Is Improving Intervention Feasible?

The experience of reading this course shows that when the definition of improving interventions is announced in the audience, there necessarily found skeptics (and the older the audience, the more of them), who consider what has been said to be a beautiful but unattainable goal. There are many reasons for skepticism. Let’s discuss the main ones.

  • “It is impossible to make everyone feel good”. This is a substitution of the thesis: improving intervention is not when “everyone is good”, but when no one is worse, and this is not the same thing.
  • “Experience shows that no solution to any complex problem is possible without creating new problems”. In reality, it often happens. But this is only a consequence of the fact that the systems technology was not realized properly. Obvious examples of this are the antialcohol campaigns conducted in United States in the beginning of the 20th century, and in the end of it in Russia during the Soviet era. Both attempts failed and gave rise to many new problems. The negative experience of solving problems is not connected with the impossibility of solving them successfully, but with noncompliance with the requirements of their systemic solution.
  • “Very often failures in solving problems are due to lack of resources and mistakes in decision-making". Therefore, it is important to consider improving intervention as an ideal to be pursued, even if it turns out to be not fully achievable. “Improving intervention is often difficult to find, but rarely impossible” (R. L. Ackoff [1]). Applied systems analysis proposes a method of moving toward a goal, although in practice this may be hindered by a lack of information, making mistakes, insufficient resources, and a shortage of time. It is important to move in the right direction as far as possible (this will be discussed in detail in Part II of this book).

Perhaps the most serious reason for doubting the feasibility of improving interventions is the contradictory interests of the participants in the problem situation, sometimes reaching the point of conflict. How can you make an improvement intervention when someone is dissatisfied only with the fact that some others are well? Or when everyone seeks to prove and assert their rightness, and their positions are different or even incompatible?

There are several possibilities to move toward the improved intervention even in such conditions (see below). In fact, in systems analysis, it is offered to refuse clarification “who is right?”, and move to the search or creation of “common agreement”, preferring common peace and efficiency (“consensus”) before their rightness.

Four Types of Improving Interventions

Returning to the ways of solving problems, it is interesting to consider the classification proposed by R. Ackoff [1]. He noted that despite the enormous diversity and the dissimilarity of problems, there are only four ways to solve them.


This term in colloquial English denotes the actions of a priest who forgives sins to parishioners: he listens to confession and does nothing. In the professional language of systems analysis, this term denotes noninterference. Please note that this does not make anyone worse. However, preference should be given to noninterference only if any proposed interventions lead to worse results. For example, actions of a doctor in case of a difficult diagnosis (placebo), behavior of a sapper when an explosive device is unknown to him, and recommended non-interference in the family problems of your friends or spouses.


In such a type of intervention, the problem is solved partially, not fully, but in an acceptable manner. There are several possibilities to do this.

The first is to use insufficient resources for completely solving the problem to mitigate discontent in having solved the problem partially. A good example of this is some increase in wages, pensions, and scholarships to public-sector employees against the background of galloping inflation or allocation of limited resources by draw, by turns, or equally.

The second possibility is to try to return to the state when there was no problem: determine what caused the problem and eliminate the cause, that is, find the culprit and punish him. This (alas common) solution is incomplete, partial, and outdated: any event in the world is the result of many factors, and the elimination of one is certainly not adequate.

Another example of such an approach is to repeat the action that was previously successful in a similar case. But this involves the risk of insufficient similarity of circumstances, which can lead to unexpected results.


In the professional language of systems analysis, this is the term denoting the best intervention under given conditions. The relevant scientific term “optimal” has already entered the spoken language and public consciousness, so it is important to understand and apply it correctly. Optimal means the best under the given constraints. For all the seeming simplicity of this definition, it requires explanation.

First, what does “the best” mean? The same objects can be ordered in different ways, depending on what quality to consider. The criterion that evaluates this quality allows you to find the best (for this quality) alternative. How to choose the best option if the alternatives are compared not by one but by a combination of several criteria? This is not a trivial question. Let us consider (as it will turn out later, erroneously) this question as purely technical and consider it in the “Choice” section of the second part of the course. Meanwhile, it’s important for us that before talking about optimality, it is necessary to specify and determine by which criterion (or criteria) the compared options will be ordered, that is, in what sense we will use the term “best”.

