CHAPTER THIRTY-SEVEN

The 10 Rules of Effective Research

SOME BUSINESSES—NOT VERY MANY—get a fiftyfold, or even a hundredfold, return on the research dollar. Many more get little or nothing. The key to success is not knowledge, intelligence, or hard work—and least of all, luck. It is following the 10 Rules of Effective Research.

1. Every new product, processor, or service begins to become obsolete on the day it first breaks even.

2. Thus, your being the one who makes your product, process, or service obsolete is the only way to prevent your competitor from doing so.

One major American company that has long understood and accepted this is Du Pont Co. When nylon came out 50 years ago, Du Pont immediately put chemists to work to invent new synthetic fibers to compete with nylon. It also began to cut nylon’s price—thus making it less attractive for would-be competitors to find a way around Du Pont’s patents. This explains both why Du Pont is still the world’s leading synthetic-fiber maker, and why Du Pont’s nylon is still in the market, and profitably so.

A Meaningless Distinction

3. If research is to have results, the nineteenth-century distinction between “pure” and “applied” research better be forgotten. It may still work in the university, but in industry it is meaningless, if not an impediment. A minor change in machining a small part, for example, may require pure research into the structure of matter. Yet creating a totally new product or process may involve only careful rereading of a standard handbook. Nor is pure research necessarily more difficult than redefining a problem so that well-known concepts can be applied to its solution.

4. In effective research, physics, chemistry, biology, mathematics, economics, and so on are not “disciplines.” They are tools. This does not mean, of course, that effective research requires universal geniuses. The most brilliant physicist or chemist today knows only a small corner of his own discipline. But effective research demands that the project leader or research director know when and how to call on what specialist. The best example may be the way in which Jim Webb, President Kennedy’s head of NASA in the 1960s, mobilized a dozen different disciplines to put a man on the moon. Mr. Webb was not a scientist but a lawyer-accountant.

5. Research is not one effort—it is three: improvement, managed evolution, and innovation. They are complementary but quite different.

• Improvement aims at making the already successful better still. It is a never-ending activity that requires specific quantitative goals, such as annual improvements of 3 percent or 5 percent in cost, quality, and customer satisfaction. Improvement starts with feedback from the front line: people who actually make the product or deliver the service; sales people; and, vitally important, the users. Then the company’s own scientists, engineers, or product designers must convert the front line’s suggestions and queries into changes in product, process, or service.

  The best-known practitioners of continuing improvement today are the Japanese. Its inventor and most consistent practitioner however was an American company, the Western Electric subsidiary of the old Bell Telephone System.

• Managed evolution is the use of a new product, process, or service to spawn an even newer product, process or service. Its motto is “each successful new product is the stepping stone to the next one.”

  The best-known practitioner is probably Sony, which has systematically evolved a dozen new products—the Walkman, for instance—out of the original tape recorder. But the most successful practitioner is probably a “no-tech” American business, ServiceMaster Co. of suburban Chicago, a multibillion-dollar multinational doing business successfully in the U.S., Japan, and Western Europe. ServiceMaster started with the systematic application of industrial engineering to hospital maintenance and the training of low-skill people. It then evolved this, step by step, into office maintenance and the care of elderly shut-ins. Managed evolution is always market driven; it often requires, however, new, or at least newly developed, technology and tools.

• Innovation, finally, is the systematic use of opportunity of changes: in society and the economy, in demographics, and in technology.

The key to effective research is to pursue improvement, managed exploitation, and innovation simultaneously though separately. The classic example is again Du Pont’s strategy in synthetic fibers. As I mentioned earlier, the company immediately began work on inventing competing fibers. But it also immediately started to improve nylon and to pursue managed evolution. Nylon was developed for women’s stockings. But soon it was modified to serve as automotive tire cord—probably the most profitable application for many years.

The first five rules are about what to do. The last five lay down how to do it.

6. Aim high! Trivial corrections usually are as hard to make and as staunchly resisted as fundamental changes. Successful research asks: If we succeed, will it make a real difference in the customer’s life or business? The Japanese control the market in videotape recorders and fax machines, both American inventions, because they set higher research goals than any American company thought attainable—in terms of product size, performance, and price.

7. Yet, effective research requires both long-range and short-range results. The efforts needed are too great to be satisfied with the short term alone. A short-term result must also be a step in a continuing long-term process. The needed balance is difficult to design. But it usually can be attained by retrospective analysis.

Researchers have long known that they should go back and read their own lab notes. Did anything happen that was pushed aside because it was unexpected or because it did not seem to lead toward the desired research objective? If so, was it actually an indication of an opportunity? Above all, was it an indication of a usable and useful short-term result? The best-known example is how Alexander Fleming came to realize that he had stumbled upon penicillin but had thrown it away as spoiling his bacterial cultures.

In improvement, where results by definition are short-term, one looks for the long-term implications. One analyzes the work of the past two or three years with this question: Did successful improvements cluster around one particular application, one particular market, one design, one process, one product? This often indicates an opportunity for fundamental, long-term innovation or change.

One large company that seems to have mastered this balancing act is Merck, one of the world’s largest pharmaceutical companies. Another is the medical-electronics business of GE. While working on such radical innovations as body scanning and nuclear magnetic resonance imaging, it has systematically fed back from these major long-term innovations to make constant, immediate improvements in its conventional X-ray apparatus.

8. Research is separate work, but it is not a separate function. Development—the translation of research results into products, processes, and services that can be manufactured, sold, delivered, and serviced—must go hand in hand with research. And manufacturing, marketing, and service all affect research from the beginning, just as much as the results of research in turn affect them. In the university, research may be the search for new knowledge as an end in itself. In industry, in government, and in medicine, research is the search for new utility.

9. Effective research requires organized abandonment—not only of products, processes and services, but also of research projects. Every product, process, service, and research project needs to be put on trial for its life every few years, with this question: Would we now start this product, process, service, or research project, knowing what we know now?

Three good clues to when to abandon:

First, when there are no more significant improvements. Second, when new products, processes, markets or applications no longer come out of managed evolution. Third, when long years of research produce only “interesting” results.

Reviewing Innovations

10. Research has to be measured like everything else. For improvements, it is fairly easy to set specific goals and to measure them. In managed evolution, too, goals can be set—e.g., one new significant product, market, or application every year. Innovation, however, requires appraisal. Every three years or so a company needs to review its innovative results. What did we innovate that made a difference in the wealth-producing capacity of this company? Were these innovations commensurate in numbers, quality, and impact with our market standing and our leadership position in our industry? What will our innovation results have to be in the next few years—again, in numbers, quality and impact—to give us the market standing and industry leadership we need?

Research expenditures in American business—flat or even declining the past few years—are starting to climb again. But spending money does not by itself guarantee results. Applying the 10 Rules of Effective Research does.

[1989]