial markets. (Interestingly academic marketing research has focused primarily on incremental product innovations aimed at consumer markets.) Neither the less frequently arising areas of radical innovation nor process innovation has received much systematic attention from academia, unfortunately.

One of my favorite visual aids is Figure 1, portraying a process view of how technological innovation occurs, and emphasizing two key generalizations. First, technological innovation is a multi-stage process, with significant variations in the primary task as well as in the managerial issues and effective management practice occurring among these stages. Figure 1 presents six stages, but the precise number and their division are somewhat arbitrary. What is key is that each phase of activity is dominated by the search for answers to different managerial questions.

[FIGURE 1 OMITTED]

At the outset, for example, emphasis is on finding a motivating idea, a notion of possible direction for technical endeavor. Coming up with one or more technical and/or market goals that stimulate initiating a research, development and/or engineering (RD&E) project is the task undertaken during Stage 1. The relevant managerial question for this stage is, how do more and better targets get generated? Which people, which structures, which strategies can be employed toward more effective idea generation for these objectives? Good managerial practice at this stage frequently involves loose control, "letting many flowers bloom," pursuing parallel and diverse approaches, fostering conflict or at least contentiousness, stimulating a variety of inputs. Critical at this early stage is ready access to small amounts of R&D financing, free of heavy and discouraging evaluative procedures. A major mistake is to set up stringent formal processes for approval of the small sums needed to try out an idea. Distributing small "pots of gold" for first- or second-level R&D supervisors to dispense at their discretion, akin to Texas Instruments' heralded "$25,000 money," makes good sense.

Later, the Stage 5 commercial development for example, the task involves in-depth specification and manufacturing engineering of ideas that have by now already been reduced to an acceptable working prototype. The managerial issues in this stage involve coordinating a number of engineers of different disciplinary backgrounds toward achieving, within previously estimated development budget and schedule, a predefined technical output ready for manufacture in large volume, reliably, and at competitive production costs. Effective managerial practice in this stage might well involve tight control, elimination of duplication, strong financial criteria for resource use accompanied by formal evaluation, single-minded even somewhat rigid adherence to plan, especially in regard to those resources--in many ways the opposite of what is encountered in Stage 1!

The next generalization embodied in Figure 1 is that innovation occurs through technical efforts carried out primarily within an internal organizational context, but involving heavy interaction with the external technological as well as market environment. Proactive search for technical and market inputs, as well as receptivity to information sensed from external sources, are critical aspects of technology-based innovation. All studies of effective innovations have shown significant contributions of external technology and have found success heavily dependent upon awareness of customer needs and competitor activity, indeed one of the most important trends in industrial innovation activity during the 1980s is the continuing increase in the use of external sources of technology as critical supplements to internal R&D efforts.

The details of Figure 1 specify a set of key flows and decision points that occur during the process of innovating. A number of major managerial elements that are embodied in those details will be treated in the remainder of this article. Two aspects of the diagram, however, are potentially misleading and deserve immediate mention.

First, for ease of presentation all stages are shown at equidistant intervals, inappropriately suggesting perhaps the similarity of these phases from a time duration and/or resource consumption perspective. This is by no means true. In particular, while each technical field is characterized by quite different schedule and resource requirements, Stage 5, commercial development, usually takes as long as the several earlier stages combined and requires more resources than most of the other stages together. That is the reason for tight financial standards being properly applied immediately prior to a project's entry to this stage.

Second, for simplicity sake no feedbacks are pictured in Figure 1 from later stages back to earlier ones. Yet, inevitably, these feedbacks exist and cause reiteration to occur among the stages. For example, involvement in the problem-solving process, Stage 3, generates new insights as to alternative idea formulations, Stage 2; and efforts at transfer into manufacturing as part of technology utilization, Stage 6, often create new requirements for problem-solving, Stage 3. Indeed a recent article by Kline (4) argues that the multiple feedback loops are the essence of the innovation process. Thus, the real process of technical innovation involves flows back and forth over time among differing primary activities, internal and external to the dominant innovating organization, with major variations arising throughout the process in regard to specific tasks, managerial issues and managerial answers.

Beyond my descriptive perspective on how innovation occurs is Peter Drucker's (5) prescriptive advocacy as to how it ought to happen: "Systematic innovation ... consists in the purposeful and organized search for changes, and in the systematic analysis of the opportunities such changes might offer for economic or social innovation." Drucker identifies seven sources for innovative opportunity, listed in Table 1. Most challenging to those of us committed to the development and use of new science and technology is Drucker's assertion that "contrary to almost universal belief, new knowledge--and especially new scientific knowledge--is not the most reliable or most predictable source of successful innovations." As with most Drucker "truths," this one is intuitively attractive, as well as unverified in any systematic manner. Whether or not new science is a critical "source" of successful innovations no doubt depends on how you define both "science" as well as "source." But there is no doubt that advances in science and technology are instrumental to the development and implementation of almost all successful product and process innovations. It is the rare case that a success stems from merely a repacking of previously existing science and technology.

Rather than focusing at this point on the claimed sources from which innovations arise, the remainder of this article concentrates on three dimensions--staffing, structure and strategy--each subject to managerial influence and/or control. (I will discuss innovation sources much more in the section on Structure.) A fourth dimension, systems support, will be treated in depth in another article in this Golden Anniversary series. Taken together, improved management of these dimensions contributes critically to achieving successful institutionalized innovation. Throughout the article I shall be seeking to blend the findings from academic research on the management of invention and innovation with the experiences and observations of successful industrial and government technology managers.

Staffing Considerations

Two primary issues arise in regard to staffing the technological organization: what kinds of people need to be involved for effective technical development, and what managerial actions can be taken to maximize their overall productivity. In regard to people requirements, as explained by Roberts and Fusfeld (6), a number of "critical behavioral roles," not just technical skills, must be practiced by the people involved in a technical development. By combining the management of technological innovation literature with our own consulting experiences we identified five key roles for achieving successful innovation, which we have been able to use to enhance RD&E organizational performance. Following us others have since added to this list, generating as many as 12 key roles needing separate monitoring and support.

Critical innovation roles.--First are idea generators, the creative contributors of new insights that both initiate projects and contribute to problem solutions throughout technical projects (7,8). ideas can be drawn from the "market pull" of sensing real or potential customer needs or demands, or from the "technological push" of envisioning the possible extension of technological performance of a material, component or system. Ideas include not just those which lead to project initiation, but also the many throughout an innovation-seeking endeavor which contribute importantly toward invention or innovation outcomes. Thus idea generators for technical projects may be scientists or engineers, sales or marketing persons, or even managers! The rare but valuable idea generators are those who come up with multiple groundbreaking ideas over their careers, such as S. D. Stookey at Coming Glass Works who among other successes came up with key ideas for photosensitive glass and Pyroceram (9). Individual differences that are either innate or developed over long periods no doubt account for many of the distinctive characteristics of effective idea-generators. But many sources of heightened idea creativity arise from managerial influences, for example from the internal organizational climate or environment and especially from supervisory practices. These are discussed in greater detail in my later section on Individual and Organizational Productivity.

There are, however, significant differences between "idea-havers" and "idea-exploiters"--those who come up with ideas and those who do something with the ideas they have generated (10,11). This holds true whether the ideas are born in universities, government labs or in industry. The generally low rate of energetic pursuit of newly created RD&E ideas mandates the requirement for the second key role in technical innovation-seeking activities, that of the entrepreneur or product champion. Entrepreneurs advocate and push for change and innovation; they take ideas, whether their own or others', and attempt to get them supported and adopted. Most major studies of factors affecting product success have found the active presence of a product champion to be a necessary condition for project success (12). For example, in recent years the late Ken Estridge gained widespread repute as the product champion behind IBM's successful development and launch of its personal computer. Lew Lehr rose to the presidency of 3M as an eventual outgrowth of his own championing of 3M's health products business (13). And despite not being a family member of the closely-held Pilkington Glass company, Alistair Pilkington became "process champion" for the revolutionary float-glass process which dramatically changed the company and the industry and again, not incidentally, led this "champion" into the chairmanship of the firm.

