(FORTUNE Magazine) – LAWRENCE BOSSIDY, CEO of Allied-Signal, predicts ''organizational revolution'' for corporate America. Says David Nadler, president of Delta Consulting Group, who works with the chiefs of AT&T, Corning, and Xerox, among others: ''CEOs feel that companies need to be structured in dramatically different ways.'' In outfits as diverse as Eastman Kodak, Hallmark Cards, and General Electric -- even the San Diego Zoo -- the search for the organization perfectly designed for the 21st century is going ahead with the urgency of a scavenger hunt.

From many quarters we hear that hierarchical organization must wither away. In this view of the future middle managers have the life expectancy of fruit flies. Those who survive will not be straw bosses but Dutch uncles, dispensing resources and wisdom to an empowered labor force that designs its own jobs. Enabled, to use a trendy term, by information technology and propelled by the need to gain speed and shed unnecessary work, this flat, information-based organization won't look like the Pharaonic pyramid of yore but like -- well, like what? Like a symphony orchestra, Peter Drucker suggests. No, a jazz combo, some say. More like a spider web, others offer. Hamlet: Or like a whale? Polonius: Very like a whale. Gee, thanks. But where's my desk? What do I do eight hours a day -- or ten, or 12? Who gives me my annual review? When do we start? Good questions, which as yet have not had good answers. Says H. James Maxmim, the CEO of Laura Ashley Holdings: ''We're just beginning to explore the post-hierarchical organization. We don't know what it looks like yet.'' Some hints, however, are emerging. The 21st-century organization arises at the confluence of three streams. One is described by the term ''high-involvement workplace,'' meaning operations with self-managing teams and other devices for empowering employees. Novelties once, these participative mechanisms have proved they can consistently deliver jaw-dropping gains in productivity, quality, and job satisfaction. A second productivity turbocharger is a new emphasis on managing business processes -- materials handling, say -- rather than functional departments like purchasing and manufacturing. Third is the evolution of information technology to the point where knowledge, accountability, and results can be distributed rapidly anywhere in the organization. The trick is to put them together into a coherent, practical design. Then you have the company yours may become, and the one your sons and daughters will work for. At the end of this rainbow, say those who have peeked, is a whole kettleful of gold. Advises Bossidy, who until last summer was vice chairman of General Electric: ''Look at GE Appliances.'' In that $5.4-billion-a-year business, redesign has brought with it a $200 million drop in average inventory. McKinsey & Co. principal Douglas Smith, one of the blue-chip consulting firm's experts on organization, figures that a company applying the new principles of organization design can cut its cost base by a third or more. Smith bases his claim on results from companies that have already reorganized parts of their operations: an industrial goods manufacturer that cut costs and raised productivity more than 50%, a financial services company where costs fell 34%, and others. Results like that come from changing a company in profound ways, not just tinkering with the boxes on an organization chart. For years, Smith says, the basic question about how best to arrange people and jobs stayed the same: ''Do we centralize or decentralize -- and where do we stick international?'' The answer was never satisfactory. Companies were set up by product, or by customer, or by territory, and then switched when those arrangements stopped working. All that rejiggering missed the point, says Smith: ''It mattered only to the top people in the company. Below them you found the same functional, vertical organization. For the 90% of the people who serve customers and make product, all that changed was the boss's name.'' No longer. The Kodaks, GEs, and their ilk have first retailored the work people do, then management structures, with startling results. To make sense out of the rush of experimentation, McKinsey's Smith and his colleague Frank Ostroff are polishing a paper that lays out what Ostroff calls ''perhaps the first real, fundamentally different, robust alternative'' to the functional organization (see box). In the months since Ostroff released an early draft to % his consulting colleagues, it has proved the document most often requested inside the firm. There's nothing new about self-managing teams -- they were ''discovered'' 43 years ago at the bottom of a coal mine in Yorkshire by a researcher from the Tavistock Institute of Human Relations in London. Since then, forms of worker self-management have been adopted at countless sites. Marvin Weisbord, an expert on organizational development, notes that all rely on one basic idea: ''The people who do the work should have in their hands