Hard Data, Soft Skills: How To Become a Great Data Science Leader

Moving into a managerial role as a data scientist isn’t just about having great tech skills. Here’s what you need to know to advance your career.

Written by Adrienne Teeley
Published on Apr. 20, 2022
Hard Data, Soft Skills: How To Become a Great Data Science Leader
Brand Studio Logo

After years of making sense of complex information, data scientists looking to move into leadership quickly find that their career of parsing information into digestible insights isn’t over — but it does look different. Instead of working with granular data, building pipelines and extracting insights, managing a team requires a set of skills that can be just as difficult to acquire as tech acumen.  

Being able to communicate effectively, for example, is vital but tricky to master. According to XSELL Technologies’ Head of Data Science Ishu Jaswal, being able to clearly communicate doesn’t just mean speaking in technical terms to a knowledgeable team. It means being able to convey complex information to non-technical audiences, like cross-functional teams and company leadership.

“People skills are important to hone in order to avoid misunderstandings, resolve conflicts, and influence opinions,” Jaswal said. “Effective communication is one of the most important skills that data science leaders need to learn and apply.”

Of course, new leaders don’t wake up and intuitively know how to be amazing in their roles. But for data scientists hoping to move into management, it can be helpful to understand exactly which tools to sharpen pending a promotion. That’s why Built in Chicago connected with Jaswal and leaders from Nordstrom, Walgreens and CNA to hear how they personally stretched their skills to better support their teams, as well as what advice they have for others looking to do the same.  


Meghan Hickey
Senior Manager Data Science & Analytics, Merchandising • Nordstrom


Nordstrom is an enterprise retail company that strives to use technology and data to improve the way consumers shop for apparel, cosmetics, accessories and more.  


What appealed to you about managing a data science team? 

I come from a non-traditional background, and most of my technical skills were self-taught. When I first started at Nordstrom as an entry-level data analyst, I found it exciting and rewarding to dive into complex problems and use data to come up with recommendations or solutions. Working in data science at Nordstrom allowed me to quickly work on high-impact projects that were driving measurable improvements in our business.

As I grew, I realized I really enjoyed helping others discover the possibilities that working as a data scientist could unlock. Ultimately, I wanted to move into a manager role so I could help beyond a project level, and really influence the strategic direction of what kinds of work the team would take on. Today, the most rewarding part of my job is being able to match my team members to projects that have a high business impact, align with their goals and help them stretch or grow in new ways. There are so many different, exciting projects going on at Nordstrom where data science can play a huge role in improving our business. 


What Nordstrom’s data team is working on

Currently, Hickey’s team is developing machine learning models that predict which items the company should mark down, when and by how much. “We created several forecasting and optimization models to see which had the best accuracy,” Hickey said. “By optimizing our markdowns, we ensure we’re offering our customers the best price at the right time.”


What skills do data scientists need to develop when they move into a management role?

Moving into a management role forced me to focus on my non-technical skills in order to best support my team, a shift from when I was a scientist working on growing my technical skills. I found I had to refine my empathy and influence. Working to always lead with empathy helps me to better understand the perspective of those who report to me, my stakeholders and our customers in order to provide them with the best experience. Prioritizing being mindful of all the complexities within the problems we face helps me recommend better solutions and guide my team members to do the same. It also sets a tone for openness and understanding across the team.

A lot of our work as data scientists is to develop new, innovative solutions to problems, and we owe it to our stakeholders to be able to explain how each project can improve their lives and the lives of our customers. We need to stick up for our work and prove it’s worth prioritizing and investing in when there are lots of other meaningful work happening across the company. As a manager, the success of your team’s work is dependent on your ability to help others clearly see the value of what you are working on and buy into your vision. 

I’ve now been a part of three different functions in our data science and analytics organization, supporting marketing, digital and merchandising. In all of them, the ability to tell a story around your team’s work has led to better outcomes for the team and our partners.

The success of your team’s work is dependent on your ability to help others clearly see the value of what you are working on.


What other advice would you give to a data scientist who is managing a team for the first time?

Be patient with yourself and your team. Switching into a people management role takes getting used to, especially for those of us used to doing technical work all day. Invest time up front into getting to know your team and what they are working on. Learn about how they like to work, what they like about their job and what they’d like to see change. 

It’s a big switch to go from doing all the work yourself and being deeply immersed in the data to zooming out and offering support and guidance to your team on how to solve problems. The more you can learn about your team’s projects, the more it helps build trust and ensures you can deliver the best results for the customer.



AJ Udechukwu
Vice President, Data & Artificial Intelligence • Walgreens


Walgreens is a pharmacy and consumer goods company operating in more than 9,000 brick-and-mortar locations across the United States. The company has a robust digital presence online and through the Walgreens app. 


What appealed to you about managing a data science team? 

I love seeing growth and transformation in processes and business functions, but also in people. Managing a data science team affords one the opportunity to shape and mentor other practitioners. There is also a special bond between former hands-on practitioners and their teams. The leader is typically better equipped to understand the challenges of the team, what’s possible within given timeframes and more.


Three main skills needed to grow into a management role at Walgreens

  • Understanding business drivers, like how your organization makes and spends money, and how the business is funded
  • Team building and assembling the best team possible given particular constraints, like geography, compensation and talent availability
  • Storytelling


What other advice would you give to a data scientist who is managing a team for the first time? 

Don’t expect individual team members to be masters of every skill — build the team as a whole. If you’re starting off with a small team, hire talent who complements your background.



CNA logo on the outside of the office building


Bill Stergiou
AVP, Data Science • CNA


CNA is a property- and casualty-insurance company designed to support businesses of all sizes. 


What appealed to you about managing a data science team?

