Since 1990, when Michael Hammer published re-engineering's manifesto in the Harvard Business Review - in an article entitled "Don't Automate, Obliterate" - nearly all big companies, and most small ones, have embarked on at least one re-engineering project. By most estimates, about half of these efforts failed to reach their goals. Here lies the re-engineering dilemma. While it is obvious to any manager that new technology enables worked to be organized very differently from the way it is organized today, there are as yet no hard and fast rules as to how the task should be approached, only compelling examples and a widespread sense of urgency.
Many executives will want to imitate Mervyn's, a middle-sized, middle-market chain of about 300 department stores spread across the USA. It added tens of millions of dollars a year to profits after spending nine-months re-engineering its inventory-control systems. By using computers to track sales more closely and to respond more quickly, it managed to both reduce inventory dramatically and to halve the frequency with which the goods are out of stock when a customer asks for them.
Mervyn's is not alone. Banks, car-makers, insurance companies, retailers and a host of others can boast about inspiringly successful instances of re-engineering. But the big picture tells a different story. Despite the billions spent on putting computers on managers' desktops, growth in the overall productivity of the white-collar workers who use the machines has generally been disappointing. Underlying every re-engineering effort is the larger process of competitive markets groping towards a new equilibrium for economics dominated by trading in ideas rather than things, and powered by computers rather than motors.
Two factors are driving re-engineering. First, as most executives have already realized, computers and networks enable work to be done radically different from how it was done before. Second, as they are now coming to realize, information, the raw material of today's service economics, is a different commodity, with different properties, from the iron and coal that fueled the industrial revolution. These two factors rock the very foundations of industrial organization.
One of the simplest and most profound consequences of the new realities of capitalism is a shift from sequential processes to simultaneous ones. Information on a computer can be copied instantly and costlessly, so many people can work on the same thing at once, instead of passing it, step by time-consuming step, from one to another. This simple realization is the source of some of the most impressive savings of re-engineering. IBM credit, for example, used this insight to reduce the time needed to approve financial terms for the purchase of big computers - and thus close the deal - from over a week to under four hours. Insurance companies have cut claims-processing times from weeks to minutes. But the new ways of working are not just faster, they are also fundamentally different.
Technology is breaking down the structures that separate companies from their customers. One of the first waves of disintermediation came in the 1970s, when people began getting their cash direct from a hole in the wall instead of a teller. Then car-makers began placing orders directly into their suppliers production-planning computers rather than picking from a warehouse, and Wal-Mart began stocking its shelves with Proctor & Gamble goods directly from P & G's inventory, managed by P&G's computers, instead of via its own warehouse.
Further disintermediation will come with the growth of computer networks, like the Internet, which enable consumers to shop electronically both cheaply and easily. Companies that distribute over the Internet, like the Internet Shopping Network, which in 1994 was purchased by television's Home Shopping Network, have two big advantages over conventional mail-order. They need neither to print catalogues, because would-be customers browse their databases electronically, nor to employ banks of people to answer telephones, because orders are placed directly into the computers that fill them. Soon such disintermediation could affect everybody who makes their living from answering a telephone, from stockbrokers to warehouse staff. With changes come the necessity of learning new ways of working.