As I reported in my first Research Management article, the entrepreneurial "role" is the same, whether carried out internally in existing organizations or "externally" in their own newly founded companies (14). But the mode of behavior and what is needed for "internal" vs. "external" entrepreneurial success may well be different, as expanded by Maidique (15). My own studies of "internal entrepreneurs" found that they needed to be sensitive to company politics and the latest corporate "buzzwords" in order to gain internal support, and as indicated below required a high level "sponsor" to lead them through the corporate jungle. Lehr (13) argues that strong entrepreneurial efforts are needed even within companies that have long traditions of fostering entrepreneurship, in order to overcome inevitable managerial resistance.

A third required role in effective innovative activities is the program manager or leader, sometimes strangely called the "business innovator," supplying the support functions of planning, scheduling, monitoring and control, technical work supervision, business and financial coordination relating to the R&D project (16,17). This is the one "role" which is also usually an assigned job in the organization, the other roles being incidental to an individual's specific work assignment.

Gatekeepers, or special communicators, are the fourth critical role identified, the link-pins who frequently bring information messages from sources outside of a project group into that group (18). These human bridges join technical, market and manufacturing sources of information to the potential technical users of that information. Gatekeepers may bridge one technical group to another within the same company, or may link university research activities to a corporate advanced technology center, or may tie customer concerns into a supplier's design team.

Tom Allen's pioneering empirical studies of the functioning of the technical gatekeeper have been extended by many other academics and broadly accepted and applied throughout industry and government. For example, a study of so-called "bridge scientists" at Stanford Research Institute found that such individuals are rare but easily identified (19). Effective "bridgers" were found to be interpersonally able (e.g., good listeners), have depth in at least one discipline, have a wide range of interests, and be oriented toward problem solving. As I have argued repeatedly (6) a vital extension of the concept is the recognition that some gatekeepers can "bridge" to market or manufacturing inputs, rather than just technical information, often bringing in raw or processed information, or points of view, that are otherwise lacking within the R&D organization itself.

The final key role is that of the sponsor or coach, performed usually by a more senior person who is neither carrying out the R&D itself nor is directly and personally aggressively championing the change. The role is one of providing encouragement, psychic support, facilitation to the more junior people involved in the task implementation, often including important help in "bootlegging" the resources needed by those trying to move technological advances forward in an organization (14). My research data affirmed the logical--the higher up in an organization a sponsor was located, the higher the probability of success of internal efforts to generate new product lines. Sponsors are often needed for idea generators, project managers, and especially for entrepreneurs. A good example of the effectiveness of the "ultimate sponsor," the corporate chief executive officer, is Chapman's sponsorship of Gorman's work on "available light motion pictures" at Eastman Kodak (20). But CEOs as project sponsors can be organizationally dangerous too! Who is to turn off the CEO's pet project when it runs amok?

These several critical roles are all needed within or in close contact with each internal working group in order for it to achieve successfully the goals of an innovative outcome. But in addition, the effective development and maintenance of a technical organization requires recognition of these differentiated roles in order to create and implement appropriate people management processes, including recruiting, job assignment, personnel development and training, performance measurement and rewards.

Individual and organizational productivity.--Beyond the people and role behaviors needed for effective staffing are the principal managerial acts that can affect staff creativity, inventiveness and productivity. The 30 years of Research Management reveal plentiful discussions of approaches for stimulating creative idea-generating among scientists and engineers, including such techniques as brainstorming, Synectics, and morphological analysis. (See 21 and 22 for extensive reviews.) Despite the enthusiastic testimony in the various articles, I remain unconvinced that systematic evidence supports the use of these methods, which frankly I view as mainly gimmicks. As documented below, effective individual and group supervision, including proper maintenance of group diversity and task challenge, seem to me more likely to produce useable ideas.

Stages of a scientist or an engineer's career, and the composition of his/her immediate work group are primary influences upon technical productivity (or creativity or inventiveness, if you prefer). This generalization rests upon a broad foundation of research into the performance of technical people and project groups. Katz (23) has demonstrated that technical professionals evolve through three career stages which he labels socialization, innovation and stabilization. As with the different stages of a project cycle, each stage of an individual's career provides a new set of managerial challenges for maximizing personal productivity. The setting of work norms, providing task direction, and joining new employees into the internal technical communications network are managerial issues confronted during the socialization or job "break-in" stage. In contrast, maintaining the employee's earlier motivation and renewing technical skills are among the very different sort of questions needing treatment in the stabilization or job maturity phase.

But personal and group productivity are not just influenced by the individual's job cycle. The nature of the immediate work group, its composition and supervision, matter greatly. In general what Kuhn (24) called "creative tensions," a mix between comfort-reinforcing stability and conflicting challenge, seems desirable. For example, multi-dimensional diversity among technical colleagues in a project team heightens technical performance (7). Variations in age, technical background, even personal values, correlate with enhanced group productivity. This need for internal challenge is further reflected in the findings that the average years a group has worked together significantly impacts upon that group's technical productivity (25). The long-term stable technical group apparently becomes too self-secure, diminishes its outside technical contacts, and decreases its performance. Supervisory intervention at the technical group or RD&E project level seems able to affect this performance, however. For example, technical skills of the first-level group leader, and not human relations skills, enhance a group's effectiveness (26). And even the stable technical team can be moved to high-performing status with proper leadership, in this case requiring strong direction and control by the project manager (17). Thamhain and Wilemon (27) support the importance of the project manager's technical expertise and reliance upon work challenge as major sources of effective technical performance.

Organization Structure

The design of organization structures that will enhance technological innovation requires focusing on both the organization's inputs and its outputs.

Effective RD&E organizations need appropriate technical and market information inputs, and their outputs need to be integrated toward mission objectives and transferred downstream toward their ultimate users.

Market inputs.--Managerial research has repeatedly demonstrated that 60 to 80 percent of successful technical innovations seem to have been initiated by activities responsive to "market pull," i.e. forces reflecting orientation to perceived need or demand (28-30). Of particular note is the recent IRI study of basic research in industry (31), which among many other interesting conclusions produced the unsought finding that "most innovations come about as a result of the recognition of a market need or opportunity. While the push of new technology is also important, it plays a distinctly secondary role." These studies less frequently indicate how technical organizations uncover these needs. Sometimes one person's personal "hobby-horse" forces "market" consciousness to initiate and sustain a technical program, especially when coupled with that individual's entrepreneurial drive and skills, as in Peter Goldmark's successful pursuit of the long-playing record. As Goldmark exclaimed, "My initial interest in the long-playing record (LP) arose out of my sincere hatred of the phonograph ... it seemed to violate what I thought the quality of music should be" (32).

In organizations less dominated by one key figure, "market gatekeepers" or customer liaison personnel frequently aid the technical organization to better understand its customers' requirements, priorities or preferences. For example, Corning Glass supposedly discovered the need for optical waveguides as a result of one of its staff visiting the British Post Office. Organizing to gain meaningful market inputs for research and engineering use may depend upon explicit assignments of such responsibilities to cooperating marketing staff or to RD&E people themselves. The product development cycle should be organized, as suggested in Figure 1, to bring market inputs into design repeatedly, during the early product specification stage, and again during prototyping through active involvement of selected customers. As a sharp contrast to desirable practice, in one consulting project for a major chemical company I found the sales organization prohibiting R&D people from visiting "their" customers, lest the R&D people agree with customer complaints!