I enjoy working with an extremely talented team. This data science group has diverse backgrounds, from actuaries to teachers to linguists, which enables us to get different points of view on our projects. 

Our team is lucky to be able to work on interesting challenges that require out-of-the-box thinking. For example, we create a lot of our own statistical techniques that are tailored to the problems we are trying to solve, instead of relying on standard statistical packages that don’t always fit our needs. The solutions our team delivers have a significant impact on our business partners’ processes, our insureds’ satisfaction, and ultimately, CNA’s bottom line.


What skills do data scientists need to develop when they move into a management role?

Data scientists at CNA require a suite of skills that enable them to manage a technical team and work closely with business partners to determine which projects to pursue, then incorporate their ideas and ensure that models have value once they are implemented. I believe there are three main skills that best enable these goals.

First, management skills are necessary for leaders to promote individual development through project work and outside training; give individuals ownership of project work — including presentations to senior leaders — listen to advice and concerns; and remove blockers.

Communication skills are needed to best understand the problem that needs to be solved and how they can contribute to the solution. Then, managers need to communicate with the team to achieve a robust, efficient solution.

Finally, statistical and machine learning knowledge is key. Managers need a deep understanding of the best techniques to impact the power of the solutions we offer our business partners, and the ability to communicate the benefits of certain techniques to a non-technical audience.

I view my role in managing a highly technical team as enabling their success.


What other advice would you give to a data scientist who is managing a team for the first time? 

Data science managers need to actively balance the management of a technical team to ensure a top-rate work product, while engaging business partners to ensure the products being built are effective and valuable. 

I view my role in managing a highly technical team as enabling their success. My job is to remove any blockers to the team’s success, whether it be organizational, technical or statistical. I make sure that each team member has ownership over a project or major project component that they can sink their teeth into, become an expert in and present their findings to senior leaders within CNA. 

At the same time, I work with business partners to help them understand the processes that our models impact through shadowing, aligning on project goals and scope, and soliciting ideas to maximize the effectiveness of the model. The more business partners who are involved and have a meaningful role in the modeling process, the more likely they are to derive significant value from the model.



Ishu Jaswal
Head of Data Science • XSELL Technologies


XSELL’s platform uses natural language processing technology to provide real-time coaching and support to contact center agents to improve sales and customer service. 


What appealed to you about managing a data science team?

I like being challenged in my work and working in a role where I can continuously add value to the company’s mission. That being said, whenever I think of the term “manager,” I think of a leader who is a great coach and can develop others. Taking my current role gave me a chance to develop leadership skills and made me a key stakeholder at the organizational level. It also gave me the ability to be a decision-maker in defining the department’s strategy and developing ideas that can make an impact on the business.

Since I was a core member of the data science team at XSELL before being promoted, I knew the gaps that needed to be filled and the bottlenecks that needed to be resolved. This opportunity motivated and enabled me to solve these challenges. Overall, it seemed like the right next step on my career ladder, and I’m really enjoying this role.

As with everything, there is a learning curve to being a good leader.


What skills do data scientists need to develop when they move into a management role? 

I strongly believe that to lead a data science team, one should have a good technical skill set, either in general or related to a specific domain, and should understand what an end-to-end data science process looks like. On top of that, a data science manager should be able to balance the technical nuances across domains of data, math, statistics, machine learning and software, and connect them to business context and value to the organization.

To excel in the management role, one needs to be a great leader. I continuously work toward better understanding the strategic objective of the company, because a leader should understand what the company vision is and how your and your team’s role aligns with it.

I also strive to think about the big picture. As an individual contributor, one thinks about their current work and how that work aligns with the project goal. But as a manager, you should be able to take a couple of steps back to see how a piece of work aligns with company goals and prioritize accordingly. This way, one should be able to put a framework to an unstructured problem, identify the gaps in the current process and identify the next steps to achieve the goal.

In business, data science leaders need to be proficient at analyzing priorities and deadlines and then clearly and fluently explaining the steps and processes to achieve the work to both technical and non-technical audiences. This critical element helps you collaborate well with cross-functional teams, as well as translate the impact of the work to business leaders in terms of value.


Resources to strengthen leadership skills

Jaswal believes that data science leaders should work toward developing their leadership and mentorship abilities to improve how they coach, communicate and listen to their teams. While leadership seminars and workshops are helpful, there are a few books he recommends as well, like, “The 15 Commitments of Conscious Leadership” by Jim Dethmer, “The One Thing” by Gary Keller, and “On the Edge” by Alison Levine.


What other advice would you give to a data scientist who is managing a team for the first time? 

A good manager is vision-oriented and should be able to take responsibility for the team to deliver value and outcomes back to the organization while retaining and developing great talent. Always listen to the team before making decisions, coach the team, empower them and inspire them to develop their skills. 

Be curious and open to learning new things. As with everything, there is a learning curve to being a good leader. Do not feel ashamed of asking for help — there is no shame in not knowing what you don't know.

Build trust by having candid and transparent conversations with your team. Engage your team in making critical decisions that help them feel empowered, ask for their feedback and take action based on that feedback to improve things.

Be a mentor. No one likes a commanding and controlling boss. People care about leaders who coach them and develop them. Do not try to convert one’s weakness into strengths, but instead focus on understanding your team’s weaknesses and maximizing their strengths. Only then will you be able to build a strong team.

Do not stop your data science learning process. Management work takes a lot of time, but continuing your learning in the data science domain is the key to being successful at growing yourself and your team’s skill set. Always try to stay up to date with the industry and state-of-the-art architectures to be aware of what is available in the market. 



Responses have been edited for length and clarity. Images via listed companies and Shutterstock.

Hiring Now
Cloud • Fintech • Software