Being a boss is hard. Overseeing a data science team can be especially challenging: Roles are still in flux, turnover is high, and companies are ironing out the best ways for teams to function. And being a technical whiz doesn’t necessarily prepare you to manage others.

On a panel discussion at the recent Rev summit for data science leaders, three experts shared their tips for hiring, retaining, and nurturing data science talent. Whether you’re managing a data science team today, preparing to launch one, or hope to do so in the future, their insights will make you a better leader.

Hiring the right people

  • Recruit leaders first. If you start with a junior hire or someone fresh from academia, they’re likely to feel lost and frustrated without mentorship. Michelangelo D’Agostino, Senior Director of Data Science at ShopRunner, suggested installing a more experienced person first to give the team direction.
  • Choose managers carefully. Don’t just focus on technical talent and experience. The panelists agreed that humility, curiosity and an ability to listen and take feedback are crucial traits for a senior role. “Someone who’s going to be in charge has to know they don’t have all the good ideas or all the answers,” D’Agostino said. He suggested asking prospective hires to describe a situation in which they failed and how they would avoid repeating it to gauge capacity for self-reflection.
  • Rethink data challenges. Given the competitive hiring landscape, onerous take-home tests can screen out qualified candidates and create a tense, exam-like atmosphere. You don’t need these challenges to make strong hires, said Patrick Phelps, Lead Data Scientist at Insight Data Science who previously led data science teams at Yelp and Pinterest. “It’s really hard to scale…[and] it takes a huge amount of time to grade,” he said. “I’d rather just put a good data scientist on my team in a room with them for an hour.” If you do include a challenge, D’Agostino suggests having candidates complete a coding exercise in the office and talk through it as in an informal code review.

Retaining your talent

  • Don’t oversell the role. Half of data scientists stay at their jobs for two years or less. To reduce turnover, be truthful about the position you’re hiring for, advised Conor Jensen, a customer success manager at Domino. “Be very realistic upfront about what the role is, what the pain is going to be, where you think the impact is going to be, and what the timeline looks like,” he said. “A lot of times we get very excited about what we’re going to accomplish as data scientists, and we can get a little ahead of ourselves.”
  • Understand team members’ motivations. Jensen recommended taking time to discover each employee’s goals, interests and personal incentives. Then you can pair them with rewarding projects and recognize accomplishments in a meaningful way.
  • Offer support. “Data science can be a discipline of failures: Models fail, processes fail, data sources turn out to be terrible,” Phelps said. He suggested offering positive reinforcement and reminding team members that it can take years to see an impact. Jensen also suggested breaking problems into manageable chunks so employees aren’t intimidated by an overwhelming project.
  • Create learning opportunities. Data scientists often leave their jobs because they’re bored, observed D’Agostino. If core projects aren’t cutting-edge, he suggested creating opportunities for team members to learn new things, such as a weekly lunch to discuss the latest research or occasional hackathons to test a new software framework or computational technique.

Nurturing successful teams

  • Build bridges to other stakeholders. Avoid friction and crossed wires by opening communication channels with other teams. Phelps suggested putting a data scientist and product manager in a room for an hour before each new project to ensure they’re on the same page. Jensen added that making data scientists attend meetings without their laptops can force them to communicate with other stakeholders. Giving data scientists opportunities to explain their work to engineers, product managers, and others can also build communication, said D’Agostino.
  • Track performance. Use a template to keep track of what you discussed, the objectives you set, and the feedback you gave during one-on-one meetings with your reports, Phelps advised. Relying on your memory won’t work.
  • Aim to take projects to production. Preparing teams to deploy “their own API services and productionalize code…helps you move faster, and you don’t get blocked on engineering resources that might not be available,” D’Agostino said.
  • Start on-call rotation. As teams get bigger, Phelps recommended setting up a weekly rotation of data scientists on call to fix models that break. That encourages better documentation and gives those not on-call time to focus on core projects.

Being a good leader

  • Ask the dumb questions. Seemingly simple questions can open the door to identifying and solving fundamental problems, Phelps said.
  • Always be learning. The panelists recommended reading prolifically to keep up with developments in this quickly evolving field. Consume not only technical material, but also insights about management and organizational psychology, Jensen suggested.
  • Get out of the way, but not forever. If you’re a new manager, Phelps recommended stepping away from coding for three to six months. “Otherwise, you will never truly prioritize being a manager, and you’ll under-serve the team.” After that, feel free to tackle non-critical projects or those nobody else wants to do. “That way I get to do stuff, but I also get to see where the team is hiding the bodies today,” he said.



To learn more about this panel discussion and its participants, check out the session details here.

To learn more about Data Science Pop-ups in a city near you, visit popup.dominodatalab.com.