Are you sure you want to become a Data Science Manager?

Don't rush into the fancy title until you have read this.

Photo by Benjamin Elliott on Unsplash

Picture this. You have just delivered a killer project, your team’s buzzing, and then—bam! You are asked, 'Have you ever thought about leading the team?' Sounds tempting, right? But hold on—do you really know what you’re signing up for?

As a Data Science manager, I have watched my teams grow from 0 to 12 data scientists and helped scale our DS discipline from 20 to 50+. I have also seen fellow managers leave for new challenges. Both instances created a vacuum to fill: the need of a manager to lead a data science squad. Filling this void can be an amazing opportunity, but I have also seen many colleagues fail to adjust to the transition.

The transition to management is no joke. Sure, it has its perks. But no one talks about the trade-offs. Most early managers walk in totally unprepared - that’s where the frustration starts.

What will you be able to read in this blog?

I will cover what are the main considerations for you to think about before making a move into management.

You will read about:

1. Running a high level self-reflection. Why are you thinking about this move?

2. Debunking the glorified “manager” title. In many instances, lead roles are a lateral move, not a vertical one.

3. 5 main differences when you make the move. I am sure there are more, but these 5 are the ones that I relate to the most.

4. Describing the danger zone. Moving back from manager to ICs is not as easy. Also, avoid the things that drain your energy.

5. But also, acknowledging the doors that a DS manager role open. A managerial role makes you grow in areas that few have a chance to experience.

First, tell me why.

The mentioned gap to lead a data science squad, opened an opportunity for those who wanted to transition from their current individual contributor role to a lead role. However, this transition is not for everyone. I have seen those who have rushed into it, regretting the move some time later. I always ask this question to potential people leads: why do you want to become a manager?

Some answers or thoughts I have heard are:

• Are you seeking recognition? Moving into management comes with increased visibility - you’ll lead decisions, interact with stakeholders, and guide strategy. But is this kind of recognition what excites you? And if so, what would you be willing to trade in exchange for recognition?

• Are you stuck at the Senior Data Scientist level? You have spent many years in a Senior DS role and becoming a Principal Data Scientist is really complicated. The next step may seem like management. But before you jump in, consider: Is this where your passion lies, or are you just looking for a way out of stagnation?

• Is it about the money? In the tech space, managerial roles actually have the same salary ranges as their equivalent IC counterparts. There are other industries where managers do get paid more though. Assuming that there is financial benefit with the new role, it is enough to trade in the technical work you love?

• Is it a requirement in your company? In many organisations, career growth seems tied to management. Maybe you feel pressure from above, as if the only way to advance is to take on a leadership role. But is management truly the only path to growth, or could you explore other ways to progress in your career (even if it is moving to another company)?

Do you relate to any of the above? If so, see that there is a common denominator in my challenge questions:

What is the trade-off?

Debunking the glorified “manager” title.

Before covering the hidden differences between a manager and an IC, let me touch upon the actual “manager” word.

Photo by Museums Victoria on Unsplash

It comes from the industrial revolution.

The term "manager" has roots in the industrial era, when work was defined by efficiency, control, and supervision. In factories, managers oversaw production lines, ensured workers followed processes, and enforced order. Every time I hear the term “manager” referring to a people lead, I cringe.

But the term is recognisable.

Part of the reason we still use "manager" is simply because it’s a catch-all term. It’s familiar. Everyone knows what a manager is, even if the responsibilities of modern managers go far beyond just managing tasks. In industries like data science, where innovation and collaboration are key, managers are more about facilitating success than enforcing authority. However, the term persists because it’s deeply embedded in corporate structures and hierarchies.

I prefer the term “guide”.

First of all, I have always disliked the term “manager” applied to people teams. My subconscious brain associates the word “manager” with authority and power, but in reality, you don’t have that much decision making power. I believe that a manager’s job is to guide and support your team, removing roadblocks, and managing expectations from stakeholders. The "big" decisions often come from higher up, so in many cases, you become more of a facilitator than a direct decision-maker.

Management is more a lateral than a vertical move.

There are people who view moving into a manager role is as a promotion. In some companies that might actually true, but in the tech space, becoming a manager is a lateral move. I have never seen a junior data scientist become a manager. It has always been those who reach to senior data scientists and have acquired enough technical knowledge who make the transition. In fact, it is a lateral move even on pay grade, as many tech companies have the same salary ranges for the same level of IC or manager roles. To make an analogy; management isn’t an elevator ride to the top, it’s more like taking a detour through a maze—with half the map missing and someone constantly yelling about deadlines.

The 5 things no one tells you about when you become a Data Science manager

As an IC considering the move to management, you’ve likely interacted with your own manager daily and observed what they do. Maybe you've had a manager and a squad lead, so you think you know the drill - I say 'maybe' because perhaps you started as the only data scientist in a startup and grew from there, though that’s rare. If I asked you to list what a manager does, your list might include:

• Do 1-1s

• Some admin stuff

• Plan the weekly and quarterly work

• Interact with stakeholders on behalf of the team

These are true enough, but they only scratch the surface. These are tasks, not responsibilities.

And responsibilities are the things that define what a manager’s role really looks like.

Let’s burst this naive knowledge bubble with five things no one warns you about.

1) Your project focus shifts from technical to strategic work

You still need to be technical, but with a different focus.

The more critical code you write, the more of a blocker you become. The temptation to roll up your sleeves and start coding is strong. Especially when a project hits a roadblock or when you know you can do the work 3 times faster. For example; reviewing pull requests (PRs) or offering feedback on the structure of an analysis in a notebook? That’s perfectly fine. These light-touch interventions allow you to stay engaged technically without bottlenecking the team. However, taking on a full data quality audit or single-handedly building a new model? That’s too much. I can guarantee that what looked like a free coding week, can easily become a crowded Outlook agenda and suddenly there is no more time to write code and you are the blocker. When you do this, you’re no longer leading your team—you’re stepping back into the IC role and taking ownership of tasks that your team should be handling.

On this point, I speak from my own experience and wrote an article talking about how my agenda was a full mess and had to review a way to become more efficient handling bigger teams and more projects.

But, you need to be technical enough to guide your team. A good manager understands the nuances of the work their team is doing. You don't need to dive into the details of every line of code or optimise models yourself, but you must be able to follow the conversation. You need to understand the “art of the possible”—what can be realistically achieved given the time, resources, and technical constraints. This understanding helps you plan the phases of long-term projects, anticipate risks, and communicate effectively with stakeholders.

2) Your impact is less clear

What got you to success as an IC, will not lead to success as a manager.

You are no longer in the driver’s seat. As an IC, your success is directly tied to your output—your code, your models, your analyses. But as a manager, your performance is evaluated by how well your team performs. This can be a tough adjustment for those used to being high performers. A good manager focuses on enabling the team, clearing roadblocks, and fostering a productive environment.

Stopping work is also part of success. As a manager, recognising when to stop on-going initiatives or not taking on too many projects is crucial. Spreading the team too thin can lead to long-term risks and create single points of failure. Remember, it’s not just about what you achieve but also about how you set your team up for sustainable success.

No performance drop in your absence is the ultimate goal. Imagine that you were to take a three-month leave, but your team maintains their output levels. That is a testament to how well you’ve set them up for success. It demonstrates that you’ve empowered your team to function independently. Your role has shifted from being the day-to-day enforcer to a strategic leader who cultivates talent, raises standards and ensures the team thrives, even in your absence.

3) Your growth doesn’t come from books

Learning to lead takes experience, not just theory