The Temporary Company
From the 1920s through the 1940s, the movie business was controlled by big studios like MGM and Columbia. The studios employed actors, directors, screenwriters, photographers, publicists, even projectionists—all the people needed to produce a movie, get it into theaters, and fill the seats. Central managers determined which films to make and who would work on them. The film industry was a model of bigcompany, industrial organization.
By the 1950s, however, the studio system had disintegrated. The power had shifted from the studio to the individual. Actors, directors, and screenwriters became freelancers, and they made their own choices about what projects to work on. For a movie to be made, these freelancers would join together into a temporary company, which would employ different specialists as needed from day to day. As soon as the film was completed, the temporary company would go out of existence, but the various players would, in time, join together in new combinations to work on new projects.
The shift in the film business from permanent companies to temporary companies shows how entire industries can evolve, quite rapidly, from centralized structures to network structures. And such transformations are by no means limited to the idiosyncratic world of Hollywood. Consider the way many manufacturers are today pursuing radical outsourcing strategies, letting external agents perform more of their traditional activities. The computer-display division of the Finnish company Nokia, for example, chose to enter the U.S. display market with only five employees. Technical support, logistics, sales, and marketing were all subcontracted to specialists around the country. The fashion accessories company Topsy Tail, which has revenues of $80 million but only three employees, never even touches its products through the entire supply chain. It contracts with various injection-molding companies to manufacture its goods; uses design agencies to create its packaging; and distributes and sells its products through a network of independent fulfillment houses, distributors, and sales reps. Nokia's and Topsy Tail's highly decentralized operations bear more resemblance to the network model of organization than to the traditional industrial model.
For another, broader example, look at what's happened to the textile industry in the Prato region of Italy. In the early 1970s, Massimo Menichetti inherited his family's business, a failing textile mill. Menichetti quickly broke up the firm into eight separate companies. He sold a major portion of equity—between one-third and one-half—to key employees, and he required that at least 50 percent of the new companies' sales come from customers that had not been served by the old company. Within three years, the eight new businesses had achieved a complete turnaround, attaining significant increases in machine utilization and productivity.
Following the Menichetti model, many other big mills in Prato broke themselves up into much smaller pieces. By 1990, more than 15,000 small textile firms, averaging fewer than five employees, were active in the region. The tiny firms built stateof-the-art factories and warehouses, and they developed cooperative ventures in such areas as purchasing, logistics, and R&D, where scale economies could be exploited. Textile production in the area tripled during this time, despite the fact that the textile industry was in decline throughout the rest of Europe. And the quality of the products produced in the Prato region rose as innovation flourished. Textiles from Prato have now become the preferred material for fashion designers around the world.
Playing a key role in the Prato textile industry are brokers, known as impannatori, who act as conduits between the small manufacturing concerns and the textile buyers. The impannatori help coordinate the design and manufacturing process by bringing together appropriate groups of businesses to meet the particular needs of a customer. They have even created an electronic market, which serves as a clearinghouse for information about projected factory utilization and upcoming requirements, allowing textile production capacity to be traded like a commodity.
The Prato experience shows that an economy can be built on the network model, but Prato, it could be argued, is a small and homogenous region. How would a complex, diverse industry operate under the network model? The answer is: far more easily than one might expect. As a thought experiment, let's take a journey forward in time, into the midst of the twenty-first century, and see how automobiles, the archetypal industrial product, are being designed.
General Motors, we find, has split apart into several dozen separate divisions, and these divisions have outsourced most of their traditional activities. They are now small companies concerned mainly with managing their brands and funding the development of new types and models of cars. A number of independent manufacturers perform fabrication and assembly on a contract basis for anyone who wants to pay for it. Vehicles are devised by freelance engineers and designers, who join together into small, ever-shifting coalitions to work on particular projects. A coalition may, for example, focus on engineering an electrical system or on designing a chassis, or it may concentrate on managing the integration of all of the subsystems into complete automobiles.
These design coalitions take many forms. Some are organized as joint ventures; some share equity among their members; some are built around electronic markets that set prices and wages. All are autonomous and self-organizing, and all depend on a universal, high-speed computer network—the descendant of the Internet—to connect them to one another and exchange electronic cash. A highly developed venture-capital infrastructure monitors and assesses the various teams and provides financing to the most promising ones.
In addition to being highly efficient, with little managerial or administrative overhead, this market-based structure has spurred innovation throughout the automotive industry. While much of the venture capital goes to support traditional design concepts, some is allocated to more speculative, even wild-eyed, ideas, which if successful could create enormous financial rewards. A small coalition of engineers may, for example, receive funds to design a factory for making individualized lighting systems for car grilles. If their idea pans out, they could all become multimillionaires overnight. And the next day, they might dissolve their coalition and head off to seek new colleagues and new challenges.
Over the past few years, under the auspices of the Massachusetts Institute of Technology's initiative on Inventing the Organizations of the 21st Century, we have worked with a group of business professors and executives to consider the different ways business might be organized in the next century. The automotive design scenario we've just laid out was discussed and refined by this group, and we subsequently shared it with managers and engineers from big car companies. They not only agreed that it was a plausible model for car design but also pointed out that the auto industry was in some ways already moving toward such a model. Many automakers have been outsourcing more and more of their basic design work, granting ever greater autonomy to external design agencies.
A shift to an e-lance economy would bring about fundamental changes in virtually every business function, not just in product design. Supply chains would become ad hoc structures, assembled to fit the needs of a particular project and disassembled when the project ended. Manufacturing capacity would be bought and sold in an open market, and independent, specialized manufacturing concerns would undertake small batch orders for a variety of brokers, design shops, and even consumers. Marketing would be performed in some cases by brokers, in other cases by small companies that would own brands and certify the quality of the merchandise sold under them. In still other cases, the ability of consumers to share product information on the Internet would render marketing obsolete; consumers would simply "swarm" around the best offerings. Financing would come less from retained earnings and big equity markets and more from venture capitalists and interested individuals. Small investors might trade shares in ad hoc, project-based enterprises over the Internet. Business would be transformed fundamentally. But nowhere would the changes be as great as in the function of management itself.
See Laubacher and Malone (1997a).