the means to change to suit the customer.'' That means workers should have the incentive and the power to respond to whoever buys their output -- at times someone else within their organization -- not just whoever cuts their paychecks. Weisbord adds that self-management typically delivers 40% increases in output per man-hour. TO SEE HOW it's done, skip the blackjack table next time you're in Puerto Rico and pay a visit to Bayamon, outside San Juan, where a new General Electric factory has been running for a year and a half. The place makes arresters, which are surge protectors that guard power stations and transmission lines against lightning strikes. Bayamon is the godchild of Philip Jarrosiak, manager of human resources for GE's capacitor and power protection operations. Once a minor-league infielder, Jarrosiak joined GE when he was 20, landing an hourly job making aircraft engines in Rutland, Vermont. In the 32 years since, he put himself through college at night and worked his way into management ranks, where he specializes in designing high-performance workplaces at both green-field and established sites. Bayamon is his newest and, Jarrosiak says, ''an opportunity to put in everything I know.'' The facility employs 172 hourly workers and just 15 salaried ''advisers,'' plus manager R. Clayton Crum. That's it: three layers, no supervisors, no staff. A conventional plant, Jarrosiak says, would have about twice as many salaried people. Every hourly worker is on a team with ten or so others; they meet weekly. Each team ''owns'' part of the work -- assembly, shipping and receiving, etc. But team members come from all areas of the plant, so that each group has representatives from both upstream and downstream operations. An adviser sits in the back of the room and speaks up only if the team needs help. What vaults Bayamon into the next century is the way it teaches its workers. Says Harvard professor Shoshanna Zuboff, author of In the Age of the Smart Machine: ''The 21st-century company has to promote and nurture the capacity to improve and to innovate. That idea has radical implications. It means learning becomes the axial principle of organizations. It replaces control as the fundamental job of management.'' Bayamon is a perpetual-learning machine. Hourly workers change jobs every six months, rotating through the factory's four main work areas. In six months they'll begin their second circuit of the plant, and everyone on the floor will know his job and how it affects the next person in line. The reward for learning is a triple-scoop compensation plan that pays for skill, knowledge, and business performance. The first time around, workers get a 25-cent-an-hour pay raise at each rotation; thereafter they can nearly double their pay by ''declaring a major,'' so to speak, and learning a skill like machine maintenance or quality control. More pay comes from passing courses in English, business practices, and other subjects. Toss in bonuses -- $225 a quarter or more -- for meeting plantwide performance goals and having perfect attendance. Promotions and layoffs will be decided by skill level, not seniority. In just a year the work force became 20% more productive than its nearest company equivalent on the mainland, and Jarrosiak predicts productivity will rise 20% more by the end of 1993. FOR YEARS plants like Bayamon existed barely connected to the organizations of which they are a part. Some Procter & Gamble factories were worker run as long ago as 1968, a fact concealed from competitors -- and sometimes from headquarters. The Gaines pet food plant in Topeka, Kansas, just celebrated 20 years of self-management. For two decades, under three owners -- Anderson Clayton, General Foods, and Quaker Oats -- Topeka has always placed first when its labor productivity was compared with that of other pet food plants within its company. According to Herman Simon, plant manager for 17 years, higher-ups who saw the numbers vowed never to mess with the plant. But they rarely went away determined to make their other factories over in its image. Says a frustrated William Buehler, senior vice president at Xerox: ''You can see a high-performance factory or office, but it just doesn't spread. I don't know why.'' One reason is that nervous executives experiment where failure won't be fatal, and thereby contain the gains too. Says Jarrosiak: ''I hate pilot programs off in a corner of a plant. You need commitment.'' < You also need to be able to envision how such operations fit into a large- scale enterprise. Says McKinsey's Ostroff: ''Executives know what teams can do. But they need a picture that links the high-performance team to the whole organization and multiplies the gains.'' It's relatively easy to oversee one of these operations when it's confined within one function, like manufacturing. For self-directed management to spread, a company must lay goals, responsibilities, and measurements across functions. Ostroff argues: ''Senior managers need to be able to say 'empowerment' and 'accountability' in