Data Analysts vs. Data Scientists

The 2016 Glassdoor Report revealed that "Data Scientist "was the top job in America last year.

The report is based on anonymous reports from individuals who rate their own job experiences including how well they’re paid, treated and their career advancement opportunities. Whilst South Africa may be a little behind the States, we’re already seeing an increase in demand from our clients for Data Scientists.

Scientists, not Analysts?

Candidates have often asked what the key difference is between a Data Scientist and a Data Analyst. Firstly, the difference is important particularly when building data science teams.

For many the chief difference is that Data Scientists are better statisticians than most programmers and better programmers than most statisticians. Add to it that they have the required people skills due to the increasing importance of interaction with customers to fully understand their needs, and you’ve got an entirely different creature that needs to be recruited.

There is a lot of ambiguity within the market currently in regards to these two terms.  High calibre analytical candidates are vying for the Data Scientist title these days as they think they will earn more money and get that next big analytics job! 

I hope to dispel this myth….and also make it easier for my colleagues in the industry to recruit these two unique skills successfully.

It comes down to what they DO

Data science is such a new field that many companies will likely have differing titles for the work that Data Scientists actually do.

For the last 20 years we have worked from a clear job description for a Data Analyst or Business Intelligence Analyst, knowing well what they do and the role they perform within an organisation. But with the massive overload of data that is currently streaming into businesses daily, the importance of this role has been escalated, primarily because of the complexity of what needs to be DONE with the information, rather than simply the analysis thereof. 

To differentiate their roles, lets take a look at each:

Data Analysts focus on using SQL to chop up/slice data and to use internal analytical systems to derive meaning from the information, usually focused on understanding business performance. They typically have intermediate statistical and IT knowledge.

Data Scientists on the other hand are focused on understanding the movement of data and manipulating it freely to make sense of it now, and determine trends going forward. They do this primarily through their machine learning skills and advanced statistical knowledge.

Both however have in common the personality traits of curiosity, the desire to derive insight from information and the ability to tell stories from the numbers.

The diversification of skillsets is primarily due to the change in what organisations need and how this has adapted the structure of the organisations themselves. In the past, Data Analysts would find themselves largely isolated, a division of their own, whereas now Data Science teams are being incorporated into various departments within an organisation.

Unique Skills Required

A Data Scientist should possess a wide breadth of abilities including academic curiosity, storytelling, product sense, engineering experience and an acute analytical horsepower capability: cleverness!

The fact that these individuals are now expected to function within diverse departmental teams requires them to have greater levels of people centricity and high levels of communication. Crucial, is their ability to shift gears from highly analytical language to presentations that take technical jargon from “goobledy gook” to business relevant stories that drive improved performance and business results.

Let’s look at each of these areas in greater depth:

Academic curiosity: The desire to go beneath the surface and distil a problem into a very clear set of hypotheses that can be tested.

Storytelling: This is the ability to distill a quantitative result from a machine learning model into something (be it words, pictures or charts) that everyone can understand immediately is actually a very important skill for Data Scientists.

Product sense: Is the ability to use the story to create new products or to change an existing product in a way that improves company goals and metrics. It moves beyond simply building a data filter or a data engine that spews out information to knowing how to transpose what the data is telling the organisation into a product, capitalising on trends and turning these into profits. 

Statistical and machine learning knowledge: This is the domain knowledge required to acquire data from different sources, to create a model, optimise it and validate it.

In a nutshell, to be considered for the role of a Data Scientist the individual should know, if nothing else, how to access information, mix it, clean it, filter it, mine it, visualise it and then validate it for the purposes of improved business performance.

Lastly it is important to mention that our clients have changed their interview techniques in order to more effectively assess this mix of new skills mentioned in this article.  We have seen that the use of case studies and problem solving thread type questions is the preferred method of assessment today.

Of course, this method requires a new way of engagement including lots of preparation from both the candidate and the recruiter working within such a specialised field. 

We are proud to offer our clients and candidates value added services including coaching, assessment preparation and validation of alignment between candidate aspirations and client requirements.