by Stephanie Collis
Head of Marketing & Communications, Zetaris
Follow me on LinkedIn & Twitter
Data science is a victim of its own success. Just like any other corporate fad, the field has become one of those things that is sometimes signed off on without close scrutiny from decision-makers who are more focused on portraying their company as a progressive market-leader. This is a tragedy, as data science can offer significant benefits to businesses across a wide range of industries when implemented correctly, potentially saving organisations significant amounts of time, money and labour. An extreme example came from management consulting firm McKinsey & Company, who estimated that big data initiatives could cut the US healthcare spending bill by as much as $450 billion. The same is true at any scale – when done right, big data cuts inefficiencies and wastage, letting you reroute resources back into growing your organisation.
However, this eager dash by some companies to implement big data without fully understanding it does raise the question – what makes a data scientist? This is an open question with a lot of potential answers, but in this blog, we’ll explore what we at Zetaris believe makes someone a data scientist, and examine what they could do to become more effective in their role.
About more than just the tools
Like revheads at the race track, tech-savvy decision-makers looking into data science can sometimes get so swept up in the range of tools on offer that they forget about the people they need behind them. So much focus is placed on machine learning systems and analytics infrastructures, with many business-owners seemingly happy to put state-of-the-art tools in the hands of amateurs. But remember, holding the keys to a cutting-edge analytics data lake does not a data scientist make.
One possible factor that separates a true data scientist from someone who just happens to leverage analytics in their work is in their ability to create value from data. While people frequently talk about ‘mining’ data, it’s not the best possible metaphor. A more accurate way of visualising the process of wringing value from data is refining oil – you start with a large quantity of raw material that could all potentially be valuable, and through a lengthy and technical process, transform it into something with commercial value or utility. A data scientist is someone who can confidently approach this great mass of raw material and give it value – whether that’s turning it into a saleable product, leveraging it to build a closer relationship with clients, or using it to plot a strategic course for the business through the market. Just as to be an artist someone has to create art – but can do it through sculpture, painting, performance, video or myriad other mediums – a data scientist has to create value, but how it is created and the nature of that value is up to them and their company.
In a fast-paced field with new innovations arising seemingly every day, data scientists need to contain multitudes. Data science is an exploding and highly fluid area of practice, with new fronts being opened up all the time. Data scientists need to be across all of this, having at least an awareness of new tools and techniques as they come out and a deep knowledge of their area of practice, coupled with a willingness to incorporate the former into the latter where required.
This said, the warning about focusing too much on tools also applies here. Just because there is a galaxy of possible ways to ‘do’ data science does not mean all of them are valuable to your needs. Finding something that works and being flexible in your approach is far more important than having certifications in every possible programming language and platform. Writing for KDnuggets, Head of Data Science at Amazon Karolis Urbonas warned against getting too tangled up in new tools.
“People tend to focus on tools, processes or – more generally – emphasize the form over the content,” he wrote.
“These are just instruments that are used to solve problems.
“The core function of any data-driven role is solving problems by extracting knowledge from data. A great data scientist first strives to understand the problem at hand, then defines the requirements for the solution to the problem, and only then decides which tools and techniques are best fit for the task.”
Data scientists need a broad awareness of the market as a whole, and a depth of knowledge about their specific area of practice, but this need should never overtake their ability to solve a problem.
Be your oracle
As with any specialist, data scientists need to be able to relate what they’ve discovered to laypeople. Unlike other areas of technology, data science is completely opaque to many people, meaning that practitioners need to be able to not just extract insights but explain them to decision-makers in a way that they can act on, as otherwise their discoveries are just inert pieces of information. FICO chief analytics offer Andrew Jennings wrote that in many cases, picking strong communication skills over cutting-edge technical prowess is often a good trade.
“Making that trade-off between best-in-class technical skills and strong communicators who can help translate the highly technical information into language that a business user can understand is a trade-off worth making,” he wrote.
“Also, going in reverse, those same people need to be able to translate a business need into an analytics investigation.”
Know how to spot a true data scientist and your business will flourish. If you’d like to learn more about where the field of big data analytics could take your organisation, speak to the team at Zetaris today.
FIND OUT MORE ON BLENDING REAL-TIME AND DEEP LEARNING BASED HISTORIC ANALYSIS OF BUSINESS DATA - DOWNLOAD THE PAPER