In-demand data and analytics skills to hire for now

As the quantity and quality of information improve, managers face increasing pressure to leverage their data to be competitive in the digital era, launch new customer strategies, and get the most out of their teams. Without the right talent analyzing that data and translating it for decision-makers, companies will be left behind.

“Everything starts with the data,” said Nadine Kawkabani, global business strategy director at MFS Investment Management and an industry host for MIT Sloan’s business analytics capstone program.

When she joined the program in 2018, the line she often heard was that data was doubling every two years. Now, it’s every 18 months. Because the number of people hired to work with that data isn’t moving as fast, Kawkabani said, the question becomes, “How can we be more efficient at bringing up the trends or what the data is telling us, or how can we rule out the noise?”

Kawkabani and hiring managers from Comcast, Netflix, and Pfizer shared the technical skills that are essential at their companies and the soft skills they look for when it comes to filling today’s data-centric roles.

Python, R, SQL, and “the right math”

While acknowledging the proliferation of data and analytics undergraduate and master’s programs, the hiring managers said they don’t require an advanced degree when looking for a data expert. But they all agreed that fluency in basic programming is a common denominator for anyone considering a data job.

“Whether it’s an engineer or data scientist or research scientist, SQL or Python are the required programming language — Python or R, depending on the candidate’s preference,” said Yichen Sun, SM ’13, who leads a team of engineers and data scientists at Netflix.

Sun said she also looks for proficiency in basic sampling techniques like A/B testing, and causal inference techniques like difference in differences.

Trace Hawkins, senior vice president of strategic analytics at Comcast, encourages people who know SQL to start learning Python and vice versa, though he said he can find a role for someone regardless of which programming language they prefer. What’s nonnegotiable is the way someone interprets and analyzes the data.

“In Python, you can generate a match pair comparison population, but do you know how to do it right? Do you actually understand the difference between a good match pair and a bad match pair? And how would you evaluate your clustering algorithm for whether your segments are mutually exclusive and collectively exhausted?” Hawkins asked. “All of the things you might do there, you need to understand methodologically how to validate that your math was the right math.”

In search of unicorns

Hawkins and Jonathan Lowe, the data science lead for Pfizer Global Supply Operations Insights, both said they look for unicorns — not $1 billion companies in this case, but data experts with coveted second skill sets to apply at their companies.

Hawkins looks for data workers who can translate their findings to a business audience. Lowe said his “super unicorns” are the data scientists who also happen to have consulting skills and love developing software.

“There’s a fourth category, too, which sometimes we make an exception and hire for even without the other [skills], which is domain expertise,” Lowe said. “If somebody says, ‘I’ve worked in a quality lab for half my career and now, for the last several years, I’ve been learning more data science,’ we will gobble those people up.”

State-of-the-art technologies aren’t always the best solution in the Netflix production environment, where Sun’s team needs to consider computational cost, consumer experience, privacy requirements, data infrastructure readiness, and more.

“We therefore need someone to be both principled and practical, make the right trade-offs, and to be able to articulate the ‘why’ behind such technical decisions,” Sun said.

Communication, curiosity, collaboration

Bridging the gap between the business and data sides of a company are top priorities for hiring managers, with each emphasizing the importance of accurately translating the information gleaned from data into actionable business strategies.

“Storytelling skills would be another way to describe this capability,” Lowe said. “[Don’t] just blurt out a bunch of technical jargon but tell a story around why the business needs this [data] support and what will happen if the business uses what you’ve built.”

Today’s data-centric roles also require curiosity, which contributes to an innovative and problem-first mindset. While a data expert with a solution in search of a problem isn’t a deal breaker, Sun said she will try to coach the person into understanding that their solution might be the right application for a problem but that there might be an “even more elegant or even simpler way to do it.”

Relatedly, Sun also looks for “someone who’s more reflective, who is able to receive this feedback in a very productive way and be adaptable in terms of what approach they use.”

These and other soft skills are examples of how data and analytics jobs — and the related culture — have changed, Kawkabani said. It’s no longer about handling data with blinders on; it’s about ensuring that the data makes sense and that the people who are touching the data also understand how they’re impacting the strategy of the firm.

“We’re all relying on each other,” Kawkabani said. “I can put the best strategy out there, but if I don’t have good data, nice graphs, accurate data, and timely, interpretable data, it doesn’t mean anything.”

Read next: What is synthetic data — and how can it help you competitively?