Let’s not sugar-coat this: Being a data science manager can be difficult. It isn’t like being a sales manager or a business development manager or any other manager of smart professionals. The objective of the data science department, for one, isn’t as clear cut–it’s not just about improving sales or identifying business opportunities or hiring a new employee. It’s about all that and more.
Data scientists, engineers, and analysts all exist on the same plane, but possess different responsibilities in their individual roles. They are in charge of a wide range of duties: creating algorithms to collect and wrangle data, interpreting and identifying important data points, presenting data to stakeholders through visualizations.
So with that in mind, what should data science managers focus on as they lead a team?
Get to Know Your Team
Given their diverse job descriptions, a wide spectrum of individuals from all sorts of backgrounds often form a data science team.
Data engineers possessing acute programming skills make use of their knowledge of machine learning and other technologies to pick out important data. Data scientists further wrangle this messy data into clean and valid lines of data. Data analysts with an eye for distilling data must then turn this pure data into tangible meaning for stakeholders.
However, the roles of these individuals can overlap in certain senses and it therefore becomes very important for a data science manager to truly understand what each member of their team does best in order to delegate the right job to the right person.
Before taking on a new team, or new team member, it is of the utmost importance to go over their strengths and weaknesses. Keep track of the performances of each member throughout their career in your team. As your understanding of your subordinates grows, you will become more and more able to see where each member shines the most–and that’s where your team will gain a great boost in their work.
Let’s talk about what a data science manager really does. Like any other manager, they have to manage their subordinates–and as previously mentioned, give the right job to the right person.
But a data science manager also is in charge of guiding the team’s vision by identifying new opportunities to leverage the powers afforded by data science. They must also help to execute that vision, by working closely with not just the team itself but with others teams, such as the engineering department or the product management department.
In other words, data science managers have to be malleable and fluid, able to fill in the gaps between their subordinates and bridge the space between their team and other groups within the company that may be more customer-facing.
Plan for the Future
Data science managers have big shoes to fill (and many pairs), from being leaders to presenters to technical support. But the biggest role they have to take on is that of a visionary.
By itself, data is meaningless. It’s also messy, and it can be difficult to pick out what matters from the sea of what doesn’t. A data science team working in the dark with no clear goal in mind will have a hard time figuring out the potential that a pure data set can hold for a particular client.
It is therefore the manager’s job to envision the path that their team must follow. They must leverage their experience in the field and their expertise to find structure and meaning where there isn’t any, and set their team toward the right goal.
Before taking on a new project, it is the responsibility of a data science manager to take the time to truly understand the parameters of the project and to have a clear idea of what kind of methods their team can undertake. In other words, do your due diligence, in order to help bring your team forward to the next level.
See the Big Picture
In many other departments, managers often gather their team on a weekly basis to catch up. They set aside time to discuss where they are on a project’s deadline and discuss what each individual team member has accomplished in terms of their own duties.
However, in data science, it’s not as important. In fact, it is the sum of parts, the whole, that matters more than each individual part. Rather than contributing directly to the final product, in the data science department, every individual contributes to a chain of puzzle pieces that connect to form the big picture.
A data science manager’s job is not to micromanage and control the way that each of their subordinates achieves their goal. It isn’t their job to be an excellent analyst or engineer or presenter and carry the team to their goal. Instead, data science managers must see the big picture and keep their team constantly on track, something that is difficult given the lack of distinct markers of progress.
Learn, Learn, Learn
With the wide range of roles that data science managers must play, it’s no wonder they have to stay on their toes. After all, technology and business constantly change according to the predominant preferences of the market. This means managers in this field have to keep themselves at the top of their game in terms of their knowledge of best practices in the field.
One important expertise for data science managers is their ability to communicate effectively. This can be difficult, especially when it comes to presenting the findings of their team to relevant stakeholders in a way that non-data science professionals can understand.
Above all, data science managers cannot rest on their laurels. They must be constantly learning and improving themselves so as to better lead their employees and contribute to the team’s progress.
As a whole, data science managers have to fill big shoes. They are the frontline technical support for their team, but they’re also the leaders and visionaries that guide their team forward. In a field as complex as data science, good managers are hard to come by, as the work can be immensely difficult. However, it can also be among of the most rewarding managerial work as it affords a unique level of involvement and engagement with employees, clients, and other departments.