A special prospective customer for innovations, often overlooked when R&D does occasionally seek market inputs, is the company's own manufacturing activity. Yet depending on the company and industry, the manufacturing organization turns out to be the eventual "customer" of anywhere from one- to two-thirds of the company's technological developments. Manufacturing, similar to an outside unrelated potential product customer, has to decide whether or not it wants to "buy" an internally developed improvement in materials, components, manufacturing equipment or overall production process, for its own internal "consumption." That prospective in-house manufacturing "customer" deserves at least the same degree of involvement with the design and development process as does an outside firm or individual. If R&D's "market-oriented" ties to its own manufacturing group can be improved, the potential for significantly impacting company performance is high, especially given the recent IRI study results that show R&D aimed at process innovation as far more likely to succeed than that targeted toward new and improved products (33). But I am convinced that overcoming the gap between the central lab and a major plant installation usually needs special efforts and sometimes creative organizational designs, in particular when important process changes are contemplated.

Rather than seeking collaboration to provide market information to the RD&E process, many companies have ill-advisedly substituted marketing-oriented control of RD&E. Organizational subordination of research and engineering to "product managers" (inevitably marketing or sales people) or tight budgetary control of RD&E by these units may force market-based criteria to dominate technical project selection. But this is usually accompanied by a short-term quick-fix orientation, erosion of technical capability, and gradual destruction of product/process competitiveness. Analyses by Souder (34) have demonstrated that strong and positive relations between R&D and marketing organizations significantly improve the track record on new product introductions. In my experience this is best achieved by welding partnerships among equals, rather than by extracting compliance from subordinates. Good examples of such R&D/ marketing partnerships are evident in the team structures used by both Hewlett-Packard and 3M in their new-product pursuits.

Market research techniques have long been used to help define consumer preferences in new product designs (35). These methods have been less helpful for developing industrial goods. Recently von Hippel (36) has demonstrated that potential industrial customers whose needs place them at the leading edge of technological demands can be used to specify detailed desired performance characteristics and features for as yet nonexisting products. Military or space research requirements are frequent sources of such "leading edge" requirements. Fusfeld (37) has long used this insight in developing forecasts of the rate of market penetration of new technologies. The problem however is to distinguish a customer demand that is truly in the vanguard of future broader market needs from the "cry for help" from what amounts to the "lunatic fringe" that exists in almost every technical field. That fringe also has needs that are real and extraordinary, but unfortunately not representative of future growth opportunities.

Technical inputs.--Despite the presumed dominant role of "market pull" as a source of innovative projects, "technology push," i.e. undertaking projects for advancing the technical state-of-the-art in an area without anticipation of the specific commercial benefits to be derived, is also the critical source of many significant product and process successes. The many studies cited earlier still show these technology-push successes to be in the minority, but unfortunately do not clearly indicate the relative worth of the two approaches. One confusion leading to arguments is to assert that if market pull is the key, then market research should be more effective than it has proven to be? Collier (38) for example quotes Barnes' listing of "Neoprene, nylon, polyethylene, silicones, penicillin, Teflon, transistors, xerography, and the Polaroid Land camera" as not resulting from "a market research study of what people said they wanted." While market research is not the only or even the primary indicator of "market pull," many research directors especially sympathize with Collier's point of view. Guy Suits (39), long-time leader of General Electric's research efforts, cites Langmuir's work on hydrogen dissociation, leading to a new type of welding, as a good example of technology push. Indeed Casey (40) goes further, arguing that misleading market research was a contributor to the long period required for commercial development of high fructose corn syrup. More logs are heaped on this fire when one cites the supposed market research studies (often claimed to have come from the same large consulting organization!) that demonstrated no meaningful market prospects for computers, instant photography or the dry copier.

When technical advancement is the goal, managers have long understood that professional depth in an organization is achieved by grouping people together in their own area of specialization, with work assigned and performance supervised by a more accomplished person of the same specialization (41). This approach is called functional or discipline-based or specialty-oriented organization. It is the traditional organization structure of the craft guild and of the university. Multiple specialists working together interact comfortably, using the same general knowledge base, analytic skills and tools, and vocabulary. When technical people are organized in functional arrays, their natural interplay brings depth of specialized capabilities to bear on technical problems. Indeed, Marquis and Straight found that technical groups organized in functional forms have the highest technical excellence (42).

But in any nontrivial technical field the vast majority of applicable technical know how exists outside of a performing technical organization. For technical effectiveness even a strong functional team needs to draw upon the pre-existing technical knowledge that is in the outside world, whether in the technical literature, in already developed products and processes, or especially in the minds of other technical professionals. For example, as illustrated in Table 2 several studies point out that for innovations eventually developed within a firm, about 60 percent of the sources of the initial technical ideas had outside origins (28). Allen (43) has demonstrated the relative differences among channels for technical information input to an organization, distinguishing what is readily accessed from what is used most effectively in coming up with high-rated problem solutions. His work as well as that of others (44,45) indicates the minor role played by the literature, especially in contributions to engineering and development, in contrast with personal contacts, experience and training.

One factor that inadvertently has significant effect on technical inputs to RD&E groups is the architectural layout of their work space. Early observations by Jack Morton, then vice president of semiconductor research at Bell Laboratories, led to his concern for the physical separation between technical organizations that were intended to relate to each other (46). Research at MIT by Muller-Thym in the early 1960s empirically established spatial effects on the frequency of communication among engineers and scientists in the same laboratory.

These concepts have been well-developed by Allen (18) into careful findings on specific design elements of RD&E architecture. The distance between two potential communicators, vertical separation, walls and other architectural features importantly influence technology flows.

Thus far I have addressed what affects technical inputs in support of an organization's internal invention activities, the first element of the two-step innovation process I defined initially. What about technical inputs not aimed at invention but rather at innovation directly? Clearly technological solutions (inventions) already exist elsewhere, and an innovating organization might merely adopt or adapt them by slight modification for a new purpose. This would permit skipping the first stage of invention and going directly to the exploitation stage. An early U.S. study determined that 22 percent of key successful innovations had been adopted or adapted (44) while comparable U.K. data indicated a 33 percent adoption rate (45). Japanese data on license fee payments for foreign technology show a long-established pattern of heavy use of outside technology. A small study of Taiwanese innovations found adoptions to have accounted for the bulk of successes (47). While specific percentages no doubt have changed in recent years, adoption or adaptation of prior outside inventions is a major source of innovation worldwide, but apparently still substantially underutilized by U.S. firms. In recent years the growth of research consortia, effective or not, and the rapidly growing number of "strategic alliances" between large corporations and new firms in areas of emerging technologies indicate that more looking to the outside for technology is taking place, even by U.S. companies. I will discuss this development further in the section on Strategy.

One unique source of potential adoptions is the user. Von Hippel (48) has shown that users frequently create and implement innovations for their own use, followed later by manufacturer adoptions of those innovations for large-scale production and distribution. His research on scientific instruments and several areas of manufacturing equipment demonstrated that heavy percentages of new products had been user-developed.

Technical organizations need to be designed to facilitate accessing these several different sources of technical information inputs, whether as contributions toward internal inventions or as sources for adoption more directly as innovations. A variety of approaches are suggested, ranging from such simple considerations as ensuring that at least some salespeople have technical skills and/or incentives so that they bring back a customer's ideas in addition to his orders. Much more ambitious are the IBM marketing department's several "applied science centers" across the United States, established adjacent to concentrations of innovative users to learn about new software and hardware developments and transfer that technical information back into IBM's product development groups. IBM's Cambridge operation in Technology Square, working closely with MIT's pioneering Project MAC, thus became the source of IBM's first commercial computer time-sharing system, a field adaptation of an innovative user's development. As one approach to overcome biases against outside sources of technology, increasingly corporations are establishing the position of Chief Technical Officer or Vice President of Technology, with broad responsibility for both internal technology development as well as external technology acquisition. Organizational experiments to enhance both technical and market information inputs are underway across a broad front.

Output-focused organization.--Just as the functional organization structure maximizes technical inputs, the project, program, mission or product organization is intended to integrate all inputs toward well-defined outputs. By placing in the same group, under a single leader, all the contributors toward a given objective, the project organization maximizes coordination and control toward achieving output goals. The Marquis and Straight study (42) cited earlier supports these findings. But project structures have a fundamental flaw that seriously affects many technical organizations. The project form tends to remove technical people from organizational groups in which they interact with colleagues of their own scientific or engineering discipline. Furthermore, the project manager may be technically expert, but inevitably in only one of the disciplines of his or her subordinates, not all of them. If the project has long duration, especially when the technology base is rapidly changing, the technical skills of the project members erode over time due to lack of stimulating technical reinforcement and supervision.