the same sentence.'' BUSINESS processes -- almost sure to become a term you will hear lots of -- can form the link between high-performance work teams and the corporation at large. Organizing around processes, as opposed to functions, permits greater self-management and allows companies to dismantle unneeded supervisory structures. It's a management axiom that crab grass grows in the cracks between departments. Purchasing buys parts cheap, but manufacturing needs them strong. Shipping moves goods in bulk, but sales promised them fast. ''I call it Palermo's law,'' says Richard Palermo, a vice president for quality and transition at Xerox. ''If a problem has been bothering your company and your customers for years and won't yield, that problem is the result of a cross- functional dispute, where nobody has total control of the whole process.'' And here's Palermo's corollary: People who work in different functions hate each other. Upon this fratricidal scene, enter the process doctor. Depending on which consulting firm he's coming from, he may describe his work as ''reengineering'' or ''core process redesign'' or ''process innovation.'' Michael Hammer, a consultant in Cambridge, Massachusetts, defines, though not exactly lyrically, what the doctor is up to: ''Reengineering is the fundamental analysis and radical redesign of business processes to achieve dramatic improvements in critical measures of performance.'' Process management differs from managing a function in three ways. First, it uses external objectives. Old-line manufacturing departments, for example, tend to be measured on unit costs, an intradepartmental number that can lead to overlong production runs and stacks of unsold goods. By contrast, an integrated manufacturing and shipping process might be rated by how often it turns over its inventory -- a processwide measurement that reveals how all are working together to keep costs down. Second, in process management employees with different skills are grouped to accomplish a complete piece of work. Mortgage loan officer, title searcher, and credit checker sit and work together, not in series. Third, information moves straight to where it's needed, unfiltered by a hierarchy. If you have a problem with people upstream from you, you deal with them directly, rather than asking your boss to talk to theirs. Reengineered processes have been in place at Kodak for more than two years. The 1,500 employees who make black and white film -- inevitably called Zebras -- work not in departments but in what's called ''the flow.'' (Black and white is big business: about $2 billion a year from sales of 7,000 products used in printing, X-rays, even spy satellites.) Headed by Richard Malloy, a 25-member leadership team watches the flow. They measure it with end-of-process tallies like productivity. Within the flow are streams defined by ''customers'' -- Kodak business units -- and scored on customer satisfaction measures such as on-time delivery. One stream, for example, is charged with making hundreds of types of film for the Health Sciences Division and works closely with it to schedule production and to develop new products, a Zebra specialty. In the streams most employees work in self-directed teams. A few functions -- accounting and human resources -- remain outside the streams. When the flow began in 1989, the black and white film operation was running 15% over budgeted cost, took up to 42 days to fill an order, was late a third of the time, and scored worst in Kodak's morale surveys. Last year the group came in 15% under budgeted cost, had cut response time in half, was late one time out of 20, and wore the biggest smiles in Rochester, New York. Why? Says Zebra Robert Brookhouse: ''When you create a flow and a flow chart, you find where you're wasting time, doing things twice. And because we own our entire process, we can change it.'' ORGANIZING AROUND a process seems to yield sterling results as consistently as high-involvement factories do. Privately held Hallmark (1991 sales: $2.9 billion) expects big gains now that Steven Stanton of CSC Index, a Cambridge, Massachusetts, consulting firm, has helped the company reengineer its new- product process. The greeting-card maker lives or dies on new stuff -- some 40,000 cards and other items a year, the work of 700 writers, artists, and , designers on what Hallmark boasts is the world's largest creative staff. The process of developing a new card had become grotesque; it took two years -- longer than the road from Gettysburg to Appomattox Court House. The company was choking on sketches, approvals, cost estimates, and proofs. Says Hallmark's Don Fletcher: ''We needed a lot of people just to check items in and out of departments.'' Fletcher's title, vice president for business process redesign, pretty much tells what happened. Starting this spring, about half the staff will be put to work on cards for particular holidays like Valentine's Day or Christmas. The birthday and get-well card folks will follow. A team of