CHAPTER EIGHTEEN

… And How They Join Together

IDENTIFYING KEY ACTIVITIES AND analyzing their contributions defines the building blocks of organization. But to place the structural units which make up the organization requires two additional pieces of work: an analysis of decisions and an analysis of relations.

Decision Analysis

What decisions are needed to obtain the performance necessary to attain objectives? What kinds of decisions are they? On what level of the organization should they be made? What activities are involved in, or affected by, them? Which managers must therefore participate in the decisions—at least to the extent of being consulted beforehand? Which managers must be informed after they have been made? The answers to these questions very largely determine where certain work belongs. It will be argued that it is impossible to anticipate what kinds of decisions will arise in the future. But while their content cannot be predicted their kind and subject matter have a high degree of predictability.

In one large company well over 90 percent of the decisions that managers had to make over a five-year period were found to be “typical,” and fell within a small number of categories. In only a few cases would it have been necessary to ask, “Where does this decision belong?” had the problem been thought through in advance. Yet, because there was no decision analysis, almost three-quarters of the decisions had to “go looking for a home,” as the graphic phrase within the company put it, and most of them went to a much higher level of management than was needed. The company’s components had been placed according to the size of their payroll rather than according to their decision responsibility so that the activities that should have made key decisions were placed so low as to be without authority and also without adequate information.

To place authority and responsibility for various kinds of decisions requires first that they be classified according to kind and character. Such standard classification as “policy decisions” and “operating decisions” are practically meaningless, however, and give rise to endless debates of a highly abstruse nature. Not much more helpful is classification according to the amount of money involved.

There are four basic characteristics which determine the nature of any business decision.

First, there is the degree of futurity in the decision. For how long into the future does it commit the company? And how fast can it be reversed?

The buyers at Sears, Roebuck have practically no limit as to the amount to which they can commit the company. But no buyer or buying supervisor can either abandon an existing product or add a new one without the approval of the head of the entire buying operation who, traditionally, is the number two or number three executive in the entire Sears, Roebuck organization.

The second criterion is the impact a decision has on other functions, on other areas, or on the business as a whole. If it affects only one function, it is of the lowest order. Otherwise it will have to be made on a higher level, where the impact on all affected functions can be considered; or it must be made in close consultation with the managers of the other affected functions. To use technical language, “Optimization” of process and performance of one function or area must not be at the expense of other functions or areas; this is undesirable “suboptimization.”

One example of a decision which looks like a purely “technical” one affecting one area only, but which actually has impact on many areas, is a change in the methods of keeping the parts inventory in a mass-production plant. This affects all manufacturing operations. It makes necessary major changes in assembly. It affects delivery to customers—it might even lead to radical changes in marketing and pricing, such as the abandonment of certain designs and models and of certain premium prices. And it may require substantial changes in engineering design. The technical problems in inventory-keeping—though quite considerable—pale into insignificance compared to the problems in other areas which any change in inventory-keeping will produce. To “optimize” inventory-keeping at the expense of these other areas cannot be allowed. It can be avoided only if the decision is recognized as belonging to a fairly high order and handled as one affecting the entire process: either it has to be reserved for management higher than the plant; or it requires close consultation among all functional managers.

The consideration of the impact of a decision and the need to prevent “suboptimization” may shift the focus of a decision decisively, as the following example shows.

In the early days of the Du Pont Company, when it was still solely an explosives manufacturer, the company was by far the world’s largest buyer of nitrate, without, however, owning any nitrate fields. Yet the purchasing department was given a completely free hand in buying nitrate. It did so, indeed, most successfully—from the point of view of purchasing. It bought nitrate when the market prices were low and succeeded in obtaining the vital raw material for the company at prices far below what the competitors usually had to pay. Yet this was suboptimization. For the low prices for nitrate and the resulting competitive cost advantage were paid for by tying up large sums of money in inventory. This, in the first place, meant that a good deal of the cost advantage of low nitrate prices was illusory and offset by high interest payments. More serious, it also meant that the company, in the event of a downturn in business, might find itself in a liquidity crisis. The decision to balance cheap raw material prices against the cost of money and the danger of illiquidity was therefore properly made as a top-management decision. But after the new limits for inventory had been established, the buying decisions again became exclusively the task of the purchasing people.

The character of a decision is also determined by the number of qualitative factors that enter into it: basic principles of conduct, ethical values, social and political beliefs, etc. The moment value considerations have to be taken into account, the decision moves into a higher order and requires either determination or review at a higher level. And the most important as well as the most common of qualitative factors are human beings.

Finally, decisions can be classified according to whether they are periodically recurrent or rare, if not unique, decisions. The recurrent decision requires the establishment of a general rule, that is, of a decision in principle. Since suspending an employee deals with a person, the rule has to be decided at a fairly high level in the organization. But the application of the rule to the specific case, while also a decision, can then be placed on a much lower level.

