The purpose of this article was to determine the success factors of decision support systems (SSD). Books and electronic sources were investigated that addressed the issue of decision support systems in an objective manner and that spoke of critical factors that could determine the success or failure of these systems. It was found that there are technical and organizational factors that affect the fulfillment of the purpose of the system, and that it is a combination of both - as principles - that suggest a greater probability of success.
It is recognized by practitioners and researchers alike that top management needs more capacity than that provided by printed reports that are the hallmark of management information systems. A wide variety of applications were developed, and these differed from other applications in the sense that they were interactive, capable of answering "what if" questions and including user-friendly interfaces. These applications were focused on problems that were faced by senior management levels.
It is at this point that decision support systems (SSD) are born, which are computer-based information systems that combine models and data to try to solve problems with the help of an extensively involved user. (Turban, McLean, & Wetherbe, 2002)
In order to understand the principles of success that an SSD requires, it is necessary to know a little more about them, specifically their characteristics and their components.
Features and capabilities of an SSD
While it is true that the term SSD has different meanings for many people and can be seen as an approach or as a philosophy, there are certain characteristics that have been recognized as ideal. However, most SSDs have only some of the following attributes:
- An SSD supports decision makers at any managerial level, whether they are individuals or groups, mainly in semi-structured and unstructured situations, through the combination of human judgment and objective information An SSD supports several interdependent and / or sequential decisions An SSD assists in all phases of the decision-making process - intelligence, design, selection, and implementation - as well as a variety of decision-making processes and styles. An SSD is customizable by the user through the time to deal with changing conditions. An SSD is easy to build and use in many cases. An SSD promotes learning, which results in new demands and application refinement, which in turn results in additional learning.An SSD usually uses quantitative models (standard and / or custom-made) Advanced SSDs are equipped with a knowledge management component that allows efficient and effective solution of very complex problems An SSD can be disseminated for use on the Web An SSD enables easy sensitivity analysis.
Components of an SSD
Apart from these ideal characteristics, each SSD system consists of at least the data, user interface, and model management subsystems, as well as users (see Figure 1).
The SSD data subsystem is made up of the SSD database, the database management system, the data directory, and the facility for making queries.
The SSD model management subsystem comprises the model base, the model base management system, the modeling language, the model directory, and the command processor, model integration and execution.
The user interface subsystem includes not only hardware and software, but also factors involved with ease of use, accessibility, and human-machine interactions (Turban & Aronson, 2001, p. 107).
Finally, the user is the person who has to make the decision that is intended to be supported by the SSD, also called the manager or the decision maker. An SSD has two classes of users: managers and staff specialists. Generally, managers expect a friendlier interface than that expected by staff specialists since the latter are more detailed and willing to use more complex systems.
More complex systems adapt other components such as the knowledge management subsystem, as well as modules tailored to solve specific problems.
Principles for a successful SSD
The theory of SSDs is vast and to a certain extent not very complex to understand, at least if we talk about the most general aspects. The most labor-intensive part comes at the time of implementation, whether it is a solution already developed or a custom application. Alter (1981) developed a series of generalizations based on information obtained from his research, as well as the research of others, to define a series of points that serve as principles that suggest the success of an SSD.
- Principle 1. The SSD should improve decision making
An SSD system should be evaluated to the extent that it improves decision-making and not to the extent that it is interactive, friendly, or semi-structured. For this to be achieved, the SSD must provide information that was previously inaccessible, it must provide better alternatives to draw inferences and better ways to explain decisions to others, among others.
- Principle 2. The SSD should contain as much "intelligence" as possible about the user's problem.
A test that determines that the SSD has sufficient intelligence is to ask the user how the system improves decision making. If the user is able to demonstrate or explain in detail how this occurs, the system is most likely smart enough to be useful.
- Principle 3. The SSD should be used through the most cost effective usage pattern.
Each usage pattern (terminal mode, broker, vendor, and subscriber) has particular benefits and costs, and it is necessary to know that each pattern can be implemented well or poorly.
- Principle 4. The SSD should be used by experts who understand its meaning and how it should be used.
Since SSDs are made up of analytical models, they require effort to be understood so that they serve their purpose. They need to be used only by people willing to take the time to understand these models.
- Principle 5. The SSD should be user controllable
The SSD user must be able to specify which reports or calculation options they want, when they want these reports, and in what way these reports should be limited in scope and level of aggregation.
- Principle 6. The SSD should contain any data, model, deployment capabilities, and human intermediaries required to improve decision making.
The user not only needs data listings in an orderly manner but also needs statistics and operations research. This information must be approached through an explicit mathematical model so that the information is valuable for decision making. On the other hand, it has been recognized that efficient graphic displays help people to perceive patterns.
- Principle 7. SSD should be implemented through whatever development strategy represents the most cost-effective and least risk-prone during establishment.
Although evolutionary approaches are appropriate in some situations, the benefits and risks of other ways of implementing systems are worth exploring.
As I mentioned earlier, meeting these principles assumes the success of an SSD, although other critical factors must be taken into account. One factor to which special attention should be paid is resistance to change. The implementation of an SSD requires a process of change, mainly from managers. Changing the traditional way of doing things can lead to uncertainty and discomfort.
Some methodologies that can facilitate the management of change in the organization are based on the development of high-performance work teams, the management of best practices, and the minimization of resistance to change through participation, communication and training (Calderas, 2001).
Conclusions
SSDs represent an important tool for managers at any level of the organization due to the capabilities they incorporate. On the other hand, its implementation process is a process that requires special attention. The technical details of the system are very important but it has been seen that the other aspects that are involved - users, control and fulfillment of the purpose - may be of greater importance. I think that the seven principles stipulated by Alter (1981) serve very well to evaluate an SSD, although on the other hand I believe that they are not a guarantee of its success in an organization. Factors such as resistance to change can become a critical aspect determining the success or failure of the system.
It is a very interesting combination that makes the implementation of an SSD more complex, which although we are talking about a computer system, has a specific and ambitious purpose. If we talk about the system as it is and its closest components, compliance with the seven principles of Alter suggests an efficient result of the implementation of the SSD.
References
Alter, S. Transforming DSS jargon into principles for DSS success (1994), in: P. Gray (Ed.), Decision Support and Executive Information System, Prentice-Hall, Englewood Cliffs, NJ, pp. 2-26.
Caldera, Borja. Decision support systems and resistance to change (2001). Netmedia, Article 2263.
Turban & Aronson. Decision Support Systems and Intelligent Systems (2001). Upper Saddle River, NJ: Prentice-Hall.
Turban, McLean & Wetherbe. Information Technology for Management (2002). Massachusetts: Wiley.
