olly - Fotolia
Data scientists are some of the most sought-after professionals today, and for a good reason; they are essential for transforming the growing volume of raw data into actionable information for organizations across all industry verticals.
Despite this, enterprises struggle to understand what educational or work backgrounds a potential data science hire should have before bringing them onto the team. The answers vary, even as organizations' need for properly utilized data does not.
Data science job requirements
Organizations invest heavily in data as a corporate asset, and there is more data available to them than ever before, said Khaled Ghadban, director of analytics and AI services at Alessa by Tier1 Financial Solutions.
With that in mind, "data science is quickly becoming something that's more than just a 'nice to have' but more of a 'need to have' now," he said.
Organizations typically want potential hires to have a high level of education and work experience that enables them to turn theoretical approaches into real-world problem-solving.
"An education in statistics, math or computer science along with training in data mining, data analytics, internal problem-solving has to be combined with experience," said Tendü Yogurtçu, CTO at Precisely, a vendor of data integration, quality and enrichment software, and herself a recipient of a doctorate in computer science.
How to hire data scientists: Necessary skills and education
When hiring for a data scientist position, most organizations seek candidates with at least a master's degree in data science, computer science, math or a related discipline. Some hiring managers want candidates who have even more advanced levels of education, namely doctorate courses or degree.
Consider the observations from Robert Half Technology, which shared its career insights in a 2020 online post: "While a Ph.D. or other advanced degree may not be essential to get hired for an entry-level data scientist role, it is likely to become more important as you look to advance in your data science career."
Moreover, enterprise leaders said they want their data scientists to continue to advance their learning even further while employed.
"One thing we know about this space is it's very rapidly evolving, so the ability to develop one's self is a key skill," Yogurtçu said.
Yogurtçu said she assesses candidates on their willingness to keep learning. She said her company offers educational opportunities, such as peer training and online learning platforms, to help its data scientists further hone their skills. This helps them boost their careers, while also advancing the organization's data program.
Data scientists don't need to look far to find supplemental training -- numerous online learning platforms, including EdX and Udemy, offer free or inexpensive data science, AI, and analytics courses. Many technology vendors also offer free online training sessions for people looking to learn their data science tools.
In fact, ongoing training is so important in the data science field that organizations sometimes see a willingness to learn coupled with the ability to put that learning into practice as enough for professionals from other disciplines to land a data scientist position even if they don't have advanced degrees in the field.
"I don't think there is a 'best' set of credentials for data scientists. I know people with master's degrees in data science and I know people with certificates from boot camps," said Pat Ryan, principal architect at the consulting firm SPR.
"Any degree of credentials will open doors, so the best set of credentials for you depends on the career opportunity you'd like to pursue," he continued. "A master's degree is best if a person wants to be more academically focused in data science. However, data science boot camps are often great for those who just want to be amazing at their jobs as practitioners."
Experience is essential when hiring for senior data science positions
While a candidate's education in important, most organizations want to fill more senior data scientist positions with candidates who have real-world experience.
Pat RyanPrincipal architect, SPR
"For entry-level positions, real-world experience doing data science is going to be optional. It's going to be more practical or academic backgrounds that would qualify you for the position," said Scott McClellan, senior director with the data science product group at Nvidia, a manufacturer of graphics processing chips.
"But for those advanced positions, you need that experience of shepherding a data science project into product. The skill sets for moving projects into production that are needed for those senior positions are more learned on the job than in academics," he said.
McClellan said he believes that a data scientist with such experience can advance their career just as successfully as those who have advanced degrees.
"Everyone is going to say they want a Ph.D., but I would skip that in favor of experience, and so would many other managers," he said. "Some of the most advanced data scientists I've had work for me, what made them successful and what made them so valuable was their solid experience and their solid knowledge of systems -- [for example], networking, storage, CPUs, GPUs, what causes bottlenecks, how you get past scaling issues."
Data scientist experience requirements
Others also said they, too, look to hire or promote data scientists into senior positions based on their work experience, and find industry and business experience particularly valuable.
Given the complexity of data and its ability to tell a story, data scientists first need to understand the potential of any set of data, Ghadban said. This can help organizations drive a roadmap and data strategy.
"Products are also becoming more sophisticated in the way they leverage, organize and visualize data so that users are able to make more informed, strategic decisions," he said. "Experience with different types of data products is playing a central role in the evolution of this type of career, along with industry and consulting experience."
Indeed, sources agreed that those data scientists with experience turning data into information that drives good outcomes for an organization will most likely go the farthest in their field.
"Since a data scientist is a combination of technical skills and business domain expertise, a person who can talk to the business [side] and understand the business problem along with the math and science aspects is well qualified to be a data scientist," Ryan said. "What gets a data scientist to be promoted is how useful a model of insight is to the business and how the business is impacted that can translate into a data scientist being promoted."