The rare decision, however, has to be treated as a distinct event. Whenever it occurs, it has to be thought through.

A decision should always be made at the lowest possible level and as close to the scene of action as possible. However, a decision should always be made at a level insuring that all activities and objectives affected are fully considered. The first rule tells us how far down a decision should be made. The second how far down it can be made, as well as which managers must share in the decision and which must be informed of it. The two together tell us where certain activities should be placed. Managers should be high enough to have the authority needed to make the typical decisions pertaining to their work, and low enough to have the detailed knowledge and the first-hand experience, “where the action is.”

Relations Analysis

The final step in designing the building blocks of organization is an analysis of relations. It tells us where a specific component belongs.

With whom will a manager in charge of an activity have to work, what contribution does that manager have to make to managers in charge of other activities, and what contribution do these other managers have to make in return?

The basic rule in placing an activity within the organization structure is to impose on it the smallest possible number of relationships. At the same time, it should be so placed that the crucial relations, that is, the relationship on which depend its success and the effectiveness of its contribution, should be easy, accessible, and central to the unit. The rule is to keep relationships to a minimum but make each count.

This rule explains why functions are not, as traditional organization theory would have them be, “bundles of related skills.” If we followed that logic, we would, for instance, put production planning into a planning component in which all kinds of planners would work together. The skills needed in production planning are closely related to all other operational planning skills. Instead we put the production planner into manufacturing and as close as possible both to the plant manager and to the first-line supervisors.

There is often a conflict between placement according to decision analysis and placement according to relations analysis. By and large, one should try to follow the logic of relations as far as possible.

If organization design has to follow the logic of decisions in order to avoid suboptimization (as is usually the case with respect to the accounting function) the work itself should be planned according to relations analysis, that is, as close as possible to the scene of action. The direction of the work, the setting of rules, of standards, but also the appraisal and evaluation of the work should be placed according to decision analysis in a central component which can see the entire business and think through the impacts.

The four analyses—of key activities, of contributions, of decisions, of relations—should always be kept as simple and as brief as possible. In a small enterprise they can often be done in a matter of hours and on a few pieces of paper. In a very large and complex enterprise, though, such as General Electric, the First National City Bank, or Unilever (not to mention the Department of Defense) the job may well require months of study and the application of highly advanced tools of analysis and synthesis. But these analyses should never be slighted or skimped. They should be considered a necessary task and one that has to be done well in every business.

Symptoms of Malorganization

There is no perfect organization. At its best an organization structure will not cause trouble. But what are the most common mistakes in designing the building blocks of organization and joining them together? And what are the most common and the most serious flaws in organization?

The most common and the most serious symptom of malorganization is multiplication of the number of management levels. A basic rule of organization is to build the least possible number of management levels and forge the shortest possible chain of command.

Every additional level makes more difficult the attainment of common direction and mutual understanding. Every additional level distorts objectives and misdirects attention. Mathematical “information theory” has a law that any additional relay in a communications system halves the “message” and doubles the “noise.” Any “level” in an organization is a “relay.” Every link in the chain sets up additional stresses and creates one more source of inertia, friction, and slack.

Every additional level, especially in the big business, adds to the difficulty of developing tomorrow’s managers, both by adding to the time it takes to come up from the bottom and by making specialists rather than managers out of the people moving up through the chain.

In some large companies there are today twelve or even fifteen levels between first-line supervisor and company president. Assuming that someone gets appointed first-line supervisor at age twenty-five, and then spends only five years on each intervening level—both optimistic assumptions—that person could not be considered for the company’s presidency until age eighty or ninety. And the usual “cure”—a special promotion ladder for hand-picked young “geniuses” or “crown princes”—is as bad as the disease.

How few levels are really needed is shown by the example of the oldest, largest, and most successful organization of the West, the Catholic Church. There is only one level of authority and responsibility between the Pope and the lowliest parish priest: the bishop.

The second most common symptom of malorganization is recurrence of organizational problems. No sooner has a problem supposedly been “solved” than it comes back again in a new guise.

A typical example in a manufacturing company is product development. The marketing people think it belongs to them, the research and development people are equally convinced that it belongs to them. But placing it in either component simply creates a recurring problem. Actually both placements are wrong. In a business that wants innovation, product development is a key activity and a revenue-producing activity. It should not be subordinated to any other activity. It deserves to be organized as a separate innovative component.

The recurrent organization problem indicates unthinking application of traditional “organization principles” such as that of the “typical function” or that of “staff and line.” The answer lies in making the right analyses—the key activities analysis, the contributions analysis, the decisions analysis, and the relations analysis. An organization problem that comes back more than a couple of times should not be treated mechanically by shuffling little boxes on a piece of paper. It indicates lack of thinking, lack of clarity, and lack of understanding.

