The decision-making process in an organization can be viewed as a sequence of stages whose completion leads to an effective solution to the organization's problems. It can be considered, with some degree of convention, that this series consists of ten stages:
- Defining the goals of the organization.
- Identifying problems in the process of achieving those goals.
- Investigating problems and establishing a diagnosis.
- Searching for a solution to the problem.
- Evaluating all alternatives and selecting the best one.
- Obtaining agreement on the decision within the organization.
- Approving the decision.
- Preparing to implement the decision.
- Managing the application of the decision.
- Verifying the effectiveness of the decision.
There is as yet no consensus on the number and composition of the stages that actually make up the decision-making process. One reason for this lies in the differing circumstances under which specific processes take place, which assign varying degrees of importance to different stages of the decision-making process in different cases.
We shall consider the terms "decision-making" and "problem-solving" as synonymous, and view the decision itself as an effective response that ensures the desired outcome given the existing or possible state or states of the organization.
Since the primary responsibility of top-level management is to create a management system, it is clearly important to establish, from the outset, the nature and character of that system. A management system can be defined as a subsystem of the organization whose components are groups of interacting people: its functions consist in perceiving certain organizational problems (inputs) and subsequently performing a set of actions (processes) that produce decisions (outputs) which increase the revenue from the activities of the entire organization (satisfaction) or optimize some function of all organizational inputs and outputs.
The management system is a normative "human–human" type system whose primary goal is the effective resolution of organizational problems. To design a decision-making system, top-level management must: identify the system's main components, build a model, and collect and use data on the system's effectiveness. Systems analysis is the process of decomposing and identifying progressively smaller parts of a system; systems synthesis is the process of regrouping these parts.
Organizations can be viewed as goal-directed systems of cooperating individuals that enhance their personal well-being. An organization needs a problem-solving function because questions arise regarding the use of organizational resources and the distribution of the aggregate fund of benefits among its members, and the organization must adapt to its environment and develop new programs of action. The management system provides the organization with a set of decisions that are applied by executors when corresponding situations arise. Unlike individual decision-making, organizational decision-making is a group process that requires agreement and adaptation among group members.
The system is unified into a whole through a network of information flows; if a decision proves unsatisfactory, the problem is sent back through the feedback loop to the management unit, which repeats the entire process. The system is partially closed. The management system can be represented as a work process that builds solutions to organizational problems, and the diagram of this work process does not differ from the diagrams of other organizational subsystems. All organizational problems will follow a predetermined path, and managers will perform only those tasks that form part of the problem-solving process. This does not mean, however, that in all cases problems will necessarily be solved in precisely this manner; different approaches to problem-solving may be required. Nor is only one process necessarily needed; an organization may provide for many different processes through which organizational problems will be solved. It is important, however, that these processes be designed in advance; once a particular process has been selected for solving a given problem, it should be applied.
Organizational problems may be identified directly by top-level management, managers, rank-and-file employees, and shareholders. Questions should be raised by managers at all levels; however, it would be ideal to use as many sources as possible, including both shareholders and rank-and-file employees, since as the number of participants identifying problems increases, so does the number of people with a stake in the organization's work.
In addition, the organization should establish an information division that will serve as the primary organ for identifying problems. This division collects, classifies, and organizes information about the external and internal environment, then compares it with pre-established organizational performance standards. If the standards are met, there is no problem; but if deviations from the standards are detected, the cause of the deviation is formulated as a problem requiring resolution. The information division collects information about specific areas of the environment in which the organization operates — for example, the economy and government policy. Critical environmental factors are then identified, and the most important of these, such as customers and suppliers, are surveyed. The activities of the information division can be centralized, in which case it is responsible for reviewing and assessing both the external and internal environment of the organization, or the performance of this function can be decentralized so that each division collects the information it needs independently. To facilitate decision-making, it is necessary to establish a document workflow and develop document forms. Using a document form for the problem identification stage can be very helpful, as it facilitates the transfer of problem-related information within the organization.