However, this is not enough for optimality. The second, no less important, integral part of the concept of optimality is the dependence of the result of choice on the specific constraints in this situation. Under the same quality criteria, the choice from the same set of alternatives under different constraints will generally be different. Therefore, only those alternatives that satisfy the imposed restrictions should be compared with each other according to the chosen quality criterion: the best alternative in the sense of the criterion that does not meet the restriction cannot be implemented.

The desire to do everything as best as possible is so natural for people that it is not surprising how quickly the abstract mathematical concept of optimality has moved from science into business, governance, and even in everyday life. Although the widespread popularity of the idea of optimality (apparently, it was a consequence of the great fashion for cybernetics in the 1950s-1980s), few people, except specialists, drew attention to the warning of N. Wiener [2], the father of cybernetics, about the need for a careful use of this concept.

The fact is that intervention in a problematic situation is based on the information we have about the situation and the degree of knowledge regarding our situation may be different. There are well-structured problems that allow the construction of quantitative mathematical models, for example, many engineering and scientific problems (such problems are called “hard” problems). But many real-life problems are described in languages far from mathematics, from everyday problems described in spoken language, to professional ones, for example, many humanitarians and naturalists (such problems are called “fuzzy” or “soft” problems). Naturally, these differences are also taken into account in a number of details of the problem-solving technology, which makes it possible to speak of “hard” and “soft” technologies of applied systems analysis.

The difference between “hard” and “soft” methodologies of systems analysis can be emphasized once again considering optimality. Both optimality components (criteria and constraints) are sensitive to this difference. The very requirement that the quality criteria and limitations are expressed quantitatively implies that the situation under consideration is so well-studied that it allows the construction of its mathematical description. The optimization problem is a formal mathematical problem, quite adequate to the problems of the “hard” type (and the course of optimization methods is one of the largest and most ingenious mathematical subjects at the university).

However, within the framework of formal mathematical models, the “fragility” of optimal solutions was revealed; often even with small deviations from the assumptions in the formulation of the problem, the quality of its solution can change very dramatically. Therefore, consideration of the problems of stability (robustness) of solutions is an important section of the theory of optimization.

In the transition to “soft” problems, the situation is much more complicated. It’s not just that for such problems that it’s harder (if at all possible) to find quantitative measures for criteria and constraints. The main thing is that the “softness” of the problem is a consequence of little knowledge about the problem; in particular, there is no possibility to list all the important limitations, which, as we have seen, radically affects the quality of choice. Hence, the optimality in this case should be considered an unattainable ideal, which is still worth striving for. Additionally, optimization attempts should be considered only as an element of the trial and error method discussed at the end of this part.


Dissolution denotes an intervention that ends in complete extinction of the problem and nonappearance of new' problems. It would seem w'hat could be better than optimal? The essential difference between the third and fourth methods is that “optimal” is the best under given conditions, and “dissolution” considers the restrictions and conditions not as firmly fixed, but as subject to change or cancellation to find new' and previously unacceptable options, among which may be options that are much more effective than previously optimal.

An important option to “dissolve” the problem is to prevent it by taking measures to ensure that it does not appear. Here, the change of the system is made not after the appearance of the problem to solve it but before that to prevent it.

A visual analogy is the change in the fire protection system. In the past, this system mainly consisted of firefighters, brave, exotically dressed, and rushing dow'n the street in their cars with flashing lights and sirens to fight fire and heroically save the lives of people and their property. In the current system, the majority of employees are fire inspectors. They regularly check the buildings under their jurisdiction without light and sound effects to prevent fires. They prescribe improving changes in objects and processes and reducing the likelihood of fire and possible damage from them. This activity is no less — if not more! — important than extinguishing fires.

In this w'ay, in far-sighted organizations, special employees are introduced into the staff, whose duty is to conduct preventive inspection of all components of the organization, as well as early detection of dangerous situations and trends. Sometimes such employees are called “internal auditors”, “thinking engineers”, “inspectors”, etc. So far, this function is implemented only in the form of an internal audit. A gap in organizations is the absence of persons responsible for implementing preventive improving intervention. This responsibility is assigned to managers who are already loaded w'ith the management of the current activities of their unit.

