The Art of Data Science

If we asked 10 different people “What is data science?”, we would probably get 10 different answers. In essence data science is what data scientists do. Not very helpful now, is it? In the following post we provide you our view on data science. Stay tuned to find out about the art of data science.

What is really data science? Is it a form of science, is it statistics? Is it programming, is it visualizing data? In reality, it includes all of the above. To look at an exact definition: data science aims at extracting insights and knowledge from a large amount of unstructured (or structured) data by using algorithms, techniques drawn from computer science, statistics and mathematics. This knowledge is then used to answer different business questions and prepare for future issues.

Reading this definition not many people would get too excited. Some might wonder: how does one end up as a data scientist? If you ask little children what they want to be when they grow up, I guarantee data scientist will not be an answer. For now.

The name data science implies that it is a form of science. But is this true? What could art have to do with it? A good data scientist holds the following skills: mathematics, statistics, programming, database management and interaction, domain knowledge and finally visualization and storytelling. These past three important skills hold the art aspect of data science. Knowing machine learning techniques, all the technicalities is really not enough. These techniques need to be applied differently in every situation, since each business challenge differs. One size does not fit all. Data scientists need to be creative and draw upon their inner artists. As we like to define it: data science is the art of engineering actionable knowledge from raw data.

The work of data scientists consists of multiple steps:

  • Problem statement: define exact challenge that needs to be solved
  • Collect and prepare all relevant data
  • Choose appropriate technique, algorithm
  • Adjust elements based on initial findings

While performing these steps, data scientists face many different problems, bugs. The same set of rules cannot always be applied and creativity needs to kick in. Problems need to be faced from different angles. This is one of the main “art” aspect of data science. If you have to get creative, you are an artist. Next to that, the unmeasurable valuable information data scientist find is not enough in itself. While machines can be told to execute certain tasks and provide result the job does not end there. Data scientists need to engage with stakeholders and all the uncovered results need to logically connect. This is where storytelling comes into the picture, and the other art aspect of data science.

What do we mean by storytelling exactly? The revealed problems and the found solutions need to be presented in a way that it speaks to the audience and does not only capture logic but also emotions. The findings can be visualized on different graphs, each type of graph fitting perfectly to the presented data. It is up to the data scientists whether to use bar charts, line plots, gauges or any other. There is never one correct way to do it, it is up to the data artists to find the best one fitting their story and their data.


For a long time there were no specific studies aimed at training data scientists, it more just evolved from already existing positions. However, with the development of the profession and the huge demand companies have developed for data experts resulted in the full separation and emerge of the data science profession. While the good base of statistics, programming and other technicalities can be taught, the art aspect comes with practice and intuition.

Interested how our data scientists can work their magic within your organization? Or do you have any questions regarding our post? Contact us for a free consultation or a discussion via LinkedIn or our website.

Back to blogs