Having received a problem, the manager can proceed to analyze the causes of its occurrence, which will enable the correct actions to be taken. Initially, the manager reviewing the problem familiarizes themselves with existing instructions to understand the nature of the already applied decision (if one exists at all). Alternatively, they may request additional information from the person who identified the problem. Problems identified by the information division or discovered as a result of decision monitoring are usually accompanied by corresponding information, and the manager should obtain it. If a ready-made decision already exists, the manager may examine the actual execution of the operation and discover that the given decision is not being applied, or is being executed incorrectly, or is ineffective, or is wrong.
The manager must have relative freedom to conduct investigations across various areas of the organization's work, to obtain data from the information division, and to collect any new data that, in their opinion, is necessary to identify the cause of the problem being solved. A rule should be in effect within the management system that all divisions provide full support to managers, and that managers may freely collect information — with the exception of classified information — in all departments of the organization. The information division is undoubtedly the primary source of information about both external and internal events, and this division should develop a set of procedures for searching and requesting the information that managers need.
After the manager has studied the causes of the difficulties, they communicate their findings to the management division. If they believe that an existing decision is not being carried out, the management division notifies the manager responsible for that area, who must rectify the situation. If it becomes necessary to repeat the stage of preparing the decision for implementation, the appropriate divisions are notified, and the management division draws up a new routing and schedule for the problem resolution and assigns responsibilities for resolving the problem; this procedure may be repeated until an effective decision is obtained.
If the decision is incorrect, the earlier analysis must be revisited to determine whether an error was made in selecting alternatives or whether the causal factors have changed. If no satisfactory decision is found, the causal factors must be redefined; when a new possibility arises, it may also be necessary to re-examine the causes of the problem. If during the investigation the manager identifies another problem, they also report it to the management division, and this new problem will be processed like any other. In cases where problems are interdependent, a dilemma may arise: the investigator cannot solve their problem until another related problem has been resolved. In such a case, the resolution of the first problem is suspended until the second is resolved.
The next stage of decision-making is the search for alternative solutions. This is probably the least formalized of all ten stages; at this stage one must rely on the creative abilities of managers. However, the manager to whom a problem has been assigned usually bears responsibility for its resolution and should possess sufficient technical competence and experience, as well as being well informed about developments in the technical literature. Managers should not only be familiar with solutions proven in practice but also create new ones. Therefore, their inclination toward experimentation and development in this direction should be encouraged. Research and development, for example, should be conducted not only in the design department but in all parts of the organization.
At this stage, the person responsible for resolving the problem evaluates the alternative solutions and selects the best among them. To standardize such calculations within the organization, one can use, for example, a solution comparison form. If the manager cannot assess the costs and output of a particular solution, the design department or the accounting department may assist. After the investigator has selected the best of the available solutions, they should plan how, when, and by whom it will be implemented, who will manage its application, and who will verify the results. They should also adjust the budget accordingly.
The next stage in the decision-making process is to identify those proposals that directly or indirectly affect more than one division; this makes it possible to determine the zone of influence of such proposals on all the affected divisions. This function is performed by the management division. After the management division has determined which divisions will be affected, it selects the managers with whom the proposal must be coordinated, draws up a schedule for their work, and sets a deadline by which they must communicate their reaction to the proposal. Each division, upon receiving the proposal, calculates the expected net benefit or effect of applying the proposal to its specific operation. Divisions also consider other aspects of the proposal, such as its implementation and verification; they may object to any phase they consider impractical. Any proposal is also reviewed by those departmental employees whose compensation and work are affected by the proposal, and their reactions are summarized by the investigator in a report on the meeting of lower-level managers.
After each division has completed its calculation of the benefit to be gained from implementing the proposal, the results are forwarded to the management division, which compiles a summary table of benefits and determines the net benefit for the entire organization. Some divisions may wish to obtain additional information or may conclude that the disagreements are so serious as to require discussion at a meeting of interested parties. Consequently, each division should be given the opportunity to propose meetings at which all interested divisions would be represented. If such a meeting results in changes to the original proposal, the benefit that the organization would derive from the proposal is recalculated.
