Do I want to be a (data science) manager?
If you are in a technical field, deciding whether you want to enter management or continue to develop as a technical specialist is one of the most daunting decisions you might face. Before even contemplating your options, I would encourage you to do two things: understand the career tracks at your current company and independently evaluate what you want.
Understanding Career Tracks
You don’t usually find the different career paths plastered on billboards for all to see. You usually need to start a conversation with your manager to understand what your options are. That said, most companies have something similar to the track pictured below.
Your career will start at a junior level, focusing entirely on technical work (yellow). As you get more experience and move up the ranks, you will naturally progress up the Technical Specialist track (blue). With this advancement comes choice: do you continue to develop your technical skills as a technical specialist or pivot to a focus on softer skills and people management? Before getting into the decision making process, let me address a few things about this chart that you may have already noticed.
The arrows unidirectionally point towards management. This isn’t to say that you can’t go back to being an technical specialist, but it’s a lot easier to move from specialist to manager than vice versa. Why? Because as a senior technical specialist you will continue to get pseudo managerial experience by being involved in project planning, scoping decisions, or mentoring others (responsibilities that a manager often takes on). The longer you are a manager though, the further removed you become from the technical granularities of the work; often dulling your skills. Thus, it’s often easier for technical specialists to jump over to management then to switch back, although it’s not impossible.
With that in mind, there’s no harm in switching later on. You don’t need to make the “right” selection as you soon as the choice presents itself. Switching to the managerial track early can help accelerate your growth in one particular subject area (usually the content of the team you’re managing). Whereas staying as a technical specialist longer allows you to get a wider amount of experience and thus widens the numbers teams you can potentially manage down the road.
Secondly you’ll notice a giant box dedicated to apprentice management. Whatever company you are at, ask what an introduction to management looks like and what resources would be provided. In my experience, becoming a manager was far more nuanced and complex than I’d anticipated, so you should be proactive in asking about the type of support and training you will get. It is very easy to underestimate the adjustment it takes to be oriented around the work and success of others instead of just yourself
Parts of a Data Science Project
So now you have an understanding of your options, but you still might not know what’s right for you. Once you are in a subject domain that you like (if you are not, have a conversation with your manager using the hierarchy of data needs as a framework), start to think about how you contribute to a given project. At a high-level, most data science initiatives can be formatted into the following setup:
(1) Planning/Scoping — Before the project starts, you will need to draft a design doc, a spec, and even a roadmap or timeline for delivery. Often this involves establishing stakeholders and determining what is within/out of scope.
(2) Execution — Whether you’re building from scratch, reusing an existing repo, or simply calling an internal (or external) service, this is where you get your hands dirty. The product could be a dashboard, a machine-learning model, or an experiment but the core portion of work will be writing code and understanding the nuances behind the numbers.
(3) Implementation — Great, you built a model or did some ad-hoc analysis, how can it be reused? Some data science projects aren’t just one-and-done, so they’ll need extra engineering work to ensure they run autonomously. Most often these take the form of models, dashboards, or ETLs; wherein you will need to automate a portion of your scripts to provide updated data on recurring basis. It may be as basic as creating a cron job, but the process should still involve getting your code reviewed and approved by other engineers.
(4) Storytelling — Most projects involve conveying to others what you learned and how that should change your strategy or product. This storytelling can take many forms: verbal presentations to your team, 1:1 retrospective with your peers or manager, or even just a whitepaper. This skill is often one neglected early on in your career, as it can be very easy to hand this work over to your manager to go advocate on your behalf.
Figuring out what you want
It’s at this point that I would ask three questions and caveat to think about them independently; that is, the answers do not need to align. Of the four areas of Planning, Execution, Implementation, and Storytelling detailed earlier,
A. Which do you enjoy working on the most? (interest)
B. Where are you strongest/weakest? (skills)
C. Which ones are most important to success in your current role? (role)
Write it down. Take note if there are major differences across the three answers. If there are, that’s fine, it means we’ve got some work to do. Let’s start by reviewing your answer for what you enjoy.
If you highlighted boxes (2) and (3), then you’ve got a bent for the technical side of the work. That’s great. As you become more senior, staying on a technical specialist track will continue to develop these skills while getting some exposure to the others. as you become a tech-lead.
If you highlighted boxes (1) and (4), you skew towards the softer skills needed for a project and ones that closely align with what a manager does on a daily basis. The more senior a manager is, the less and less exposure there is to the technical categories since they’re typically doing planning and storytelling for multiple projects under their purview. This is why managers often find it hard to go back to being a technical IC.
Aside: I’m drastically oversimplifying the responsibilities of the manager by omitting the part of the role that is about people management and development. If you are interested in learning more about that experience please stay tuned, I’ll follow-up with a post about that topic in the near future.
Now, let’s talk about your skills (B). It’s critical to be honest with yourself about what sets you apart and where you need to develop. Having a manager and peers who are open and honest is key to ensuring you don’t have a biased perspective. The biggest thing to take note of is where there’s a gap between what you enjoy working on and what your strengths are. The best thing to do to close that gap is find people who have that strength and talk with them about how they honed their skills. Whether they took a course on the side or developed good habits that compounded over time, you should work to create a development plan that helps your skills better match your interests.
Now onto (C), your current role. Below you can find a self-made chart that roughly illustrates how different roles spend their time.
How did your answers about your interests and strengths align with your current role? If your interests match your skills, but not your role then, start talking to your manager and others about roles that better suit you. Be patient though. You don’t find your ideal role overnight and the process of growing the skills to be successful in it takes time too. The point is to slowly steer your role to more closely match both your skills and interests, bit by bit.
If your self-assessment didn’t fall into one of the nice little scenarios I described above, that’s fine too. Hopefully you at least gained a sense of what you want and how that aligns with different career paths. Don’t feel like you need to rush any decisions. Careers are many decades long, so a couple extra months — or even years — experimenting and figuring out what suits you best is perfectly fine.
Summary
At the end of the day, there are no wrong choices; the key is just taking time to be introspective and to learn from experience. Technical specialists will get the opportunity to really hone their craft, developing larger parts of more complex data products over time. Managers will spend their time balancing developing their team and planning projects efficiently to reduce waste and storytelling to create alignment across organizations. Having talented individuals in both types of roles is critical to a data science organization’s success.