artists, writers, lithographers, merchandisers, bean counters, and so on will be assigned to each holiday. Team members are moving from all over a two-million-square-foot office building in Kansas City so they can sit together. Like a canoe on a lake, a card will flow directly from one part of the process to the next within, say, the Mother's Day team; before, it had to be portaged from one vast department to the next. This should cut cycle time in half, which will not only save money but will also make the company more responsive to changing tastes. Hallmark hasn't eradicated departments. There will be ''centers of excellence'' to which workers will return between projects for training and brief, special stints, a bit like homerooms in high school. For now, department heads remain the senior managers of the business. But the head of graphic arts, which makes separations and proofs, has told Fletcher that he hopes the department infrastructure will eventually dissolve in the flow. That's the right idea, say hard-core process managers. If you reengineer a process, pocket a one-time gain, and return to your desk, says McKinsey's Smith, ''the barnacles you scrape off will just grow back.'' The way to keep them off, says Hammer, is to obliterate the functions: ''In the future, executive positions will not be defined in terms of collections of people, like head of the sales department, but in terms of process, like senior-VP-of- getting-stu ff-to-customers, which is sales, shipping, billing. You'll no longer have a box on an organization chart. You'll own part of a process map.'' Can a whole company literally lie on its side and organize horizontally, by process? You got it, says Allied-Signal's Bossidy: ''Every business has maybe six basic processes. We'll organize around them. The people who run them will , be the leaders of the business.'' An industrial company might select processes like new-product development, flow of materials (purchasing, receiving, manufacturing), and the order- delivery billing cycle. Into these process flows will go management teams to tend subprocesses and teams of workers to carry out tasks. Whoever is needed will be there: The materials-flow group might have finance folks but no marketers -- but the marketers will be plentiful in the new-product process. There are no departments in Bossidy's 21st-century corporation: ''You might have a CFO, but he won't have many people who report to him.'' If metallurgists and actuaries are taken out of departments and clumped around processes, what happens to their specialized skills? A minor problem, argues James Champy, CEO of CSC Index: ''State-of-the-art knowledge comes from a small group of people. Most people in a function don't contribute expertise. They execute.'' Put the innovators in a stafflike or lablike group. Create a house Yellow Pages so functional expertise is easy to find even though dispersed. Link experts in a real or electronic network where they can keep each other up to date and can get training and career development help. ''That's okay,'' says Bossidy. ''The engineers can have a club. But they can't work in the same room, and they can't sit at the same table at the company banquet.'' His vision is somewhat radical, he admits, understating the case. ''So corporations will first try to make the matrix work. Boy, that will drive employees and managers nuts.'' One trouble with breaking down the walls: In most companies functional and hierarchical walls are load bearing. Remove them and the roof caves in. A big burden they bear is to collect, evaluate, and pass on information. Another is to determine employees' career paths -- to define ambition, reward, and sycophancy. In a flat shop of teams and processes, both information flow and careers will have to be different. Walk around futuristic companies and you see odd sights: suppliers who work in their customers' offices; widely available, easy-to-read charts tracking scrap, on-time delivery, and other data that rivals would kill for; hourly workers logged onto PCs, reading their E-mail. They're all part of an effort to put information where it can be used at the moment it's needed. Says Delta Consulting's David Nadler: ''In the organization of the future, information technology will be a load-bearing material -- as hierarchy is now. You can't | have self-management without it.'' That is, computer networks and the information they carry will help define your corporate structure. Let information flow wherever it's needed, and a horizontal self-managed company is not only possible, it's inescapable. Building computer highways that can transport cost and other data sideways within a process, as well as vertically to top management, is a step in this direction. Other steps include training that teaches workers how their actions affect overall business performance and measurements that direct tasks at optimum outcomes, such as rewarding salespeople for gross margin, not gross sales. You have to transport power as well as knowledge. In a hierarchy, rank defines