Data Scientist is a job which is closely related to buzzwords such as industry 4.0, digitization and big data. The Harvard Business Review has chosen the Data Scientist as “Sexiest Job of the 21st Century” for a reason. But why, in spite of all this, are so many companies lacking representatives of this highly sought-after profession? How do you become a Data Scientist? And what skills and knowledge does a Data Scientist need?
Different types of Data Scientists
In order to answer the question why there are still too few Data Scientists, it is necessary to clarify how the term Data Scientist has to be interpreted. A rough distinction can be made between two different job profiles, whereby representatives of both groups would define themselves equally as Data Scientists.
Firstly, there is the Enterprise Data Scientist. He works in a company and is a mixture of business economist, IT specialist, statistician and communication expert. On the other hand, there is the Academic Data Scientist. He can rather be assigned to the pure algorithm development. Furthermore, he works with “ideal” data and is less practice-oriented but rather methodologically oriented. In the enterprise environment, it is rarely the case that a completely new algorithm has to be developed. Rather, it is a matter of adapting or extending existing concepts to the concrete problem definition. A complete redevelopment of modelling methods would often take too long.
What skills does a Data Scientist need?
Put simply, a Data Scientist must have technical know-how and communicative strength. The basic equipment always consists of very good knowledge of computer science, business administration, mathematics and statistics. A Data Scientist must also be able to understand business processes. Without them, he can hardly interpret and understand the results of analyses and the data-generating processes. In addition, a profound understanding of data structures, databases and models is also a compelling competence.
The standard also includes programming skills in order to be able to work and interact with data. This includes, for example, mastering very large amounts of data, creating complex queries and linking different data sources. Statistical and analytical skills come into play when predictions of future events are to be derived from historical data. But not only the ability to understand and analyse processes is very important. Visualizing analysis results is also part of the toolbox.
The widespread image of the computer nerd doesn’t fit the Data Scientist. His profile is characterised by a high level of problem-solving competence and excellent communication skills. The latter are necessary to communicate complex issues and models in such a way that management, users and customers understand and trust the solution. Data science projects can only succeed if the customer’s perspective and vision are not lost on the way through the data jungle. After all, at the end of the day it’s all about telling the stories that are in the data – suitably and relevantly packaged for each target group.
Why there is no perfect Data Scientist
The sheer amount of what a perfect Data Scientist should be able to master is inhuman. And even if the sum of all key skills were to be combined with rapid technological development, the perfect Data Scientist would never exist. Maybe that’s not even necessary. In most cases, it is “only” about being able to solve data-driven problems from start to finish within the framework of concrete data science projects. This usually only works in a team. Experts such as statisticians or data engineers must be employed for certain subtasks. They contribute their respective strengths and advantages during a project – and a Data Scientist also develops focal points in his work that are useful for his field of application.
Detective work is not for everyone
If you look at this comprehensive and very demanding requirement profile, you will immediately see why there is a lack of data scientists. The combination of excellent communication skills and extensive technical know-how is a major hurdle. Not everyone is equally prepared for the challenges of everyday work. Because the core business of Data Scientists insists on hypothesis formation, which can be confirmed, but must always be rejected. On the one hand, this hypothesis-driven and experimental way of working is similar to scientific work, which explains the concept of the Data Scientist. On the other hand, the Data Scientist often acts like a kind of detective. Time and again, courage is needed because problems have to be questioned: What should be achieved and why?
Data Scientist: A career with a future
Due to the so far insatiable need for Data Scientists, many places in Germany, Switzerland and Austria are developing (advanced) courses of study and further education opportunities. In addition, companies – such as Alexander Thamm GmbH in Munich, for example – offer trainee programs for data science. The prerequisites for a Data Scientist are still excellent: very good earning opportunities, a diverse and multifaceted range of tasks and, above all, great future potential.