Equally common and equally dangerous is an organization structure that puts the attention of key people on the wrong, the irrelevant, the secondary problems. Organization should put the attention of people on major business decisions, on key activities, and on performance and results. If, instead, it puts attention on proper behavior, on etiquette, on procedure, let alone on jurisdictional conflict, organization misdirects. Then organization becomes a bar to performance.

Again, this is the result of mechanical rather than organic organization building. It is the result of slapping on so-called principles, instead of thinking through what organization the strategy of the business demands. It is the result of focusing organization on symmetry rather than on performance.

No organization chart is likely ever to be displayed in a major art museum. What matters is not the chart but the organization. A chart is nothing but an oversimplification which enables people to make sure that they talk about the same things in discussing organization. One never makes organizational changes for the sake of the chart. This always results in malorganization.

There are a number of common symptoms of poor organization which, usually, require no further diagnosis. There is, first, the symptom of too many meetings attended by too many people.

There are, especially in large organizations, managerial organs which do their work in and by meetings. The top committees in General Motors are examples. And so are the boards of directors composed of the top officers which govern both Standard Oil of New Jersey and Du Pont. But these are exceptions— deliberative organs which do not have operating functions and, as a rule, do not have decision-making functions either. They are organs to guide, to reflect, to review—and perhaps their most important function is to compel the operating top managers who sit down with the committee to think through their own direction, their own needs, and their own opportunities.

But apart from such deliberative bodies, which discharge their functions in meetings, meetings should be considered as a concession to organizational imperfection. The ideal is the organization which can operate without meetings—in the same sense in which the ideal of the machine designer is to have only one moving part in his contraption. In every human organization there is far too much need for cooperation, coordination, and human relations to have to provide for additional meetings. And the human dynamics of meetings are so complex as to make them very poor tools for getting any work done.

Whenever executives, except at the very top level, spend more than a fairly small fraction of their time—maybe a quarter or less—in meetings, there is evidence of a case of malorganization. An excess of meetings indicates that jobs have not been defined clearly, have not been structured big enough, have not been made truly responsible. Also the need for meetings indicates that the decisions and relations analyses have not been made at all or have not been applied. The rule should be to minimize the need for people to get together to accomplish anything.

An organization in which people are all the time concerned about feelings and about what other people will like is not an organization that has good human relations. On the contrary, it is an organization that has very poor human relations. Good human relations, like good manners, are taken for granted. Constant anxiety over other people’s feelings is the worst kind of human relations.

An organization that suffers from this—and a great many do—can be said unequivocally to suffer from overstaffing. It might be overstaffed in terms of activities. Instead of focusing on key activities, it tries to do a little bit of everything—especially in advice and teaching activities. Or the individual activities are overstaffed. It is in crowded rooms that people get on each other’s nerves, poke their elbows into each other’s eyes, and step on each other’s toes. Where there is enough distance they do not collide. Overstaffed organizations create work rather than performance. They also create friction, sensitivity, irritation, and concern with feelings.

It is a symptom of malorganization to rely on “coordinators,” “assistants,” and other such whose job it is not to have a job. This indicates that activities and jobs have been designed too narrowly, or that activities and jobs, rather than being designed for one defined result, are expected to do a great many parts of different tasks. It usually indicates also that organizational components have been organized according to skill rather than according to their place in the process or according to their contribution. For skill always contributes only a part rather than a result. And then one needs a coordinator or some other such nonjob to put pieces together that should never have been separated in the first place.

“Organizitis” as a Chronic Affliction

Some, indeed a good many, businesses, especially large and complex ones, suffer from the disease of “organizitis.” Everybody is concerned with organization. Reorganization is going on all the time. At the first sign of any trouble, be it only a spat over a specification between a purchasing agent and the people in engineering, the cry goes up for the “organization doctors,” whether outside consultants or inside staff. And no organizational solution ever lasts long, indeed few organizational arrangements are even given enough time to be tested and worked out in practice, before another organization study is put in train.

In some cases this does indeed suggest malorganization. “Organizitis” will set in if organization structure fails to come to grips with fundamentals. It is, especially, the result of not rethinking and restructuring the organization when there is a fundamental change in the size and complexity of a business or in its objectives and strategy.

But just as often “organizitis” is self-inflicted and a form of hypochondria. It therefore should be emphasized that organizational changes should not be undertaken often and should not be undertaken lightly. Reorganization is a form of surgery, and even minor surgery has risks.

The demands for organization studies or for reorganization as a response to minor ailments should be resisted. No organization will ever be perfect. A certain amount of friction, of incongruity, of organizational confusion is inevitable.