This dilemma has led to the creation of an organization that is intended to be a "compromise"--the "matrix" structure in which technical performers are supposed to maintain active membership in two organizations, their original discipline-based functional group as well as the focused project group. In theory the "matrixed" person thus has two bosses, one functional and one project, each of whom will extract his appropriate "due," thereby attempting simultaneously to maintain the technical skills and performance of the individual, more or less, while orienting his loyalty and contributions toward the project's output goals, more or less! However, most technical "matrix" organizations are only "paper" matrices, not "real" matrices--they appear to be matrices on organization charts but do not strongly pull the engineer between two conflicting masters.

If one wanted to obtain truly matrixed individuals, the influences that push a technologist's time and attention toward competing sets of objectives (e.g., functional excellence vs. project schedule demands) would have to be roughly balanced between those objectives. A technical contributor's priorities are influenced by: (a) who is responsible for his/her performance evaluation and reward distribution; (b) who makes the individual's specific task assignments; (c) where is the individual physically located relative to the two "competing" managers; (d) what is the longer-term career relevance of the competing groups; and (e) what is the relative persuasiveness (whether based on personality or power) of the two managers. Achieving even a rough balance among these influences would only be practicable by dominance of the functional manager on some of these dimensions, dominance of the project manager on others, and perhaps rough equivalence of the two managers on still other influences upon matrixed persons. The absence of reasonable balance in most "paper matrixed" cases leads the actual situation to its "default" condition, with the achieved results reflecting the characteristics of the dominant organization form, either functional or project but seldom both. Recent studies suggest that certain patterns of dominance among these contending influence sources achieve better performance of matrix organizations (25).

Output transfers.--But in addition to generating outputs, the technical organization needs to be designed to enhance output transfer downstream toward eventual customers and users. Downstream is where innovation takes place and where benefits are realized! A consulting survey of prestigious major corporate research laboratories has indicated a high degree of dissatisfaction with the extent and effectiveness of transfer of results to potential recipient groups (49,50). Three different clusters of bridging approaches were found helpful in increasing transfer in those labs--procedural, human and organizational. Most organizations used a variety of these approaches, often several simultaneously. My findings have been reaffirmed by recent comparative case studies by an internal task force at IBM (51) and by a consulting project at Union Carbide (52), among others.

Procedural methods include: joint planning of RD&E programs by the performing group and the organization that is expected to be the receiver, often resisted by R&D as an "invasion" of its turf; joint staffing of projects, especially pre- and post-transfer downstream; and joint project appraisal after project completion, done cautiously if at all after failures in order to avoid destructive fingerpointing.

Human bridges are the most effective transfer mechanisms, especially the upstream and downstream transfers of people. Movement of people upstream: (a) brings with them information on the context of intended project use; (b) establishes direct person-to-person contacts that will be helpful in later post-transfer troubleshooting; and (c) creates the image that the project eventually being transferred has involved prior ownership and priority inputs from the receiving unit. Later movement of people downstream: (a) carries expertise for post-transfer problem-solving; and (b) not unimportant, conveys the risk-reducing impression that the receiving unit will not be stuck with solving post-transfer problems by itself. Other human bridges that are widely used include rotation programs, market gatekeepers, joint problem-solving sessions, and other formal and informal meetings.

Organizational techniques for enhancing transfer are usually more complicated to design and implement than procedural or human bridge approaches. "Integrators," sometimes named "transfer managers," or integrating departments are frequently appointed to tie together the sending and receiving organizations. This person or unit is given the responsibility for moving the project from the sender into operating condition in the receiver organization, either lacking authority in one or the other organization or being matrixed between both.

More ambitious organizational approaches include dedicated transfer teams, established solely for the period during which technical results are being transferred to their "customers," done especially for moving purchased process technology. Venture teams, discussed further below, are also employed to reduce functional organizational transfer issues, shifting leadership responsibility among the many-disciplined team members as the primary phase of the project shifts from research to engineering to manufacturing to sales.

Strategy

Strategic management of technology includes both strategic planning and strategic implementation aspects at either of two levels: (a) overall, for the entire technology-dependent firm, government agency, division or product line; or (b) more focused, for just the technology development/acquisition process/ department/laboratory of the entire organization. As recently as ten years ago neither of these levels of strategy was the subject of much serious scholarship, or even management consulting practice. Few researchers carefully studied the overall management of the technology-intensive company. And fewer still addressed the questions of how to incorporate technological considerations into overall business strategy.

Strategic planning focuses upon the formulation of an organization's goals and objectives, and upon developing the policies needed to achieve those objectives, including identification of the organization's primary resources and priorities. But developing corporate strategy with such a global perspective, including technological dimensions, is quite new. Indeed the evolution of corporate strategic planning as a field of practice is divisible more-or-less into three decades: the 1960s, during which multi-year budget projections became the earliest forms of financial planning, sometimes mislabelled "long-range planning"; the 1970s, when market growth/share matrices and market attractiveness considerations added a new dimension to strategic analysis; and the 1980s, during which technology as a strategic factor became so widely acknowledged as to cause firms and even countries to realize that financial, marketing, and technological considerations needed to be integrated in overall strategy development (53).

Strategic thinking and planning.--Horwitch and Prahalad (54) provided an early set of perspectives at the overall strategic level, differentiating the key issues of technology-oriented strategic management among three modes: the small, usually single-product, high-tech firm; the large, multi-market, multi-product corporation; and the multi-organization, even multi-sector societal program. For each of these, Horwitch and Prahalad find a primarily non-overlapping set of strategic issues and priorities. More recent writing has focused upon similarities between the first two "modes," the entrepreneurial smaller firm and the successfully innovative larger corporation (55-58; see also 14). Maidique and Hayes conclude that to be innovative the large corporation needs to manage the "paradox" of chaos versus continuity, similar to the "creative tensions" required for the innovative technical person (7,24).

In moving from strategic thinking toward strategic planning we need principles for developing more detailed technology strategies. But what are the underpinnings of technological change, especially as it relates to the corporation, upon which overall technology strategy should be based? Three general observations seem critical here, all linked to the dynamics of technological innovation processes: (1) there are characteristic patterns over the life cycle of a technology in how frequently product versus process innovations occur; (2) each stage of technology has differing critical implications for innovation, including type, cost, degree of invention, and source; and (3) an organization's efforts to generate technological innovation create almost inevitable internal dynamics in the allocation of R&D efforts, generating multiple management problems. Each of these is discussed more fully below, with suggestions of related technology planning and strategy development approaches.

Utterback and Abernathy (59) demonstrated that a technology tends to evolve in three stages. Most technologies move from an early "fluid stage," dominated by frequent product innovations, through a "transition stage," characterized by significant process innovation and the emergence of a dominant product design, into a "specific stage," featuring lower rate of and more minor product and process innovations. While variations in this pattern of course occur, some of which are already well understood (60, 61), this generalization becomes one important basis for developing a company's or a product line's technology strategy.

One of the most significant findings from this research has been the reaffirmed role of the smaller firm as the dominant source of innovation during the earliest emerging stage of a technology, with the locus of innovation shifting toward larger companies in the transitional and more mature stages of a technology (59). Most studies that have sought to find differences in R&D productivity as a function of company size have not made this critical distinction as to the stage of technology or type of innovation. Consequently, the findings of these economic analyses have varied unconvincingly all over the lot, from some that have asserted the large company is most productive of innovations to others that have claimed the exact opposite, to still other studies that have found nonlinear ties between size and R&D results.