authority: A manager can okay deals up to $50,000, his boss to $100,000, her boss to $250,000 . . . That's obsolete, in Harvard business school professor Quinn Mills's view. The question isn't how high the money gets; it's how high your customer's blood pressure gets. ''Does he need an answer immediately? Do you have to be able to be flexible? If so, you have to empower the person who talks to the customer.'' If you can't entrust such matters to the folks in the field, maybe you should switch places with them. They can have your desk, where the decisions, obviously, are less important. What happens to the career ladder? CSC Index Chief Executive Champy suggests that law firms, with only three levels of hierarchy -- associate, partner, and senior partner -- might provide the very model of a modern career path. Says Champy: ''A lawyer's career is a progression to more complex work -- tougher cases, more important clients. Titles don't change, but everyone knows who has the highest status.'' The oldest art in organization design -- carving out strategic business units -- will still matter in this new world. The goal, as Nadler sees it, is to create ''enterprises with clear customers, markets, and measures, and few internal boundaries.'' That means letting sets of customers or customer needs define business units, and grouping into businesses the people and processes necessary to serve them. THAT'S HOW Xerox designed its new horizontal organization. Till this year, Xerox was set up in the usual functions -- R&D, manufacturing, sales, and the like. The new design creates nine businesses aimed at markets such as small businesses and individuals, office document systems, and engineering systems. Each business will have an income statement and a balance sheet, and an identifiable set of competitors. New manufacturing layouts will permit so- called focused factories dedicated to specific businesses. Most of the businesses will sell through a new Customer Operations Group, a mingling of sales, shipping, installation, service, and billing, created so customers can keep just one phone number on their Rolodexes. In fact, the businesses will sell to Customer Operations -- that is, negotiate contracts -- so that market forces extend deeply into the company. Teams lead the businesses, whose building blocks are what CEO Paul Allaire calls ''microenterprise units'': complete work processes or subprocesses. Says Allaire: ''We've given everyone in the company a direct line of sight to the customer.'' In a functional hierarchy, job descriptions, career paths, and information flow are all geared toward control -- of work, workers, and knowledge. Compare that with the evolving 21st-century company, where work is lined up with customers, not toward bosses. Senior executives have charge of the handful of processes that are critical to satisfying customers. Self-directed, the work force does most of the hiring, scheduling, and other managerial tasks that once ate up kazillions in indirect labor costs. The few people left between the executives and the work teams spend their time trying to change the organization, not to control it: They are reaching out to grab a new technology or a new customer, or to respond to a new demand from an old one. Jobs, careers, and knowledge shift constantly. The boundaries of the company will be fluid too. The growing number of strategic alliances suggests as much. So do the actions of companies like Wal- Mart Stores and Procter & Gamble, which have interwoven their order-and- fulfillment process so that the bells of Wal-Mart's cash registers in effect ring in P&G warehouses, telling them to ship a new box of Tide to replace the one you just bought. In the view of Harvard economist Robert Reich, the boundaries will become so fluid that corporations will become temporary arrangements among entrepreneurial cadres. Except for high-volume, capital-intensive work, says Reich, ''every big company will be a confederation of small ones. All small organizations will be constantly in the process of linking up into big ones.'' That may be more fluidity than most people can accept, at least as long as mortgage applications ask, ''How long have you been with your current / employer?'' But the new flexible organization will be a powerful competitor. Smith finds a metaphor in Terminator II, the movie where Arnold Schwarzenegger faces a metal monster that liquefies, then hardens again in a new shape -- now a man, now a machine, now a knife. Says Smith: ''I call it the Terminator II company.'' How'd you like to have to compete with one of those?

CHART: NOT AVAILABLE CREDIT: TONY MIKOLAJCZYK FOR FORTUNE SOURCE: MCKINSEY & CO. CAPTION: A new view of organization by McKinsey consultants Frank Ostroff and Doug Smith is meant to help clients hung up by the old template. Says Ostroff: ''They needed a clear architecture'' to show how a functional pyramid (top) could become a process-oriented, horizontal organization. FROM THE VERTICAL ORGANIZATION. . . . . .TO THE HORIZONTAL ORGANIZATION