All conversations

Women in Data · Key takeaways

Leading through empathy, breadth and trust

Elisabeth Kant and Morena Bastiaansen examine what changes when a data scientist moves into leadership, why empathy is foundational and how leaders can create impact without knowing every technical detail.

Data leaderElisabeth KantHead of Data Science
Building towards leadershipMorena BastiaansenData Scientist

What stood out

Ideas worth carrying forward.

01

Leadership trades depth for a different kind of complexity

An individual contributor can go deeply into a defined technical problem. A leader works across people, products, business goals, imperfect data and resource constraints, where there may be several defensible answers rather than one clear solution.

02

Empathy is the foundation of data leadership

Understanding what motivates people and what an audience needs makes better communication, mentoring and collaboration possible. Skills can be developed, but genuine care for people and curiosity about problems are essential starting points.

03

Strategy still needs technical grounding

Early-stage leaders need enough hands-on experience to guide developing practitioners. As their scope grows, they can step away from every algorithmic detail while still understanding which approaches fit which problems, including governance and ethical constraints.

04

Underrepresentation has no single cause

Bias, limited role models and sponsors, unequal caring responsibilities and unreliable childcare interact with confidence and appetite for risk. Individual action matters, but it should not be mistaken for a complete answer to structural barriers.

05

Challenge gender expectations in both directions

A fairer workplace gives women genuine access to influence without assuming every woman wants leadership. It also gives men freedom to prioritise family or reduce working hours without being judged against a narrow model of success.

06

Generative AI changes the toolbox, not the purpose

Data science still exists to create value from data. Leaders and practitioners need to understand where language models help, then integrate, evaluate, monitor and safeguard them rather than treating the model itself as the business outcome.

07

The leader owns the why and what; experts shape the how

As the field and team expand, the leader cannot remain the deepest expert everywhere. Giving specialists context, boundaries, ownership and autonomy usually produces better technical decisions and more meaningful work.

08

Buffer pressure without pretending it does not exist

Economic constraints, hiring freezes and difficult decisions cannot always be kept away from a team. Transparent communication, thoughtful buffering and support such as coaching help a leader protect psychological safety while facing reality.

Take part

Bring another perspective to the conversation.

If you lead a data team or are building towards leadership, I'd love to hear from you.

Start a conversation