The potential stability or predictability in patterns of technology evolution is the rationale for attempting to use technological forecasting techniques as part of technology planning and strategy development. Most technology forecasting methods are simple, often inadequate for the task (62, 63). Indeed, despite recent "rediscovery" by some consultants of technology S-curves for forecasting and planning (64), the intellectual development of the technology forecasting field more or less stopped over a decade ago (65, 66). Yet some corporations have benefited enormously from thoughtful application of technology forecasting methods to their strategic analyses. Tracy O'Rourke, the chief executive officer of Allen-Bradley, for example, cites a comprehensive technology forecast as the basis for planning his company's successful transition from electromechanical to solid-state electronic devices (67).

Each stage of a technology is associated with different strategic implications. The earliest stage in a technology's life cycle tends to feature frequent major product innovations, heavily contributed by small entrepreneurial organizations, often closely tied to lead user needs. The development of frozen orange juice concentrate by the National Research Corporation and its spinoff companies is one such example (68). The present rash of biotechnology discoveries are coming primarily from university laboratories directly or from young small enterprises, leading irresistably to the explosion of biotech alliances between large companies and the new startups. The same alliance pattern has evolved in the areas of machine vision and artificial intelligence.

The intermediate stage of a technology's life cycle may include major process innovation, with continuing but lessened product variation occurring, with increasing numbers of competitors, both large and small. To achieve the dominant product-process design during this stage, large corporations sometimes undertake long-term development programs that combine many elements of applied research and engineering. For example, General Motors' successful efforts in developing its two-cycle diesel engine included more than ten major developments needed for the final system (69).

The late stage of a technology features less frequent minor product and process innovations, contributed primarily by large corporations, motivated mostly by cost reduction and quality improvement operational objectives. As illustrated by Hollander's (70) careful analysis of DuPont rayon innovations, shown in Table 3, these numerous minor innovations can produce dramatic cumulative impact upon costs. In fact the so-called "learning curve," i.e. decreasing unit manufacturing cost as cumulative production increases, results primarily not from the volume itself but rather from the usual continuing allocation of engineering efforts to incremental cost reduction projects as a product line's volume increases. Management of the technical investment is the primary source of the so-called "learning curve" competitive advantage, not the share of market.

These key dimensions of a technology described above should strongly influence choices made by a firm or government agency in developing its technological strategy. A company's detailing of its "product innovation charter" (71), or its application of project selection principles or techniques (72) as part of technology planning, ought to reflect at least general consideration of the current stages of its principal technologies. In particular, the late stage of one technology usually corresponds to earlier stages of other potentially threatening technologies. Most corporations fail to anticipate or even appropriately respond to these technological threats (64, 73).

Technology life cycles occur in an industry as a whole, thus providing an "environmental" set of influences upon a single organization's strategy. A different kind of cycle, however, is produced within a firm by its own strategy has been ignored thus far. Yet government regulatory activities in regard especially to health and safety have had significant positive and negative influences on technological innovation (90-92). But, as pointed out by Abernathy and Chakravarthy (93), government's strategic role has also included actions to create technologies directly (via the Horwitch/Prahalad Mode III, for example, 54) as well as indirectly through market modifications (94). in a sense, the variety of alternatives facing governments for influencing technological change are equivalent to the corporate venture alternatives described previously.

In Conclusion

Recent work by Gobeli and Rudelius (95) provides a fitting basis for finishing this article. In their integrative comparative analysis of five firms in the technology-intensive cardiac-pacemaker industry they observe the differing competitive impacts that have come from the multiple stages of the innovation process. Managing at the creativity phase is not enough, nor even is managing manufactured quality sufficient, nor is managing that is focused primarily upon any other single aspect of innovation. They reaffirm the importance of key innovation-supporting people roles. Gobeli/Rudelius describe the importance of market-technology linkages, effective program management, government intervention, and appropriate goal-setting, planning and risk-taking for firms in this medical electronics industry.

Technological innovation can provide the potential for altering the competitive status of firms and nations. It can contribute to increased corporate sales and profits, as well as individual and national security and well-being. But its purposeful management is complex, involving the effective integration of people, organizational processes and plans. Only recently have some companies undertaken bold and broad action steps to try to institutionalize an effective product and process innovation program. Two firms in particular, long known for effective innovation, have publicized ambitious multi-faceted endeavors. 3M has based its attempts heavily upon a so-called "intrapreneurship" approach, while Coming Glass has developed a more broad-based effort, including redesigning organization structures, changing incentive programs, and undertaking widespread management involvement and educational change activities. In both cases leadership for these multi-year institutional change programs came from the CEO's office. Other companies would be wise to consider whether top-down company-wide commitments to accelerate and enhance effective innovation might not also apply to them.

This article has argued a host of generalizations about managing the process of invention and technological innovation, each supported by literature, empirical research, and practitioner experience. Some of these generalizations have already been widely diffused into practice, such as recognition of the gatekeeper's importance to information flow and nearly everyone's efforts to stimulate internal entrepreneurship. Some of my other contentions may still be subject to debate, modification and even rejection as we learn more.

Both academics and technology managers need to join in this continuing search for clearer managerial insights about technological invention and innovation and more effective organizational performance. The National Research Council's recent report, Management of Technology: The Hidden Competitive Advantage, summed up the goals: "Effective work in the field of Management of Technology can play a crucial role in devising the strategies and imparting the skills and attitudes to U.S. engineers and managers that they will need in the future technology-dominated economy" (96). It listed eight challenges of critical importance to industrial competitiveness: How to integrate technology into the overall strategic objectives of the firm; How to get into and out of technologies faster and more efficiently; How to assess/ evaluate technology more effectively; How best to accomplish technology transfer; How to reduce new product development time; How to manage large, complex, and interdisciplinary or interorganizational projects/systems; How to manage the organization's internal use of technology; How to leverage the effectiveness of technical professionals. Hopefully more light will be shed on these key industrial needs prior to the IRI's Diamond Jubilee!

References

(1.) H. R. Bartlett. The Development of Industrial Research in the United States. Washington: National Research Council, 1941.

(2.) J. Jewkes, D. Sawers, and R. Stillerman. The Sources of Innovation. London: Macmillan, 1958.

(3.) Research Management. "Top Research Managers Speak Out on Innovation," November 1970.

(4.) S. J. Kline. "Innovation is Not a Linear Process," Research Management. 1985.

(5.) P. F. Drucker. Innovation and Entrepreneurship: Practice and Principles. New York: Harper & Row, 1985, 35-36.

(6.) E. B. Roberts and A. R. Fusfeld. "Staffing the Innovative Technology-Based Organization," Sloan Management Review, vol. 22, no. 3, Spring 1981.

(7.) D. Pelz and F. M. Andrews. Scientists in Organizations, Revised Edition. Ann Arbor, Michigan: University of Michigan Press, 1976.

(8.) F. M. Andrews. "Innovation in R&D Organizations: Some Relevant Concepts and Empirical Results," in E. B. Roberts et al., editors. Biomedical Innovation. Cambridge: MIT Press, 1981.

(9.) S. D. Stookey. "History of the Development of Pyroceram," Research Management, Autumn 1958.

(10.) D. Peters and E. B. Roberts. "Unutilized Ideas in University Laboratories," Academy, of Management Journal, vol. 12, no. 2, June 1969.

(11.) E. B. Roberts and D. Peters. "Commercial Innovation from University Faculty," Research Policy, vol. 10, no. 2, April 1981.

(12.) A. H. Rubenstein, A. K. Chakrabarti, R. D. O'Keefe, W. E. Souder, and H. C. Young. "Factors Influencing Innovation Success at the Project Level," Research Management, May 1976.

(13.) L. W. Lehr. "Stimulating Technological Innovation: The Role of Top Management," Research Management, November 1979.

(14.) E. B. Roberts. "Entrepreneurship and Technology: A Basic Study of Innovators," Research Management. vol. 11, no. 4, July 1968.

(15.) M. A. Maidique. "Entrepreneurs, Champions, and Technological Innovation," Sloan Management Review, vol. 21, no. 2, Winter 1980.

