Capabilities of a data scientist

From Olav Laudy Data Science
Jump to: navigation, search

The term data scientist is a buzzword these days. There are companies calling for data scientists while they do not need them and there are people who have some experience with Excel and call themselves data scientists. The real need is not one person who has all data science skills, but an analytical ecosystem, both horizontal (from database experts to visualization experts to analytical end-users) and vertical (the data science career framework). Some data scientists are deep into model development, programming or other backend tasks, while other data scientists are evangelizing and educating the masses. Without the pretention that the skills below define the one and only data scientist, I found the following characteristics important:


  • Identifying opportunities to use data in order to drive business results.

A data scientist has strong business acumen. They are intimately familiar with the goals, and they can't stop pondering how data can enhance that. Status quo is not enough. Standard solutions are old. Driven by curiosity, formulating testable hypotheses on the way and conducting clever experiments, the data scientists uncovers new opportunities to have the widespread use of data to enhance decision making.


  • Creating compelling story lines around the use of data to drive the adoption of analytics.

A data scientist knows that 80% of the impact of his work is achieved by communicating. Although deeply engaged in the mathematical reality of the data science, the data scientist understands that people need stories. The data scientist is a master in showing what can be found in the data, and how those findings can enhance the business.


  • Gaining trust by explaining data science concepts in business terms.

The world is not populated by mathematicians. The data scientists understands that, in order gain acceptance of the methods used, he (or she) needs to speak the language of the business. That is: showing the results of models in terms of business impact. Showing how the models work, but rather than focusing on the technical part, being able to visualize the model outcomes and cleverly show how the predictions make sense.

Proof of Value

  • Quickly turning around innovative data science solutions using IT technology to prove ROI

The data scientist is hands on; someone with a hacker's mentality. Data is dirty and in order to use it, you need to make the right assumptions: too refined, and you never reach your goal, too coarse and the results do not reflect the true state of affairs. Nevertheless, the data scientist needs to find his way around IT architecture, coding and presenting the results. Maybe the skills are not as a full IT professional, a full stack programmer or a Powerpoint hero, yet, the data scientist get things done.


  • Embedding data science results in IT systems to generate continuous value.

The data scientist recognizes: a model is as strong as its deployment. The focus, therefore, is always on how any data science result can generate results that are useable for more than once. Typically this is done by scoring engines that are embedded in IT architecture.


  • Training data scientists to drive the widespread use of data in order to improve business results.

The data scientist is enthusiastic about his skills and always tries to get people to understand some interesting data science principle. The data scientist recognizes that (currently) there is a shortage of data science related skills and educates those who are interested to get them up to speed in the area of data science, just right for everyone's level.


  • The ability and desire to collaborate with a wide range of people with different skill sets to manufacture advanced value chains.

The data scientist recognizes that results are achieved alone, but is an effort of many people with different skills sets together. The data scientist is capable and enthusiastic to work with people with a wide range of skills to drive results.