A manager who disagrees with a particular decision or believes that a given decision will adversely affect their activities must prepare a convincing quantitative justification for their objection. After it has been established that net benefit can be derived from the proposal's implementation and that the proposal is accepted by the organization's managers and rank-and-file employees, the management division sends the proposal for approval.
Centralization of authority in an organization is especially effective at the decision approval stage, since individuals vested with authority can in this case devote more time to reviewing and endorsing decisions. In fact, no other duties can be imposed upon them. And the management division, of course, knows in advance which officials approve particular decisions. The person approving proposals may exercise one of four options: they may approve the decision, reject it, request additional information, or suggest that the analysis of the problem be continued. If the manager does not approve the decision, it means that despite a properly conducted analysis, additional information is required or alternative solutions need to be considered. There may be decisions with great uncertainty regarding potential benefit, and the top-level manager will have to choose one decision from four or five alternatives. In this case, they may also return the decision for further clarification before final approval. If the decision-making process has been carried out correctly, the approval of the decision usually becomes a formality. A complete refusal to approve a decision is a relatively rare occurrence, because it indicates serious deficiencies in the decision-making process itself.
The process of preparing the decision for implementation is set out in the decision itself. Therefore, we shall describe only the individual tasks that must be carried out when modifying organizational resources. Upon receiving the approved decision, the management division notifies the managers that preparation for its implementation may begin. It once again draws up a routing and schedule for the preparation of the decision's implementation through various phases and assigns the managers responsible for them. Some parts of the decision may be executed simultaneously, while others must be executed sequentially. In the course of executing the work schedule, the management division sends instructions for preparing the implementation of the decision to various managers; these instructions specify, among other things, the expected date of completion. In the participating divisions, a copy of this decision is filed with the current instructions, and on the established date, the decision goes into effect.
To ensure that the organization is prepared for certain events, some decisions are programmed in advance, and the management division activates these programs as described. This mechanism of pre-programming decisions makes it possible not only to plan the organization's future activities but also to coordinate the actions of the information division and the operational divisions of the organization. If the organization has planned certain changes in the event of special situations in the external environment, the information division can notify all interested departments of the need to implement the pre-programmed decision.
The principal duty of a middle-level manager heading a division consists of maintaining the division's ability to carry out the decisions prescribed to it. The decisions contain instructions on how to resolve these problems, but it is necessary to fill out requisition forms and forward them to the appropriate divisions. If the manager fully knows all current decisions, they will continuously ensure that employees carry them out and, if necessary, make minor adjustments. When serious problems arise, the manager will look for an appropriate decision in the decision manual. If no appropriate decision exists, they will fill out the form for recording the identified problem and send it to the management division. It is precisely at the stage of managing the application of a decision that the decision is transformed from a document into action; and the duty of managers is to serve as the connecting link between the instruction and the employee.
There are three methods for verifying the effectiveness of a decision. The first method determines whether the expected and actual returns are equal and whether the difference between them falls within acceptable limits. This comparison can be made monthly, weekly, or at any other interval. A significant deviation is noted, and information about it is sent to the management division in the form of an identified problem.
The second method carries out a behavioral check, analogous to that used in equipment preventive maintenance programs. At randomly selected times and places, a check is performed to directly observe whether the corresponding decision is being carried out and whether it is proving effective. Here again, deviations become problems entering the management division. Moreover, since both employees and managers raise problems, there is a third method of verifying decision execution, which consists of the following: one division reports that another division is not properly carrying out a decision. This also becomes a problem subject to investigation.
The stage of verifying decision effectiveness constitutes the feedback, or the operation that closes the loop.
Although this stage is not the only process in the organization that ensures the detection of problems, it is the part of the organization that evaluates existing programs from the standpoint of whether they are being carried out correctly. For the decision effectiveness verification mechanism using the first method, a centralized data processing system is required. Under the second method, the manager solving the problem is charged with periodically checking all decisions that have been put into effect. To this end, a schedule for reviewing all such decisions is drawn up. The third method represents, to a greater or lesser degree, an ad hoc procedure.