Leading Data Science Teams

Chapter 2. Managing the Work

Data science teams are a pretty busy bunch. There are constant requests for new reports, new models, and past analyses to be redone with new data. For most data science leaders, this means having to make thoughtful decisions around what to actually work on and what to deprioritize or not do entirely. It also makes it very important to manage how long tasks are taking because if one particular task takes longer than expected, something else has to be pushed back.

To be able to successfully manage the workload of your team, you’ll first need to have a very clear view on what the goal of your team even is. Are you helping make the product better with machine learning, providing strategy advice with data to a particular department, or something else? You’ll then need a clear project management process for how to keep track of the work, and while data science is similar to software engineering, some software engineering principles, such as Agile, don’t always directly map to data science. You’ll also need to have clear communication with your stakeholders so that, as new developments arise, everyone has the information they need.

The task of managing the work of a data science team must be a concern for data science leaders on teams of all sizes. On smaller teams, the leader may do the project management tasks directly. On larger teams, there will be a project manager who is in charge of managing the tasks. But as the data science leader, you’ll still be ...