According to technology news, the role of data scientist is one of the most rapidly growing careers today. The need for data scientists is projected to grow exponentially in the future. But what do data scientists do?
Well, as a recent article in Harvard Business Review pointed out, a great deal of what they do is finding, organizing, and cleaning data. Eighty percent of their time is spent doing this; the remaining 20% is spent analyzing that data.
Products Across the Spectrum
Sure, but when people ask “what do data scientists do?” they often aren’t literally asking what a data scientist does at the desk. They are asking, or meaning to ask, a larger question related to business strategy: what do data scientists do that will help business leaders achieve success?
Well, despite the association of data science with Google search engine optimization and honing the capabilities of self-driving cars, data science is used in all sectors, from health to transportation.
One of the easiest ways to see what data scientists do is talk to individual data scientists, as a podcast titled DataFramed does. One example is worth a thousand words, and multiple examples provide a sense of the reach of data science in business.
Data scientists, for example, are producing the methods by which online publication Buzzfeed produces such catchy headlines. They are building experimental product development systems for retail and travel endeavors like Etsy and booking.com. At Airbnb, they are working on models for machine learning. At Starbucks, data learning systems make recommendations.
But the reach of data science is not restricted to online publications and platforms—far from it. Data science has great potential to change paradigms in trucking across North America. Data science is enormously helpful in research on how to fight cancer. Companies like Alluvium use data scientists to comb their industrial operation data streams into insights.
Data scientists spend 80% of their time finding and organizing data.
Communication Skills Are Highly Important
And perhaps this multifaceted use of data scientists explains why 80% of their time can be spent in finding and organizing data rather than analyzing it. Data has to be, in many instances, pulled from data streams that can be large and abundant, but not necessarily targeted to the business goal.
After that, data scientists experiment to find the best methods for analytics. They may create pipelines and personalized streams of data to see how machines best learn, what gives the best data for a desired goal, and what the results are. Infrastructure, experiments, decision-making, and products all come into play.
What does that translate to in terms of the products of data scientists? It means they will produce dashboards, reports, data visualizations, and statistics.
In fact, at least one data science leader, when asked whether deep learning expertise or PowerPoint ability was more important, actually choose the latter, because communication of data science findings was, he felt, of paramount importance.
But lest that sound surprising, it’s also true that machine learning and other data science functions are increasingly being automated. Machines will help data scientists guide them in the future even more than they do now. So communication abilities will remain important, as well as learning new techniques and approaches quickly.