Guest blog post by Bernard Marr.
It depends entirely on how broadly you categorize them. In reality, of course – there are as many “types” of data scientist as there are people working in data science. I’ve worked with a lot, and have yet to meet two who are identical.
But what I have done here is separate data scientists into groups, containing individuals who share similar skills, methods, outlooks and responsibilities. Then I grouped those groups together, again and again, until I was left with just two quite distinctly different groups.
I’ve decided to call these two types strategic data scientists and operational data scientists.
Just to be clear, individuals that fall into either of these groups will doubtless have a lot in common. But in order to best examine who these two types of data scientists are, and how they bring value to an organization, it’s obviously useful to focus on the differences.
Broadly speaking, a strategic data scientist will have a firm understanding of business performance and growth, strategic thinking and communication skills, but be less well versed in the technical, nitty-gritty of setting up database systems and defining or selecting algorithms.
On the other hand, the operational data scientist is more likely to come from a background of programming, statistics or mathematics, and will use these skills to implement systems to probe and interpret the data and draw out the most relevant results.
In other words – and here we get to the crux of the difference, and see why both are essential – the strategic data scientist sets the questions, and the operational data scientist provides the answers.
In order to drive positive organizational change (or business growth) through data analysis, asking the right questions and arriving at the correct answers are both essential parts of the process. Both are equally worthless without the other.
Once organizations have both the questions and answers needed, both “types” of data scientist will then work with other members of the organization to put them to practical use.
(Some of you, perhaps if you have limited experience in the corporate world, might ask – is it not the responsibility of senior management to ask the questions? Well – unless you happen to have data scientists within senior management, which is often not the case, particularly at larger organizations – no. They will generally just identify problems).
Pair a great strategic data scientist with a great operational data scientist and you have an unstoppable team, capable of crunching their way to the most useful and innovative insights. You might occasionally stumble into someone who has the qualities to fill both roles exceptionally well – but in my experience this is rare!
Of course, it isn’t always essential to break down data scientists into these two types. Especially in smaller companies that might only employ one data scientist the distinctions will become much harder to make. Here is it particularly important to ensure any data scientist has the strategic business understanding as well as the data crunching skills.
As a final thought, I would like to say don’t be fooled into thinking that the split is along the lines of “suits” and “geeks” - this would be a lazy interpretation. Both are very much scientists, as long as they are formulating experiments in order to test hypothesis and record conclusions. Those looking to work their way up the career ladder in an organization with a large data science operation should certainly look to develop skills relevant to both top-level “branches” of the data scientist tree – while remaining mindful of where their own particular strengths lie.
I hope you found this post useful. I am always keen to hear your views on the topic and invite you to comment with any thoughts you might have.
About : Bernard Marr is a globally recognized expert in strategic metrics and data. He helps companies and executive teams manage, measure, analyze and improve performance.