From our last post about the role of the CMO, it is obvious that the Big Data Revolution has created a lot more new roles. One role in particular is that of the Data Scientist. It is easily the most hyped occupation of this decade, described by Hal Varian as « the sexiest job in the next 10 years. » But is there any science actually involved in the role? Or is it simply a more ‘sexier’ way to describe a Data Analyst?
In essence, a scientist can be described as someone who attains knowledge through the application of specific methods. However, if you were to learn analytical tools the process would be much faster than simply attaining knowledge. Moreover, the term « data scientist » may appear new, but in actual fact it first appeared in the work of John Tukey in the 1960’s. IBM gave a more recent definition of the role as someone who « represents an evolution from the business or data analyst role. The formal training is similar, with a solid foundation of typically in computer science and applications, modeling, statistics, analytics and math. » The key characteristics of a Data Scientist are: business knowledge, artistry, problem solver, taking multiple perspectives, and being influential. There could easily be more ways that describe the role, and that goes to show how well defined it has come to be over the years.
In reality Data Scientists are scarce. Not every company wants to engage in Big Data therefore they feel there might be no particular position in their company for them. Only recently has there been any form of third level education to ever formally become a ‘Data Scientist.’ It is not surprising that the course is being offered by UC Berkeley, who in the past linked up with IBM and others to develop a center for Open Innovation. Given that they will be taught the basics of being a Data Scientist, will this be the going trend for other well respected Universities? Will there be as many as 20 or 50 ‘Data Science’ Masters available in 2014?
The role seems quite difficult to uphold and maintain, as true Data Scientists have a lot of requirements. These include: advising executives and product managers on the implications of data; telling stories with data from having good skills at writing code; being able to structure large amounts of formless data and make sense of it all; and requiring their own time to work on projects, with the intention of building close relationships with executives. If Data Scientists can cut a lot of these requirements down into key, performance-measuring responsibilities then more companies will look into taking them on.