Three Essential Qualities of Successful Data Leaders

Introduction

As organizations seek to harness the potential of data, it’s important to remember that the key to unlocking this potential lies in finding and growing great data leaders.

Great data leaders are the ones that make it happen. They are technical enough to supervise the data lifecycle from extraction to visualization. They have enough business acumen to strategically impact the organization. They find the right structure and positioning for the data team. They evangelize data across the company. They speak everybody’s language and act as a glue between different teams. In fact, that’s a lot on their plate.

So, what does it take to be a good data leader?

This question is important. As a data leader yourself, you might be looking to understand the skills you need to develop for the job or the skills you can expect to acquire during this experience. As an organization, you might be looking to hire or grow a data leader. You might also just be curious about good data leadership.

No matter your motivation for exploring this subject, I interviewed 5 data leaders I admire a lot and tried to identify the qualities that make up their success:

Steven Longstreet built the data analytics practice at Hilton, drawing on his experience as a sales and operations person.

Serge Descombes led the first Analytics Center of Excellence at Criteo and later became the Director of Global Operations at the company.

Matt Brems, now the Principal Data Scientist at DataRobot, built upon his experience as a Data Science Instructor.

Alexandre Lenoir, created and supervised the data department at Spendesk before transitioning to the position of a staff researcher within the organization.

Arnaud de Turckheim built Payfit’s data team from scratch and then co-founded the data catalog CastorDoc.

These leaders from vastly different backgrounds agree on the soft skills you need to be successful as a data leader: prioritization, business acumen, and pedagogy.

In this piece, we dive into what these skills are and why they are needed. We also share some tips on how to develop them, drawing on the experiences of our five interviewees.

‍

I - Prioritization

Successful data leaders are prioritization gurus. In fact, as a Head of Data, you have a ton of projects that arise and require attention. There usually are three components to what your team does:

Operational runs: This is the support role of your team. This includes answering tickets for other teams and helping them with their immediate needs.

Operational build: Build is what helps you make the run faster. It’s about automating what can be automated, empowering other teams to answer their tickets themselves, etc.

Strategy: Strategy is about taking a proactive approach towards projects and providing clear directives regarding key decisions.

These three pillars are vital. If your team is only running, then it is doing a poor job. If it is solely focused on strategy, it will create frustrations with the rest of the company. But there’s only so much you can do in a day, so you need to prioritize ruthlessly to ensure you are running things the way you want. This involves being able to say “no” or “later” most of the time.

But prioritization extends beyond actions. It also encompasses decisions about personal development and the knowledge you decide to acquire.

As a data leader, you often need to find the right balance between business acumen and technical skills. Of course, you cannot be a business guru and a tech expert. So what do you do?

Steven Longstreet recommends acknowledging any gaps in business or technical skills and identifying the essential knowledge required for success. Most of the time, it boils down to understanding how to look things up and apply them effectively.

Another important part is learning to rely on experts around you. Serge Descombes explains that you should seek to become an expert yourself, for two reasons:

When you are a data leader, the experts are the people on your team. Your role is to create this team of experts and help your team grow in their expertise. You should not be seeking to replace them. Instead of trying to be an expert, you should focus on understanding experts and translating their language to other people in the company.

Secondly, when you strive to become an expert, it becomes harder to see the business aspects of the company. You are locking yourself in a technical world. It will surely make it harder to transition to the business side afterward. If this is something you are considering, then be careful about cultivating expertise.

If prioritization is your strength, then it will prove invaluable in your role as a data leader. If not, it’s a skill you can expect to develop throughout your experience.

‍

II - Political Acumen

"As a data leader, I don't command but collaborate. Navigating priorities of others, I focus on fostering relationships with other teams, nudging them to self-realize the importance of their role in our shared goals." - Matt Brems, Principal Data Scientist, Technical Excellence and Product, DataRobot

The data world is more political than one might think. There are two scenarios in which you might expect to navigate the power dynamics of your organization:

1. You are building the first data practice in your organization
2. You are not satisfied with the positioning of the data practice

Most data leaders are in one of these two scenarios. Scenario one is common, as a lot of organizations are at the start of their data journey. Scenario two is also popular, as it’s not uncommon for data teams to be left on the sidelines.

Positioning your team in a landscape that can get political goes through articulating the value of your team’s work, and building the right relationships. Let’s look at these two elements.

A. Framing your team’s role

"When you come in to build the data practice, you face constraints. It’s hard to pitch new methods to people who are used to working independently. It might be perceived as infringing their territory” - Arnaud de Turckheim, CPO, CastorDoc.

