REPLACE Big Data Show logo
April 30 - May 1, 2025
North Javits Center | New York City


Aug 09, 2023

Upskilling Existing Staff May Be Best For Filling Data Science Positions

By D.J. Murphy, Senior Editor, Digital Content for Data Universe

Ask any hiring manager or HR professional in any business vertical about the difficulty of recruiting quality talent to fill open roles at their company and you will likely encounter grim expressions. For tech-based roles and businesses, the problem is not new. In a 2019 PwC report, the share of global CEOs who identified lack of available skills as a threat to their business had been rising steadily throughout the 2010s (from 53 percent in 2011 to nearly 80 percent in 2019). And, since the pandemic made every company a tech company, the challenge has only intensified.

For data science roles specifically, the problem is even more acute. According to the U.S. Bureau of Labor Statistics, between 2021 and 2031 the number of data scientists will increase by 36 percent, much faster than the average for all jobs. As increased utilization of AI automates many jobs out of existence, recent Pew research indicates nearly one-fifth of all U.S. jobs are at risk of being replaced or assisted by AI), data professionals will be in higher demand.

And, while undergrad and graduate degree programs in data science, data analytics and other related fields are proliferating quickly, those programs are relatively new and are not turning out trained graduates fast enough to fill all the available positions.

But where can businesses find the rare candidates to tackle these increasingly numerous and vital openings?

“It isn't clear that traditional recruiting methods are failing,” says Adam Ross Nelson, a career services provider and author in the data science space. “Instead, I think it is clear that simply recruiting folks with degrees in data science will not be sufficient. Instead, employers will need to build and recruit talent from multiple sources and through multiple tactics and strategies.”

While identifying and adding external candidates will always be important for ensuring a dynamic work environment, as companies become more tech-centered and automation begins to assert itself, there may be more candidates already working within your walls than you think.

Benefits of Looking Inward

Companies have always reorganized, shuffled personnel and promoted from within, but rarely have those employees needed upskilling to the extent prospective data scientists do. And retraining employees to transition from one expertise to another can be expensive. So why do it?

Aside from an obvious way to address a serious shortage of an increasingly important resource, Nelson says there are other benefits that accrue to a company that shifts workers for this purpose.

“The benefits of growing internal employees into data science is that the internal approach will mean the new data scientists have pre-existing knowledge of the company, its operations, its challenges, strengths, weaknesses, and opportunities,” he explains. “The internal approach will mean that those who grow into data science have insights as to how the company can use data science.”

Not having to go through the onboarding process can mean an important team member can be up to speed in a new job faster. And existing familiarity with corporate culture can ease the transition, too.

Look Beyond Data Science for Data Scientists

But how do you identify possible candidates who are already in other positions and may not be thinking about transitioning to data science?

Nelson notes that because it takes time for colleges and universities to recruit students, enroll them, and graduate them in large numbers from majors that have not existed long, recruiting young talent out of schools can only be part of a hiring strategy. But people who have studied statistics, computer science, economics, engineering, math, and other fields are already well-prepared or likely possess good baseline skills to work as data scientists.

Many people like this could already be working for you. So, assess your current staff based on their educational background for potential data scientists. And once you find them, let them know about the opportunity.

How Upskilling Starts

Growing an internal employee into a data scientist is not inexpensive. It can take weeks of nearly full-time commitment for an internal employee to fully train. So making sure an existing staff member is interested and capable is a prerequisite.

Establishing and effectively promoting internal training programs are a good place to start. Some of the best at retraining current staff for data science offer full-scale, bootcamp-style data science educational programs for employees.

Nelson says employers can model internal data science bootcamps after existing online programs, or contract with third parties that run them professionally, such as Flatiron, General Assembly and others.

“I'm a fan of the work these organizations have provided the field,” said Nelson, who is an instructor for Data Society, another company in the space providing educations businesses can employ. “They specialize in providing corporate training, including smaller days, or week-long programs in addition to full-scale multi-week or multi-month private internal bootcamps.”

As with any new training vendor, employers need to sort through the costs and benefits associated with providing this training to current employees over the costs of recruiting new talent. Nelson thinks no single strategy will provide a comprehensive solution and employers will need to pursue both strategies.

Open to Everyone

While an educational background in a related field is a plus, Nelson believes a commitment to learning and a true interest in data science is more important—don’t discount an internal candidate who seems passionate just because they don’t meet certain criteria.

“The field of data science is open to anyone who is willing to put the time and effort into learning the skills,” he concludes. “You have to learn math (but often not as much as is thought necessary), you have to learn computer programming, you have to learn statistics, and you have to learn both the predictive and business theory rooted in data science that can make predictions.”

Committed and engaged learners exist in your company and, given the right training, they can help you address a challenging hiring environment in data.

For free resources, blog posts and advice on starting or enhancing a career in data science, visit