CHAPTER THIRTY-FIVE
R&D: The Best Is Business-Driven
ONLY A FEW YEARS AGO, security analysts routinely ranked stocks by the percentage of sales a company spent on research and development. But R&D spending no longer correlates with business results.
Hoffmann-La Roche, the Swiss pharmaceutical giant, has a research budget second to none in the industry. It hasn’t had a significant new product since the ’60s. Siemens, the German electrical giant, though renowned for both its research quality and its research budget, has not come up with new products for years. America’s best-known research center, AT&T’s Bell Laboratories, continues to give birth to one scientific “spectacular” after the other—in acoustics and optics, in computers and in mathematics. But unlike the Bell Labs discoveries of the past, these new scientific breakthroughs have so far not generated great commercial successes. Bell Labs’ Tokyo counterpart, the lab of NEC, similarly has not been able to turn its huge investment in telecommunications research into salable products.
But other companies—often spending much less—show signal R&D results. Merck in the U.S. and some British firms, such as Glaxo and Wellcome, are bringing out one successful new drug after another. The smallest of the old-line electrical manufacturers, Sweden’s ASEA, has so successfully innovated as to become a world leader in three highly competitive fields: electrical locomotives, direct-current transmission, and industrial robots. To blame “outside factors,” such as “excessive government regulation,” for what ails R&D simply will not do.
Prescription for Failure
For a hundred years it has been axiomatic that R&D has to be a separate function, doing its own scientific and technical work by itself. But the successful innovations in both pharmaceuticals and computers—and in other fields—are now being turned out by cross-functional teams with people from marketing, manufacturing and finance participating in research work from the very beginning.
Most companies still believe that the better an innovation—whether a product or a service—fits what the customer already is doing, the more successful it will be in the market. The need for the customer to buy costly new equipment, they think, should be kept to a minimum. But this approach is becoming a prescription for failure.
Federal Express, a couple of years back, correctly foresaw the boom in facsimile transmission. It came out with a new service that offered most of the benefits of facsimile without the customer’s having to spend a nickel on expensive new equipment. The service was a resounding flop—but the same customers are now lining up to buy their own fax machines. NEC’s failure to become a major factor in the world’s telephone markets was caused in large measure by its embodying the new technology in a hybrid: its switchboards were only half electronic, thus enabling potential customers to keep on using their existing electromechanical equipment. The customers instead are buying fully electronic switchboards from NEC’s competitors.
Since the industrial research lab was created some 90 years ago—by the chemical industry in Germany and by GE’s Charles Steinmetz in the U.S.—successful R&D has been “technology-driven.” Mr. Steinmetz, for instance, began his pioneering work on the fractional horsepower motor by spelling out the technical specs: power inputs and outputs, running speeds, friction, temperatures, durability, and reliability. This then enabled him to define what new scientific knowledge and new engineering competence would have to be produced to obtain the desired end-product. This method gave us the major technical achievements of this century up to and including NASA’s putting a man on the moon. But except in very young technologies such as biogenetics, the technology-driven approach is becoming unproductive. We increasingly need a “business-driven” R&D strategy.
R&D in the American semiconductor firm is still largely technology-driven. The Japanese have overtaken us in large measure because they went into semiconductors with the question: What is the right business strategy? They concluded that research and manufacturing have to be integrated with a large user—i.e., a computer manufacturer—that provides a substantial captive market and thus a shield against the violent price fluctuations that are inherent in the industry’s economics. Though the Americans pioneered in robotics they are way behind today; they based their strategy on the potential of technology. ASEA based its R&D on business strategy—which led it to develop very different tools for very different markets.
The best example of a business-driven R&D strategy is still the earliest one, and it is American: the way David Sarnoff, the builder and long-time chief executive of RCA, created color television. In the mid-’40s, when black-and-white TV was just starting commercially, Mr. Sarnoff foresaw the color-TV-set market, thought through what the product would have to be to satisfy customers—in price, color fidelity, channel capacity, appearance, and size—and then worked out the science and technology required to produce such a set.
The resulting requirements were almost the exact opposite of what were then considered the promising technological directions, and most of RCA’s technical people considered them absurd. But Mr. Sarnoff persisted, put to work extremely small teams of highly competent people, and had color TV in 12 years.
The Japanese, quite consciously copying Mr. Sarnoff, later used the same strategy to develop the videocassette recorder; they started out with a business goal and a business strategy and then put very small teams of highly competent people to work on the new science and technology needed. The Americans who had first put pictures on tape were technology-driven; they ended up with industrial applications for small, not-very-profitable niche markets. The Japanese now own a multibillion-dollar VCR consumer business worldwide.
Serious though these developments are, they may be only symptoms. The very concept of the company lab is becoming questionable. It assumes that one material, product, or service is uniquely right for a given market.
Steelmakers, for instance, still assume that their product has natural markets that it is destined to dominate. There may be substitutes, but they are second-rate—what the Germans contemptuously call Ersatz. But increasingly there are competitive products offering better—or at least, different—performance in practically every single “natural” steel market—plastics, for instance, for the automobile.
Forty years ago, manual typewriters were being replaced by electric ones. But they were still typewriters, only with motors added, and produced by the traditional makers. Today word processors are taking over. They come out of computer technology and are being made by computer companies.
Sixty years ago, the Supreme Court decided that the telegraph was the “natural” way to transmit writing and the telephone the “natural” way to transmit the spoken word—and it divided the electronics universe accordingly. Today telephone, videotex (growing like wildfire in Europe), telex, facsimile, and electronic mail compete in offering electronic transmission. Each represents a different technology with a different industrial base.
In practically all businesses today, the R&D lab primarily is concerned with what goes on in its traditional technology, be it steel, paper, or typewriters. What increasingly is needed is awareness of, and concern with, science and technology outside of one’s own lab, outside of one’s own field, outside of one’s own industry.
Does the traditional lab still make much sense? It assumes that all the technology needed by the company can be produced by its own lab, and, conversely, that most everything its lab produces can be put to profitable use by the company. This is simply no longer true—to the point where one of the country’s leading research administrators, William Miller, the head of Stanford Research Institute, now says that “a first-rate lab is too productive to belong to any one company.”
A Free-Standing Business
Bell Labs was founded to produce all the technology the telephone would require—and did so for many years. Conversely, its parent company, AT&T, was expected to be able to use whatever new technology the lab would come up with: it did so for many years. But since World War II, many of the technological breakthroughs to come out of Bell Labs—e.g., the transistor—have found their main applications outside of telephony. And more and more of the new telephone technology is coming from outside the telecommunications industry and its labs.
Siemens’s one truly successful new product—the body scanner—was created not in its lab, but by a British producer of phonograph records.
The technological streams no longer run parallel. They increasingly cross each other, with frequent spillovers from one to the other.
Increasingly, therefore, the research lab may become a free-standing business, doing research work on contract for a multitude of industrial clients. Each client would then need a “technology manager” rather than a “research director”—someone who can develop business objectives based on the potential of technology and technology strategies based on business and market objectives, and who then defines and buys the technical work needed to produce business results. But no one today—and surely neither the engineering nor business schools—knows how to teach technology management nor, indeed, even where to start.
[1988]