Product development cycles have been accelerating in many industries throughout the 1980s. Management frequently responds by decentralizing RD yet critics outside the R&D community are rare. After all, who wants to be perceived as being against close customer relationships? Nevertheless, I am concerned that in pursuing the laudable aims of both research productivity and customer focus, decentralization of R&D in the name of the customer is being applied as a general solution across businesses and industries, without thorough questioning of whether it fits the individual technology strategy needs of a particular business. In Sweden, where there are many conglomerates of companies in multiple businesses, the decentralized R&D model is especially popular. When we see, as we have, a VP for technology of a major pharmaceutical company within a conglomerate who cannot make $20,000 internal R&D decisions without approval from a business unit head, there is cause for concern. I have seen the same effect in oil, chemicals, cars, and, increasingly, in electronics and telecommunications, both in Europe and the United States. TRADING FUTURE FOR PRESENT At its worst, placing R&D under business unit control may resemble placing the chicken coop under the control of the fox. For the R&D manager, being close to the customer means being organizationally under the thumb of the business unit manager. The business manager, in turn, is under pressure to produce earnings fast, in which activity R&D is of limited use. Business unit managers have very different incentives, motivations, and often personalities, from R&D managers. For example, commercial managers may terminate long-term or highly uncertain projects that they perceive as risky, even when they require relatively few resources. The R&D manager has an intuitive understanding that scientifically difficult projects frequently have the highest potential value, even though they require more time and are less certain. This clash of values can result in business managers refusing to fund early-stage research, and the departure of key scientists who prize their intellectual freedom, leaving a smaller, more docile R&D staff. This arrangement can function effectively during down-cycles, but when business picks up, the real risk to the business unit may be revealed as the absence of significant new products. It takes time to recover from this situation: high-quality R&D organizations are much easier to cut than to build, and if a sharp increase in R&D is demanded, it will take more than simply writing a larger cheque. Furthermore, an R&D organization that has been trimmed to the point where it lacks esprit de corps and is accustomed to taking orders will not produce results that satisfy anyone--least of all the customer. THE CUSTOMER'S ROLE Being close to the customer provides many benefits in terms of mutual understanding and shorter response times: As soon as a customer feels a need, a business can start addressing it. However, such relationships are essentially ones of incremental development, and rarely inspire breakthroughs, simply because the customer tends to have an evolutionary view of his needs and rarely supplies a visionary spark. His utility is largely confined to fault-finding with existing products, where he outlines opportunities rather than possibilities. For instance, people may find their hi-fi equipment bulky, but they do not know that they need a Walkman until it becomes available. Car companies rely on interviewing consumers and styling beauty contests to determine what customers want, without making allowances for consumers' strong preference for their existing vehicle, and their unwillingness to acknowledge that something different may be better. …