(16.) D. G. Marquis and I. M. Rubin. "Management Factors in Project Performance." MIT Sloan School of Management Working Paper, 1966.

(17.) R. Katz and T. J. Allen. "Project Performance and the Locus of Influence in the R&D Matrix," Academy of Management Journal, vol. 26, 1985.

(18.) T. J. Allen. Managing the Flow of Technology. Cambridge: MIT Press, 1977.

(19.) J. Gartner and C. S. Naiman. "Overcoming the Barriers to Technology Transfer," Research Management, March 1976.

(20.) L. J. Thomas. "Available Light Movies--An Individual Inventor Made it Happen," Research Management, November 1980.

(21.) W. E. Souder and R. W. Ziegler. "A Review of Creativity and Problem Solving Techniques," Research Management, July 1977.

(22.) H. Geschka. "Introduction and Use of Idea-Generating Methods," Research Management May 1978.

(23.) R. Katz. "Managing Careers: The Influence of Job and Group Longevities," in R. Katz, editor. Career Issues in Human Resource Management. Englewood Cliffs, N.J.: Prentice-Hall, 1982.

(24.) T. S. Kuhn. The Structure of Scientific Revolutions. Chicago: University of Chicago Press, 1963.

(25.) R. Katz and T. J. Allen. "Investigating the Not Invented Here (NIH) syndrome: A Look at the Performance, Tenure, and Communication Patterns of 50 R&D Project Groups," R&D Management, vol. 12, no. 1, 1982.

(26.) G. F. Farris. "The Technical Supervisor: Beyond the Peter Principle," Technology Review, 1973.

(27.) H. J. Thamhain and D. L. Wilemon. "Leadership, Conflict, and Program Management Effectiveness," Sloan Management Reviews, vol. 19, no. 1, Fall 1977.

(28.) J. M. Utterback. "Innovation and the Diffusion of Technology," Science, vol. 183, no. 4125, 15 February 1974.

(29.) A. Gerstenfeld. "A Study of Successful Projects, Unsuccessful Projects, and Projects in Process in West Germany," IEEE Transactions on Engineering Management, vol. EM-23, no. 3, 1976.

(30.) R. Rothwell, C. Freeman, A. Horsley, V. T. P. Jervis, A. B. Robertson, and J. Townsend. "SAPPHO updated-project SAPPHO phase II," Research Policy, vol. 3, 1974.

(31.) W.C. Fernelius and W. H. Waldo. "Role of Basic Research in Industrial Innovation," Research Management. July 1980.

(32.) P. C. Goldmark. "How the LP Record was Developed Or the Case of the Missing Fuzz," Research Management, July 1974.

(33.) N. R. Baker, S. G. Green, and A. S. Bean. "The Need for Strategic Balance in R&D Project Portfolios," Research Management, March-April 1986.

(34.) W. E. Souder. "Effectiveness of Product Development Methods," Industrial Marketing Management, vol. 7, 1978.

(35.) G. L. Urban and J. R. Hauser. Design and Marketing of New Products. Englewood Cliffs, N.J.: Prentice-Hall, 1980.

(36.) E. A. von Hippel. "Lead Users: A Source of Novel Product Concepts," Management Science. July 1986.

(37.) A. R. Fusfeld. "How Not to Fall on Your Face in Technological Forecasting," Inside R&D, vol. 7, no. 2, January 1978.

(38.) D. W. Collier. "More Effective Research for Large Corporations," Research Management, vol. 12, no. 3, May 1969.

(39.) C.G. Suits. "Selectivity and Timing in Research," Research Management, vol. 5, no. 6, 1962.

(40.) J. P. Casey. "High Fructose Corn Syrup--A Case History of Innovation," Research Management, September 1976.

(41.) P.R. Lawrence and J. W. Lorsch. Organization and Environments. Boston: Harvard Business School, 1967.

(42.) D. G. Marquis and D. L. Straight. "Organizational Factors in Project Performance." MIT Sloan School of Management Working Paper#133-65, 1965.

(43.) T. J. Allen. "Performance of Information Channels in the Transfer of Technology," Industrial Management Review. vol. 8, no. 1, Fall 1966.

(44.) S. Myers and D. G. Marquis. Successful Industrial Innovation, Washington: National Science Foundation, 1969.

(45.) J. Langrish, M. Gibbons, W. G. Evans, and F. R. Jevons. Wealth from Knowledge. London: Macmillan, 1972.

(46.) J. A. Morton. Organizing for Innovation. New York: McGraw-Hill, 1971.

(47.) A. Gerstenfeld and L. H. Wortzel. "Strategies for Innovation in Developing Countries," Sloan Management Review. vol. 19, no. 1, 1977.

(48.) E. A. von Hippel. "Has a Customer Already Developed Your Next Product?", Sloan Management Review, vol. 18, no. 2, Winter 1977.

(49.) E. B. Roberts. "Stimulating Technological Innovation: Organizational Approaches," Research Management, vol. 22, no. 6, November 1979.

(50.) E. B. Roberts and A. Frohman. "Strategies for Improving Research Utilization," Technology Review, vol. 80, no. 5, March/April 1978.

(51.) H. Cohen, S. Keller, and D. Streeter. "The Transfer of Technology from Research to Development," Research Management, May 1979.

(52.) J. J. Smith, J. E. McKeon, K. L. Hoy, R. L. Boysen, L. Shechter, and E. B. Roberts. "Lessons from 10 Case Studies in Innovation," Research Management, vol. 27, no. 5, September-October 1984.

(53.) E. B. Roberts. "Strategic Management of Technology," in Global Technological Change: Symposium Proceedings. Cambridge: MIT Industrial Liaison Program, June 1983.

(54.) M. Horwitch and C. K. Prahalad. "Managing Technological Innovation--Three Ideal Modes," Sloan Management Review, vol. 17, no. 2, Winter 1976.

(55.) J. B. Quinn. "Technological Innovation, Entrepreneurship, and Strategy," Sloan Management Review, vol. 20, no. 3, Spring 1979.

(56.) T. J. Peters and R. H. Waterman. In Search of Excellence. New York: Harper & Row, 1982.

(57.) M. A. Maidique and R. H. Hayes. "The Art of High-Technology Management," Sloan Management Review, vol. 25, no. 2, Winter 1984.

(58.) J. Friar and M. Horwitch. "The emergence of technology strategy: a new dimension of strategic management," Technology in Society, vol. 7, nos. 2 and 3, Winter 1985/1986.

(59.) J. M. Utterback and W. J. Abernathy. "A Dynamic Model of Product and Process Innovation," Omega, vol. 3, no. 6, 1975.

(60.) J. M. Utterback. "Systems of Innovation: Macro/Micro," in W. N. Smith and C. F. Larson, editors. Innovation and U.S. Research. Washington: American Chemical Society, 1980.

(61.) J. M. Utterback and L. Kim. "Invasion of a Stable Business by Radical Innovation," in P. R. Kleindorfer, editor. The Management of Productivity and Technology in Manufacturing. New York: Plenum Press, 1986.

(62.) E. B. Roberts. "Exploratory and Normative Technological Forecasting: A Critical Appraisal," Technological Forecasting. vol. 1, no. 2, Fall 1969.

(63.) A. R. Fusfeld and F. C. Spital. "Technology Forecasting and Planning in the Corporate Environment: Survey and Comment," in B. V. Dean and J. L. Goldhar, editors. Management of Research and Innovation. TIMS Studies in the Management Sciences, vol. 15, North-Holland Publishing Co., 1980.

(64.) R. N. Foster. Innovation: The Attacker's Advantage. New York: Summit Books, 1986.

(65.) E. B. Roberts. The Dynamics of Research and Development. New York: Harper & Row, 1964.

(66.) J. P. Martino. Technological Forecasting for Decision Making. New York: Elsevier, 1972.

(67.) T. O'Rourke. Presentation at Pugh-Roberts Associates, Inc., Workshop on Critical Issues in Technology Management, April 15, 1986.