R. Ackoff gave a vivid example of applying all four methods to solving one real problem.

A problem has arisen in a bus company in a large city: after the introduction of performance premiums for high quality, conflicts between drivers and conductors began. The quality of the drivers’ work was assessed by the accuracy of the observance of the timetable, and that of the conductors’ by how w'ell they serve the passengers. During rush hours, the conductors delayed the sending signal (they had to check not only the ticket availability but also the correct payment depending on the distance, and the incoming ones to sell the tickets, each to his destination station), and this adversely affected the drivers’ surcharge.

At first, the company’s management ignored the problem (ABSOLUTION), expecting that everything would settle down by itself. But the problem continued to escalate, and the trade unions were involved in the conflict. Subsequently, the leadership tried to return to the old system of payment (RESOLUTION), but both trade unions protested as this would mean the abolition of surcharges. The management then suggested to the trade unions to agree on the division of the surcharge fund (SOLUTION), but they did not want to cooperate with each other.

The problem was DISSOLUTED by a visiting systems analyst who discovered that during rush hours the number of buses on the line (regulated during the day) was greater than the number of stops. During these hours, the conductors began to be removed from the buses and instead were assigned to each stops. They started selling tickets before the bus arrived, had the time to check the tickets for those leaving, and began to send a departure signal on time. At the end of the rush hour, the conductors returned to the buses, and the extra buses were removed from the line. In addition, the company hired a smaller number of conductors.

This interesting example should not give the impression that the four types of solutions are arranged in order of absolute preference: in this case, “dissolving” (removal the restriction “conductors must always work in the bus”) turned out to be the best of them, but in other problems any other option can turn out to be the best.

Moreover, the preference for a particular type of intervention depends on the mental orientation of the manager. R. Akoff proposed [1] to distinguish four types of managers engaged in planning, decision-making, and implementing decisions:

  • 1. Reactive management is dissatisfied with the current situation and where everything is going; it prefers what was in the past; and its efforts are aimed at returning to the previous state by eliminating the causes of the changes. The preferred type of problem-solving is resolution, and the methods used are past experience, common sense, qualitative assessments, the choice of a “good enough”, and “acceptable” solution. An example of a satisfactory application of such an approach is the clinical practice of healing, but even then there are fatal failures. In management, such an approach is associated with authoritarian management, planning from the top down, aimed at solving separate problems and eliminating the undesirable without considering its connection with the other components of the situation (which often leads to even more undesirable problems).
  • 2. Inactive (passive) management is satisfied with the present and wants neither a return to the past nor future changes; impedes changes and appreciates stability; believes that if nothing is done, then nothing will happen, and that is good; believes that it is necessary to act only when there is a threat or a crisis.

At the same time, unlike reactivists who are trying to eliminate the causes, inactivists are engaged in the suppression of symptoms (“crisis management”). The preferred type of problem-solving is “absolution”, ignoring or denying the problem, and the hope that it will disappear or be resolved by itself.

  • 3. Proactive (preventive) management is convinced that the future will be better than the past and the present; therefore, it tries to accelerate changes and use the opportunities associated with them. Forecasting the future, the ability to learn and adapt to changes in the environment, planning, and creating changes becomes important. The preferred type of problem-solving is “solution”, finding the optimal solution, that is, the best in given conditions. Technologies are mainly quantitative, such as methods of optimization, operations research, mathematical (more often, linear) programming, risk analysis, balance of expenses and income, etc.
  • 4. Interactive management not only does not want to return to the past and the perception of the present but also to accept the impending future. It is sure that the future can be created by the efforts aimed at it. The preferred type of solution is dissolution, the implementation of changes in the system and/or its environment, leading to the disappearance of the problem. The technology of this is idealized design.

More about Applied Systems Analysis

This chapter has two objectives: to specify the concept of the problem and methods of solving it, and, moreover, to give a general idea of the applied systems analysis itself. The first goal may be considered as achieved (to the extent we need now). To achieve the second goal, two more features of applied systems analysis should be discussed, which have not yet been mentioned.

Consider the typical sequence of actions in time during the systems analysis (Figure 1.7).

