A lot of my articles, as well as much of the writing on data science in general, focus on the work of individual data scientists. In this article, though, I want to focus on something different: the data science team. But first, let's define what such a team usually consists of. Although this configuration isnt set in stone, here is an example of a data science team: a few data scientists, a data engineer, a business/data analyst and a data science manager. 

The specific composition of the team is less important than how the team works together, however. With that being said, let’s look at the tools and methods you can use to improve collaboration among your data science team, whether you are a data scientist, a manager or possibly a technical recruiter.

3 Ways to Become a Better Data Science Team

  1. Planning and grooming.
  2. Stakeholder updates.
  3. Retrospectives.

More From Matt PrzybylaHow to Use SQL in Python

 

Planning and Grooming

This first tool is a combination of planning and grooming. These terms can be a little muddled, though, so let’s define them first.

Grooming falls under the umbrella of organization, but what sets this process apart from planning (in certain companies) is that it serves as the first review of whatever is in your backlog. This queue may be composed of several Jira tickets or other general tasks that your team has come up with over time but has not yet prioritized into an active process.

You can think of planning as more specific on a sprint level. Even if you don’t use Jira, you can still plan weekly, bi-weekly, or on whatever cadence you prefer, and log it with more check-ins. Typically, in these check-ins, you’ll discuss upcoming projects. More importantly, though, you’ll address the digestible tasks of a particular project for that given week or time period.

Here are a few takeaways and benefits that can come from collaborating on planning and grooming:

  • Assigning the level of effort for a particular task.
  • Assigning importance or priority.
  • Avoiding redundancies.
  • Highlighting for yourself what you’re focusing on for the week.
  • Discovering whether anyone else has worked on something similar and can help you or make the task more efficient.

Once again, these terms might switch meanings or be interchangeable depending on your company’s processes or if you’re working in an agile environment. What’s important, however, is improving your team’s overall organizational ability.

 

Stakeholder Updates

Stakeholder updates are not often discussed in the process of becoming a data scientist since the training is usually more focused on learning algorithms, coding, and the respective, underlying concepts of each of those.

Stakeholders are the people (or the single person) who assign your tasks or projects or who will digest your final project and its impact on the business. That being said, stakeholders do not all have the same role; they may be data science managers, product managers, sales engineers or in some other position, depending on the company.

You can always update stakeholders through Jira tickets, Slack messages, Google Slide decks and many other methods. The point is not the platform you use; it’s the way in which you share your information and updates.

Here are some ways that you and your team can effectively update stakeholders:

  • Data science topics can be confusing if you aren’t a data scientist, so make sure you remove unnecessary jargon that only a data scientist would know.
  • Explain a project in the why/why/how/result method. Here is an example of questions that you could raise: What is the project? Why is this project happening at all? How is it being solved? What are its results?
  • Use before and after comparisons, like the following: What was the target before, and what is it now?

Also look at breakdowns of specific groups of data: You may organize it geographically, by type and so on.

There are many ways to explain and update your data science projects, but the most important thing is how you articulate them.

 

Retrospectives

Finally, retrospectives are crucial to your data science team. This meeting is usually a thorough discussion of a few situations that your team has faced and can be held bi-weekly or monthly.

Planning and grooming take place before the project or task, and stakeholder updates occur during and often at the end of the task. The retrospective, however, encompasses everything that happened in the project’s entire timeframe.

You will typically look at a few things in this retrospective discussion:

  • What went well?
  • What could be improved?
  • What are new opportunities moving forward?
  • Specific callouts for those who did well.
  • What are the action items that we should follow up on between now and the next retrospective?

All these questions and areas will generally cover everything important that has happened over the given timeframe. You will gain a better sense of what’s important to the team, the company and yourself.

Start Your New Career in Data ScienceCareers in Data Science: How to Get Started

 

Improve Your Collaboration

While improving your own work will improve the team, you can focus on other, more team-centric items to make your data science team even better overall. To summarize, here are three ways that your data science team as a whole can improve:

  • * Planning and Grooming
  • * Stakeholder Updates
  • * Retrospectives
Expert Contributors

Built In’s expert contributor network publishes thoughtful, solutions-oriented stories written by innovative tech professionals. It is the tech industry’s definitive destination for sharing compelling, first-person accounts of problem-solving on the road to innovation.

Learn More

Great Companies Need Great People. That's Where We Come In.

Recruit With Us