(68.) D. H. Peters, "The Development of Frozen Orange Juice Concentrate," Research Management, vol. 11, no. 1, January 1968.

(69.) R. A. Richardson. "Research Toward Specific Goals: Development of the Light-Weight, Two-Cycle Diesel," Research Management, Summer 1958.

(70.) S. Hollander. The Sources of Increased Efficiency. Cambridge: MIT Press, 1965.

(71.) C. M. Crawford. "Defining the Charter for Product Innovation," Sloan Management Review, vol. 22, no. 1, Fall 1980.

(72.) M. R. Baker and W. H. Pound. "Project Selection: Where We Stand," IEEE Transactions on Engineering Management, vol. EM-11, no. 4, December 1964.

(73.) A. C. Cooper and D. Schendel. "Strategic Responses to Technological Threats," Business Horizons, vol. 19, no. 1, February 1976.

(74.) E. B. Roberts. "Research and Development System Dynamics," in E. B. Roberts, editor. Managerial Applications of System Dynamics. Cambridge: MIT Press, 1978.

(75.) H. B. Weil, T. A. Bergan, and E. B. Roberts. "The Dynamics of R&D Strategy," in E. B. Roberts, editor. Managerial Applications of System Dynamics. Cambridge: MIT Press, 1978.

(76.) K. G. Cooper. "Naval Ship Production: A Claim Settled and a Framework Built," Interfaces, vol. 10, no. 6, December 1980.

(77.) A. R. Fusfeld. "How to Put Technology into Corporate Planning," Technology Review, vol. 80, May 1978.

(78.) M. E. Porter. Competitive Advantage: Creating and Sustaining Superior Performance. New York: The Free Press, 1985.

(79.) S. B. Johnson. "Comparing R&D Strategies of Japanese and U.S. Firms," Sloan Management Review, vol. 25, no. 3, Spring 1984.

(80.) T. Hirota. "Environment and Technology Strategy of Japanese Companies." MIT Sloan School of Management Working Paper #1671-85, June 1985.

(81.) M. A. Cusumano. "Diversity and Innovation in Japanese Technology Management," in R. S. Rosenbloom, editor. Research on Technological Innovation, Management, and Policy. Greenwich, CN: JAI Press, 1986.

(82.) E. B. Roberts. "New Ventures for Corporate Growth," Harvard Business Review, vol. 59, no. 4, July-August 1980.

(83.) D. Dimanescu and J. W. Botkin. The New Alliances: America's R&D Consortia. Cambridge: Ballinger Publishing, 1986.

(84.) Business Week. Strategic Alliances: New Competitive Muscle. New York: October 6 7, 1986.

(85.) E. B. Roberts and A. Frohman. "Internal Entrepreneurship: Strategy for Growth," The Business Quarterly, vol. 37, no. 1, Spring 1972.

(86.) N. Fast. "A Visit to the New Venture Graveyard," Research Management, March 1979.

(87.) P. E. Haggerty. "Strategy, Tactics, and Research," Research Management, vol. 9, no. 3, 1966.

(88.) E. B. Roberts and C. A. Berry. "Entering New Businesses: Selecting Strategies for Success," Sloan Management Review, vol. 26, no. 3, Spring 1985.

(89.) M. H. Meyer and E. B. Roberts. "New Product Strategy in Small Technology-Based Firms: A Pilot Study," Management Science, vol. 32, no. 7, July 1986.

(90.) W. M. Capron, editor. Technological Change in Regulated Industries. Washington: The Brookings Institution, 1971.

(91.) Y. J. Allen, J. M. Utterback, M. S. Sirbu, N. A. Ashford, and J. H. Hollomon. "Government Influence on the Process of Innovation in Europe and Japan," Research Policy, vol. 7, no. 2, April 1978.

(92.) O. Hauptman and E. B. Roberts. "FDA Regulation of Product Risk and its Impact Upon Young Biomedical Firms," Journal of Product Innovation Management, vol. 4, No. 2, June 1987.

(93.) W. J. Abernathy and B. S. Chakravarthy. "Government Intervention and innovation in Industry: A Policy Framework," Sloan Management Review, vol. 20, no. 3, Spring 1979.

(94.) J. M. Utterback and A. E. Murray. "The Influence of Defense Procurement and Sponsorship of Research and Development on the Development of the Civilian Electronics Industry." MIT Center for Policy Analysis Working Paper CPA-77-2, June 1977.

(95.) D. H. Gobeli and W. Rudelius. "Management Innovation: Lessons from the Cardiac-Pacing Industry," Sloan Management Review, vol. 26, no. 4, Summer 1985.

(96.) National Research Council. Management of Technology: The Hidden Competitive Advantage. Washington: National Academy Press, 1987.

Awareness of customer needs plays a powerful role in invention and innovation. For example, shortly after transistor co-inventor William Shockley started at Bell Labs in 1936, he was alerted to the need to replace relays in telephone exchanges by research director Mervin J. Kelly. Kelly's "eloquent pep talk" started Shockley on the solid state research that paid off with the invention of the point-contact transistor

in December 1947. The photo at left, taken a few months later, shows Shockley seated at the apparatus used in his investigations with co-inventors John Bardeen (standing left) and Walter Brattain. Another 1947 invention--"one-step photography "--is shown in the photo at right. It is being demonstrated by its inventor, Edwin H. Land, who has credited his little daughter with asking why she could not see at once the picture he had just taken of her on a sunny day vacationing in Santa Fe, New Mexico. In his 1965 Charles F. Kettering A ward address, Land recalled how he immediately "undertook the task of solving the puzzle she had set before me. "Within the hour, as he walked around the town, the camera, the film and the physical chemistry became clear enough to Land that he was able to describe the dry camera in detail to Polaroid's patent attorney. Three years of "timeless intensive work" were then required before this demonstration could take place at a meeting of the Optical Society of America on Feb. 21, 1947. Photos courtesy of AT&T Bell Laboratories (left) and Polaroid Corporate Archives.

RELATED ARTICLE: An inventor considers the process.

I think the processes of invention are like creativity in the arts. The artist does not start with a blank piece of paper. He may have a blank piece of paper in front of him, but he starts with his prior training, his culture, and a bag of tricks based on his own past experience. A poet uses the words he has learned, the words of his language, and no other. His appreciation of beauty is the result of his training and culture. He is driven by ambition, ego, or desire for glory and reward.

This is identical to the drives and mechanisms of the inventor. The inventor also starts with a blank piece of paper, but he uses his training, experience, and intuition not only to create a solution, but often to invent a problem. The invention of a problem is a far greater achievement than inventing a solution. Great inventions are the mothers of necessity and not the other way around.

An inventor may recognize a need before the rest of the world does, or he may create the need because he recognizes that his invention could make life richer, easier, or in some other way, better. He uses his technology as an artist uses his. Technology is only a means to an end. The elegance and beauty of a simple solution to a complicated problem and the appreciation of this elegance, beauty, and simplicity is as much a form of art as putting together notes, words, or colors by an artist.

The question has often been raised as to whether public or even private incentives, drives, and promotional schemes can increase the output of inventions. I have no doubt that the answer is a resounding "yes." A climate of good music for an appreciable time would produce great musicians and composers. There is a great deal of potential talent of all kinds in the genes of the human race. If one wants to produce great soccer players, one has to encourage the playing of soccer, and now and then a great Pele does result. If you teach a great many young people how to play the violin, you would get a few great violinists; if you teach them how to sing, you will get some great singers. Similarly, you can encourage inventions.

This requires creation of glory, rewards, "hoopla" in the media, and all the things that encourage people to pursue any field of activity--Jacob Rabinow, holder of more than 200 patents including the magnetic particle clutch, the automatic clock regulator and, most recently, a pickproof lock; writing in Chemtech, March 1980, p. 145.

RELATED ARTICLE: "Managing Invention and Innovation" revisited.

