According to research conducted by Accenture, 80 percent of the data scientist positions that were created in 2010 and 2011 in the U.S. have not been filled. Indeed, the problem has become acute for many companies as the number of graduates with the required skills isn’t keeping pace with rising demand. In order to meet the demand for these skills by employers, the U.S. will need to increase the number of graduates with requisite skills by as much as 60 percent, according to McKinsey Global Institute.
Not all companies have the deep pockets or resources to engage in a war for talent. According to a 2013 Data Science Salary Survey conducted by O’Reilly Media, the average respondent had a median income of $100,000 and had experience using 10 open source and commercial tools such as R and Python. Those using 15 or more tools had a median annual income of $130,000.
To help tackle the skills shortage, some companies—such as AIG—are taking a creative approach by building a “science team” of statisticians, business analysts, project managers, systems architects, and engineers for its property and casualty business. AIG has assembled a team of people who individually lack the full skill sets of data scientists, but who collectively possess them all. For its part, Monster Worldwide has amassed a team of “data crunchers,” statisticians, business analysts, computer scientists, and “navigators” who can explain analytical findings to managers, according to Accenture.
Because so many companies are hard-pressed to find and retain the data science talent they’re looking for, there’s a growing movement afoot for applying crowdsourcing to tackle work demands. Organizations that extend analytical opportunities to a collection of data scientists can utilize the collective insights of a group. Meanwhile, companies that tender data science competitions can incent participants to take part in contests to work on challenges as they vie to create solutions that outperform their peers.
For earlier insight on data scientists, check out this post on whether the need for data scientists will die out.