
Almost all decisions are made in the face of uncertainty, since on very few occasions do we have all the information to decide without any doubt. We all know that uncertainty is an intrinsic element in the life of any organization, it represents a source of threats and opportunities that can destroy or create value.
However, determining an acceptable degree of uncertainty in order to maximize value creation is one of the main challenges that managers of all time have had to face.
Decision-making based on Data Analytics is a "new" form of management that allows us to respond more effectively to these threats and opportunities, increasing management’s confidence and significantly improving the value creation capacity of the company.
To understand how this "analytical" form of management works, it is important to know two simple definitions.
The first is the concept of model, which can be of various types: business models, process models, predictive models…, but ultimately all models have the same purpose, they are used to better understand problems and make better decisions.

Another important definition to consider is Decision Analysis (DA). It is a systematic, quantitative and graphical approach, generally used to evaluate multiple and sometimes complex options.
One way to deal with certain difficult decisions is through the combination of these two previous definitions, which brings us to the widely known Decision Models, these models incorporate aspects of psychology, management techniques, statistics, and economics to define the best alternatives in a decision.
Keep in mind that models are not representations of reality. British statistician George E. P. Box said:
" All models are wrong but some are useful".
This aphorism is a great truth, since the best models are vague approximations of reality, but they are very useful for understanding problems and analysing alternatives.
On the other hand, for the success of any Decision Model there must be two main actors, which are not necessarily individuals, but they could be work teams.
One is the decision maker, normally represented by the manager or director of the company, and the other is the data analyst or data scientist, who is in charge of building the model and supporting the manager throughout the decision process.
Therefore, the data analyst must be able to understand and combine the process or business model with the appropriate quantitative methods that allow him or her to represent, as much as possible, the nature of the decision. And as usual in business, this should be done in a relatively short period of time.

Due to the differences between the profiles of decision makers and data analysts, misunderstandings are very frequent when defining the models that best suit the decision.
These drawbacks can be avoided if the manager works closely with analytics professionals to develop a simple model that allows a preliminary understanding of the main factors involved in the decision. After the manager has become familiar with this model, additional details and greater sophistication can be progressively added until the desired level of satisfaction is achieved.
Some of the most frequent factors that make it necessary to incorporate more details into a model, is the fact of having multi-objective decisions.
When there are two or more objectives, many times they conflict with each other, that is, a strategy can be optimal with respect to one objective, but be the worst option with respect to the others. Such decision problems have to be evaluated through more complex analytics models, such as " Multi-objective optimization"

This type of decision analysis used to be used exclusively to evaluate difficult decisions involving multiple variables or having many possible outcomes or objectives, such as those that are frequently observed in mining or oil and gas projects. Fortunately, new technologies now allow a large amount of information to be available at any time, as well as the combination of multiple analytical approaches, which facilitates the use of these techniques for any type of decision in any industry.
Regardless of the complexity of a decision, a progressive construction of models, that involves the main actors, is the most important factor in the development of an agile and flexible process that leads to successful decisions.
Data analysis should be considered as an integral part of any decision-making process, this even begins little by little to be incorporated into the new regulations of companies in some countries. Where the company’s information systems must be fully accessible and auditable, and if required, a complete presentation with the details of the project must be presented to a technical committee.
The use of Data Analytics to support Decision Making should not be seen as an impediment to the agility of an organization, on the contrary, this is the most effective form of management to improve business control, increase productivity and profits and project more successful strategies for the future.