At the moment “Problem”, the client turns to the system analyst with his problem, which he could not solve on his own. After signing a contract that imposes a number of obligations on both sides (which we will discuss later), work begins in accordance with the technology (described in the second part of the book). After a series of operations, there comes a moment, “Model”, when we get a sufficiently adequate model (the exact meaning of these terms will be given later) of the problem situation. Now comes the period of using the model to obtain the results of certain interventions. At the end of this period, a (usually multicriteria) selection of the most appropriate option “Decision” is made. From the decision to its implementation, the path is not easy, and requires a fairly strict adherence to technology (in modern language, this is called management). With diligence and luck, we can reach the “End” point when

the problem is solved. A more detailed description of the operations involved in each stage, with constant care to maximize the probability of success in the presence of traps, the possibility of error, limited resources, lack of time, and incompleteness and inaccuracy of information will be the subject of the second part of the course. Meanwhile, let’s pay attention to two more features of applied systems analysis.

The first follows from the fact that there was an “O” moment in the past (Figure 1.7) when there was no problem at all. If the client then turned to a systems analyst, one could subject analyzing the course of the future and predicting the appearance of a problem while maintaining the firm’s style and tactics. But it would also be possible to design an intervention that would prevent the occurrence of the problem. This is reflected in a slightly humorous saying: “The best systems analysis is one that does not come true”. Therefore, the forecasting technique is included in the arsenal of applied systems analysis. So far, this function is realized in organizations only in the form of internal audits. It is a fact, however, that clients most often turn to analysts after their own attempts fail to solve an already urgent problem.

The second very important and fundamental feature of applied systems analysis is indicated in Figure 1.7, coverage of the scope of analysis beyond the limit “End” of the problem. It allows you to discuss the question: what will happen after the problem is solved? Of course, the former client will again have some problem. Not as a result of the solution of the previous one if we tried to implement an improving intervention (in principle not creating new problems), but in case of unavoidable changes in the environment and in the system itself. Whether shall we go back to the consulting firm? This will not be necessary due to the specific feature of the systems analysis.

In fact, the problem-solving process cannot be performed only by the system analyst alone. To build a model of a problem situation, information possessed only by its participants themselves is needed. Therefore, a mandatory element of the technology is their involvement in the process of working on a problem. The system analyst knows what questions to ask, in what form, and in what sequence to ask, to build an adequate model of the situation on the basis of the information received, and only the participants in the situation can answer these questions. Moreover, it is they, and not the analyst, who will have to implement the developed intervention.

Consequently, the process of systems analysis will be performed by the employees of the client’s company themselves. Doing some work on their own is the most effective form of training them for this activity. Thus, in applied systems analysis, it turns out to be a naturally built-in and integral part of training in systems analysis itself. As a result, the need to reapply to a consulting firm is significantly reduced.

Questions and Tasks

  • 1. Explain the differences between the concepts of “problem situation” and “problem”.
  • 2. What does “to solve the problem” mean?
  • 3. Which three ways to influence a subject without changing reality can (under certain conditions) lead to the solution of his problem? What are these conditions?
  • 4. What is the main difference between a subject and an object?
  • 5. How to determine the meaning of the assessment expressed by a certain subject?
  • 6. Why do we have to rely on some ideology when intervening in reality to solve the problem?
  • 7. Reproduce the classification of ideologies into three types. What is the main difference between them?
  • 8. The goal of the applied systems analysis is to create an improvement intervention. List at least three reasons why this may not actually happen.
  • 9. Name four types of improvement interventions.
  • 10. Optimality is ensured only when two requirements are met together. What are these requirements?
  • 11. What is the important result of the applied systems analysis of a specific problem, besides solving the problem itself?
  • 12. This chapter introduces special concepts (and corresponding terms) that are included in the professional language of applied systems analysis. Some of these terms are used in the spoken language, but in a different, vaguer sense. Others have a special meaning. Check whether you can reproduce professional definitions for all the concepts listed below (if there is not one, be sure to find such a definition and try to remember it):
    • - a problem situation;
    • - evaluation (of something by the subject);
    • - a problem;
    • - solution to the problem;
    • - intervention;
    • - improving intervention;
    • - applied system analysis;
    • - optimality; and
    • - “hard” and “soft” problems.
< Prev   CONTENTS   Source   Next >