This post is an adapted excerpt from our newly-released report, The Burtch Works Study: Salaries of Data Scientists 2017, which examines updated compensation and demographic data for data scientists.
Download the full study to see how data science salaries vary by experience level, region, industry, and residency status, plus our insights on how these trends will shape the future of the market.
With more companies developing quantitative teams, more professionals entering the field of data science, and more blending between data scientists and others in predictive analytics (trends that we covered more extensively here), there are several further developments expected over the coming years:
Increasing Specialization of Roles
As additional companies cultivate data science and analytics teams, and many established teams grow larger, data science roles are starting to become more specialized. Larger teams are able to focus on hiring specialists who can collaborate, and therefore concentrate on fewer tasks, rather than all tasks at once. A firm hiring its first data scientist may, however, still seek the well-rounded (and expensive) “unicorn” who is able to complete every task including building infrastructure, data acquisition, modeling, communicating with the C-Suite, and more.
Many companies used to require that all of their data science hires had PhD’s, with a deep background both in math/statistics and computer science. Now, it is becoming more common to see many professionals on a large team with differing backgrounds – some in computer science, some in math and statistics, and others in engineering – all working together on the same data science project. With many quantitative teams continuing to grow and the data science “unicorns” being so difficult to hire, we expect the trend of specialization to continue.
Blending of Predictive Analytics & Data Science
This specialization is also somewhat influenced by predictive analytics professionals, many of whom have begun picking up the computer science skills necessary to transition into data science roles, which have a heavier focus on coding and unstructured data. The bigger teams are creating an environment which can support data scientists that may have less experience with typical data science tools, since they can work together with other team members, such as data engineers, to fill in the gaps.
We anticipate that, as the two disciplines continue to blend, the suppression of data science salaries will continue (you can download our study for more info on that) as they converge with predictive analytics salaries. Given the continuing popularity of the “data scientist” title, it’s possible that data science teams of the future will be home to modelers, data scientists, statisticians, and data engineers, and all of them will be paid on the same corporate scale. At some point, there may be no clear distinction between data scientists and other predictive analytics professionals in pay or in title, but it will likely be a number of years until we reach that point.
Data scientists and their teams must remain aware of the rapid and constant evolution of data science technology and tools. Continuous education for these professionals is absolutely critical to success, since new developments are being introduced much faster than in years past. We have always advocated that quantitative professionals stay close to the data and keep their skills sharp, and, now more than ever, it is crucial, especially for data science management. Data science managers who transition away from hands-on work altogether may quickly find their knowledge out-of-date and their skills irrelevant.
As always, we will continue to keep a close eye on developing trends in the data science and analytics hiring market. For more of our insights between the annual Burtch Works Studies, keep an eye on the blog!