When the company culture is not inherently data-driven, the data team can be seen as a threat. In fact, you are disrupting a system that worked before you got in the picture, essentially telling people that they should incorporate a new dimension to their work. This is not always well received. What’s more, your team becomes a bottleneck to all the information in the company. Information is power. If you become a bottleneck to information, you might be perceived as a threat as well.

The key to success lies in framing your team’s role in the right way, early in the process. Arnaud de Turckheim recommends putting in place the following strategies when framing the position of your data team.

1. Position your team as enablers, not disruptors: Show that your team’s goal is not to disrupt but to empower others. Your expertise can provide them with tools and insights to improve their performance while maintaining their autonomy.
2. Maintain Transparent Communication: Regularly share what your team does and how it benefits the organization. Transparency minimizes fear and uncertainty.
3. Build a Collaborative Approach: Engage other teams in your processes. Show respect for their domain expertise, and request their input. This promotes a sense of ownership which usually reduces the feeling of threat.

You can go a lot of different ways about this. The important thing to note is that the data team won’t naturally have a place until it is crafted by the leader. And crafting the right positioning for your team will require some political acumen.

‍

B. Building the right relationships

“As a data leader, you are co-dependent on all other teams” Alexandre Lenoir, former Head of Data, Spendesk.

The adage “your network is your net worth” surely applies to data teams.

As a data leader, building the right relationships is essential. You are finding yourself in a position of co-dependence with other team leaders, and you need their help to make things happen. As Matt Brems noted, “As a data leader, you must navigate your way up the priority list of teams who have other concerns. It involves educating managers on the importance and downstream effects of the data team’s work.”

But that’s not it. If you want to disseminate data in all the departments of the company, you also need to find data champions. This is a big part of data leadership.

Steven Longstreet advises looking out for people in the company who are curious and willing to learn more & play with data. Once you connect these curious folks with the right data context, then your initiative becomes very powerful.

As a data leader, the first step is to identify the power dynamics that might exist in the organization. Navigating these power dynamics usually come through positioning your data team in the right way, with the right messaging, and creating strong relationships with key players in the company.

Pedagogy

Does your job sometimes feel like you’re a master coordinator? Well, it’s because you are.

That’s why pedagogy is a vital component of data leadership. As a data leader, pedagogy means being able to simplify complex topics and communicate in a language that everyone in the organization can understand.

Pedagogy serves to educate your team, bridge the communication gap between different teams and empower other teams with data. We’ll look at each facet, in turn.

First, an efficient team is a team that operates with a shared language, tools, and methods. To this end, teaching will always be an important part of your role. Serge Descombes, former leader of the Analytics Center of Excellence at Criteo, explained that a significant part of his work involved enabling the company’s 300+ analysts to work following consistent methods and tooling. Ensuring everyone works within the same framework helps you maintain uniform quality across teams and divisions.

However, your pedagogical responsibility is not limited to your team. As a data leader, you act as a glue between teams that do not speak the same language. You need to act as a translator, making sure that teams speaking different technical languages can still be on the same page. What does that look like in a real-world setting? It can take a lot of different shapes. In the case of Serge Descombes, it involved understanding the requirements of data analysts and then giving the right recommendations to data engineers responsible for the ETL process. For this to happen, Serge needed to understand both the language and analysts and engineers and be able to translate between the two groups.

Beyond this, you are also responsible for ensuring that data-driven thinking powers all departments. In fact, if business departments are not using the data to make decisions, the data might as well not exist. To ensure data is democratized, you must make data accessible, understandable, and actionable. This involves breaking mental barriers and evangelizing data through initiatives such as training sessions, tool presentations, and other means of evangelization.

The significance of pedagogy in data leadership cannot be understated. In fact, it is not uncommon to witness professionals, such as Matt Brems, transitioning from teaching roles to data leadership positions.

Conclusion

Data leadership is a complex, nuanced, and dynamic discipline. In this piece, we’ve identified three soft skills that are vital to good data leadership: prioritization, political acumen, and an aptitude for pedagogy.

By developing these skills, you not only enhance your abilities as a leader in your current position but also pave the way for the future. As a data leader, the future might not always look straightforward, as the role opens a lot of doors. My next article on the topic will focus on what comes after data leadership. If it’s a subject you’re interested in, make sure to receive it straight in your inbox by subscribing below.

Subscribe to the Newsletter

About us

We write about all the processes involved when leveraging data assets: from the modern data stack to data teams composition, to data governance. Our blog covers the technical and the less technical aspects of creating tangible value from data.

At Castor, we are building a data documentation tool for the Notion, Figma, Slack generation.

Or data-wise for the Fivetran, Looker, Snowflake, DBT aficionados. We designed our catalog software to be easy to use, delightful and friendly.

Want to check it out? Reach out to us and we will show you a demo.