In rereading "Managing Invention and Innovation," I find little to regret and nothing to retract. The framing approach of Staffing, Structure and Strategy seems reasonable and I still believe in the advice I provided almost 20 years ago. However, two critical aspects have changed over these years that bear comment:

First, today I would focus more on the big differences between managing incremental innovation and major/ breakthrough/radical innovation. We understand well how to manage the incremental innovation efforts that consume most R&D budgets and keep mid-to-large size firms moving forward, albeit with relatively slow pace and occasional hiccups. But we do not understand well how to manage major/breakthrough/radical innovation efforts, and that inability creates large failures, cyclical company performance, stagnation, and even erosion.

Second, the world of invention and innovation has become far broader than it had been, and internal-external integration has become today's most important managerial challenge.

Incremental Innovation

Established companies spend 80-90 percent of their technology budgets on upgrades, modifications, extensions, and add-ons to present product lines. The strategy behind this is the conventional drive for continuity and reasonable growth, if possible, of the existing businesses. The inputs that call for the incremental changes come largely from the existing marketplace.

I hasten to specify that I regard the "market" as having the three dimensions of customers, competitors and regulators. Those "actors" provide clear guidance, sometimes incorrectly, as to what is demanded in terms of alterations of current offerings. (Upon reflection it's even amazing that only 80 percent of successful innovations are seen as resulting from "market pull." Given the market dominance of incremental innovation, I can imagine that more like 90-95 percent of innovations are in effect generated by "market pull"!)

With this demand situation and pressure for changes in primarily familiar technological areas, the internal projects to accomplish the innovations are usually simply staffed with technical people, a linkage to the market from regular sales/marketing channels, and a skilled project manager. No unique internal entrepreneur or senior sponsor is needed to move these projects forward. The existing product organization smooths the way for accomplishment and then downstream integration of incremental advances. The organization structure to do this can be the straightforward functional organization, with project coordination, or it can be a dedicated project team because the usual short-term nature of an incremental project will not produce technical deterioration of the group during the project life.

Most companies know how to manage these efforts. The result is that most incremental innovation attempts succeed, but frequently only incrementally in terms of market size and profitability. And eventually, overdependence upon this strategy leads to failure of the product line and perhaps the firm.

Major/Breakthrough/Radical Innovation

In contrast, established firms devote a small fraction of their resources to trying to pull off big things, whether they are big in scale (i.e., major projects) or big in degree of market or technical change (i.e., breakthroughs or radical undertakings). Here, managerial understanding is weak and failure dominates. The staffing for such efforts needs to follow the arguments made in my 1988 article: full innovation teams of idea generators, entrepreneurs and champions, technical and market gatekeepers, project managers, and senior sponsors. The organization structures might be large-scale complex project groups, led by "heavyweight project managers" (including possibly the most senior executives as the "heavyweights" or at least as the "sponsors"), or they might be matrix stuctures with their own internal conflicts to resolve.

The strategic issues here are those three that 1 argue in my recent article in Research-Technology Management (1): senior executives need to provide linkage between their own insights to corporate vision and direction and tie them to the choices and priorities of major undertakings; the firm's internal resources need to be leveraged by selected elements of the vast numbers of alternative outside sources of especially technical capabilities; and the firm must push for technological leadership, not followship, of at least its key competitors.

Two of the more stiking approaches being taken toward large-scale innovation objectives are IBM's "Emerging Business Opportunities" and P&G's "Connect and Develop" programs. Both follow all three dimensions that I have proposed, but persistence of CEO commitment and follow-through may well become the determinant of their eventual impacts.

The Outside World of Invention and Innovation

The past 20 years have dramatically changed the way that firms of all sizes, along with university and government labs, now collaborate in pursuit of successful innovation. Along every stage of the life cycle discussed in my reprinted article are now many key external "venture-oriented" alternatives and complements (2).

In my 1988 article, these venture approaches earned merely a few ending paragraphs, despite my long-held enthusiasm. But today more large companies than ever before are operating their own corporate venture capital groups, and a new organization in which I am personally involved, Synchrony Ventures, is trying to bring the same strategic benefits to equity-based alliances with emerging technology firms to Fortune 2000 companies. Indeed, these corporate mechanisms may well replace the traditional venture capital firm, which has backed away from early-stage investing.

Alliances for both incremental and breakthrough innovation are now commonplace globally. "Open Innovation" at the community level has moved us in several fields far beyond the more limited assumptions of external user contributions (3). The resulting challenge is that now invention and innovation must be managed across organizations rather than just inside of them, as well as around the world. What will the next 20 years bring?--E. B. Roberts, Jan 2007.

References

(1.) Roberts, Edward B. 2004. Linkage, Leverage and Leadership Drive Successful Technological Innovation. Research-Technology Management, May-June, pp. 9-11.

(2.) Roberts, Edward B. and Liu, W. Kathy. 2001. Ally or Acquire? How Technology Leaders Decide. MIT Sloan Management Review. Fall (43, 1), pp. 26-34.

(3.) von Hippel, Eric. 2005. Democratizing Innovation. The MIT Press, Cambridge, MA.

Edward Roberts has been a leader for over 40 years in research and education on the management of technological innovation and its' practical applications to management. He is the David Sarnoff Professor of Management of Technology at the Massachussetts Institute of Technology, where he long chaired the Sloan School's Management of Technological Innovation and Entrepreneurship Group. He co-founded and co-chaired for nearly 20 years the MIT Management of Technology (MOT) Program, and also founded and chairs the MIT Entrepreneurship Center. He developed and heads the new Entrepreneurship & Innovation Program within the MIT Sloan School's MBA Program. He also served as co-director of the MIT International Center for Research on the Management of Technology.

In his entrepreneurial activities Roberts' co-founded and was CEO of Pugh-Roberts Associates, an international management consulting firm specializing in strategic planning and technology management, now a division of PA Consulting Group. He co-founded and is a director of Medical Information Technology, Inc., a producer of healthcare information systems, and also co-founded and is a director of Sohu.com, Inc., a Chinese Internet firm. In addition, Roberts co-founded and was for 20 years a general partner of the Zero Stage Capital and First Stage Capital Equity funds, a group of venture capital funds investing in early-stage technology-based firms. He has been a co-founder and/or director of numerous emerging technology companies, including at present Advanced Magnetics, PR Restaurants, Interactive SuperComputers, and Visible Measures.

Professor Roberts has authored 160 articles and 11 books, the most recent being Entrepreneurs in High Technology (Oxford University Press, 1991) and Innovation: Driving Product, Process and Market Change (Jossey-Bass/Wiley, 2002). A member of RTM's Board of Editors, he first published in the journal in July 1968. Roberts has four degrees from MIT in electrical engineering (B.S. and M.S.), management (M.S.), and economics (Ph.D.). eroberts@mit.edu.

Table 1.--Seven Sources for Innovative, Opportunity

The unexpected
The incongruity
Process need
Changes in industry or market structure
Demographics
Changes in perception, mood, and meaning
New knowledge, both scientific and nonscientific

From P. F. Drucket, Innovation and Entrepreneurship." Practice
and Principles. New York: Harper & Row, Publishers, 1985.

Table 2.--Sources of Ideas for Innovations
Developed within the Firm

                                          % from Outside
Author              Study            N       the Firm

Langrish et al.     Queen's Award    51         65
Mueller             DuPont           25         56
Myers/Marquis       5 Industries    157         62
Utterback           Instruments      32         66

From data contained in J. M. Utterback, "Innovation in Industry
and the Diffusion of Technology," Science 183, February 15, 1974.

Table 3.--Cost-Reducing Innovation
in DuPont Rayon Plants

                         Contribution of Minor Technical
                         Change to % of Net Reduction in
                                Unit Costs Due to
Plant                            Technical Change

Spruance II-A                           83
Spruance I                              80
Old Hickory                             79
Spruance III                            46
Spruance I                             100

From data in S. Hollander, The Sources of Increased Efficiency.
Cambridge: MIT Press, 1965.