Junk Your `Linear’ R&D!

To capture innovation from diverse sources, don’t rely on traditional, centralized programs. The parallel paradigm speeds-up market-driven ideas.

January 1 1989 by Donald N. Frey


The general perception in the U.S. is that we are in danger of being out-innovated by foreign competitors -particularly the Japanese-and that formal research and development may no longer be the key to leadership. Europe shares a similar perception about the U.S. and Japan. Despite a lack of formal R&D programs, the Japanese have risen to become innovative, competitive economic powers. They dominate world markets by successfully commercializing some of our basic technology. This realization has led Europe and the U.S. to a renewed recognition of the importance of commercial innovation. Sources of innovation are diverse and do not necessarily include R&D laboratories. For example, studies have shown that customers of industrial concerns are an important source of innovation, and salespeople of these concerns play a key role as both communicators and originators of innovation. Getting innovative products to the marketplace today requires a new concept of R&D.

The assumption that economic success stems from acquiring-unilaterally at the company level-enough proprietary research findings, is no longer generally true.

There will continue to be rare exceptions, such as biotechnology and superconductive ceramics, based on new and revolutionary scientific findings. But they will soon mature into broadly available technologies. The spread throughout the world of science and technology, plus the sheer volume of new research, makes it almost impossible for any one firm to corner the market. “Airplane ticket” research is the new order of the day, and an attitude of “not invented here” is likely to prove fatal. Science and technology need to be considered a pool from which all companies can extract what they need for a particular innovation.

The growing universality of science and research makes innovation more frequently market-driven instead of technologically driven. External forces affecting industrial R&D include the restructuring or rationalization of American industry through takeovers (hostile or friendly) and other forms of recapitalization. Restructuring and rationalization tend to sharpen the focus of R&D on the short term, which translates into short-term product or process development and little long-term research. Managements concentrate too much on survival in the takeover world-frequently with excessive debt. They do not make long-term investments because they see the risks as excessive. In the short run, these rationalization forces are probably irresistible. In the long run, the restructuring fad will likely run its course. Meanwhile, however, less costly and more time-efficient methods for innovation, including innovation produced through R&D, are needed to help correct this short-term behavior.

In recent years, reductions or decentralization of large industrial research labs to accomplish this goal have been well-publicized. Intermediate-sized firms, such as Borg-Warner and Cincinnati Milacron, are doing the same thing. In 1975, I dissolved the central research laboratories of Bell & Howell to decentralize R&D and thus innovation development. The result? A number of successful major innovations followed in the next ten years, compared to none in the previous decade. The rationale of these actions is that central R&D is too far from the marketplace to avoid failures of new products ill-suited for that market, too removed from customer ideas or influence, and too slow to exploit new technologies-some of which may not even have been considered in an inward-directed R&D organization. While not all R&D laboratories are being totally decentralized, all are under various forms of pressure to change. Specifically, there are three emerging trends in industrial R&D: a stronger focus on short-term development, with less focus on long-term research; the creation of more innovation-oriented R&D in a different context; and a different mode of R&D operation, with more emphasis on speeding up market-driven innovation. 

THE PARALLEL PARADIGM

While some centralized R&D organizations are being replaced by decentralized functions, most R&D organizations have also changed in functional content, becoming multi-functional, and in degree of permanence, more flexible. The term “multifunctional,” rather than “multi-disciplinary,” is used because the latter has become identified with science rather than with parallel business functions used to develop the various elements of new products. The old “hand-off,” or linear methodology, is too slow and fraught with downstream organizational barriers which contribute to failure. Simultaneous and parallel development involving engineering, manufacturing and marketing, is gaining sway. (See cover story for Philips’ Cornelis van der Kluges vision of “co-marketing.”) Given the market-driven character of most new innovation, marketing must play an integral role from the beginning, as must engineering and manufacturing, in order to achieve timely, competitive cost and quality.

Impermanence has become the new style in R&D. Ad hoc organizations or teams are being created for specific innovation. If successful, a team can be the core of a new business or product line. If unsuccessful, it is dissolved, with members sent to new ad hoc teams. The team can report to various authorities: If innovation is vital to the survival of the firm, the team often reports to an operating entity (if management can be trusted to innovate). Occasionally, it may temporarily report to a staff entity-if no existing operating business can handle the assignment, or if the innovation has no fit to existing operations.

What remains of an “old” R&D, whether centralized or decentralized, can be better used as a methodology for solving technical problems that arise in the operation of innovation teams. So a new R&D paradigm which can be called “multi-input,” “multifunctional development in parallel,” or “parallel” for short, emerges. In the process, management structure organization, reward systems, and methods for choosing innovation investments are in flux. A great deal of experimentation will occur as the new paradigm becomes more fully developed. 

SERVICE-SECTOR CONSIDERATIONS

The emergence of the service sector as a dominant force in the U.S. economy has raised some additional and specific R&D issues. This sector has questionable productivity growth and little identifiable or formal R&D. A service-sector firm has two aspects with two different R&D needs: the service or professional function performed, such as healthcare, law, banking, etc., and support information systems. While R&D-to further the professional purposes of the firm-is being conducted extensively in such major sectors as healthcare and education (at both the Federal and firm level) the comprehensive study of information technology and systems for the service industries is just now starting. Much of that research has not yet fully considered multirecording media, including paper and microfilm, in addition to electronics or multitransmission, whether through the post office, deliveries, or electronic means. As with R&D in the manufacturing sector, studies of successful information system innovations for service industries indicate that decentralization with multifunctional teams at the operating level, rather than at centralized MIS (with the possible exception of corporate standards setting), is vital for effective and timely execution. 

INFORMATION SYSTEMS “R&D”

Information systems “R&D” is two-tiered, relying on customized software as well as commercial hardware and software. The growing and intense operational nature of economically successful information systems-which contrasts with the more traditionally centralized and isolated MIS systems-is resulting in the creation of ad hoc teams to develop new information systems. These teams are multi-disciplinary and multi-functional because of the often intense customer and people contact involved in service sector information systems. But the service sector needs more than just efficient information systems, it needs effective ones. This can be achieved by designing a system to suit the specific strategy or purposes of the firm-for example, to improve market share, customer satisfaction, responsiveness, or cash flow.

The R&D paradigm is shifting from the increasingly unsuccessful and unrealistic “linear” model to a “parallel” model incorporating ad hoc multi-functional teams. Regardless of whether these teams are decentralized to operating units, they have a single project orientation with multiple units. Any part of the more traditional centralized R&D which remains, can be used in two ways: by ad hoc teams external to the laboratory to solve problems which are uncovered in the team’s innovation development; or to technically improve the innovation once it is an established product or business in the marketplace.

Major issues in technology transfer-the interaction between public R&D and private innovation-are now emerging. The private sector has a much bigger hand in development than research, but research is increasingly done better in the broad-based science and applied-science research institutions. These include private and public research universities, government research laboratories, and profit and non-profit private research laboratories. Government research laboratories present special problems in technology transfer, but efforts are being made to correct these problems.

Science today is simply too broad, too international and changing too fast for any one company to master it all. Increased private-sector financial support to institutions can come directly through grants and contracts, or, for tax-supported laboratories, indirectly through tax dollars. How well each side understands the other’s needs will determine the degree of the financial support they receive. Greater understanding will also produce more realistic expectations of economic growth on the part of our political institutions.


Frey, dismayed with the barriers to commercialization of his own research, when research engineer at Ford and then former Professor of Engineering at the University of Michigan, became a strong advocate of cross-educating students in engineering and manufacturing. Formerly chief executive of Bell & Howell, he presently teaches at Northwestern University.