From Data Point to PowerPoint: Delivering Data People Can Understand

January 23, 2019

Successful data science teams require a mixture of talents, from data wrangling and analysis to storytelling.

We have more access to data than we’ve ever had – more, in fact, than we know what to do with. Reams of it. So then in come the data scientists. But as Scott Berinato so succinctly states in HBR’s Data Science & the Art of Persuasion: “[Organizations] expect data scientists to wrangle data, analyze it in the context of knowing the business and its strategy, make charts, and present them to a lay audience. That’s unreasonable. That’s unicorn stuff.”

Unicorn stuff, indeed. As leaders, we see this all the time – we have so much data at our disposal, yet we see our teams struggle to take that data and convert it to information that helps us make more informed decisions. Or, if we do, it’s hard to actually see or articulate the results, because we haven’t framed it in a way that our teams  can understand. Our data experts can only do so much.

Don’t get me wrong; we’re getting better. But there’s a missing ingredient in taming and learning from this colossal mass of data – combining people who possess the right talents across the data science spectrum, from the first data point to the final PowerPoint.

datapoint to powerpoint

The right combination of talents 

We’ve all done it – complete your part of a project and then “throw it over the fence” for the next person/team to pick up. When the “it” is data, that’s a real problem. Your data scientists need specialized support from people who can collaborate with them to build narratives and visualizations catered to the audience(s) you need to inform. Do you have those kinds of people on your team? Would you even know it if you did?

And that brings us to a subject near and dear to my heart – talents. By taking a talent-based approach to the full lifecycle of your data operation, you can find and elevate team members to fill that often tenuous gap between data and decision-making.

Berinato suggests taking these 4 steps to build a better data science operation:

  1. Define talents, not team members. What talents do you need to be successful?
  2. Hire to create a portfolio of necessary talents. Talents aren’t roles, so focus more on making sure the necessary talents are available on the team.
  3. Expose team members to talents they don’t have. Don’t let your team turn into a middle school dance, where people separate themselves into cliques and maintain eye contact only with those most like themselves. Make sure your meetings include a mix of talents.
  4. Structure projects around talents.  This will require strong project management skills. (Hint: that’s a talent.)

Crossing the finish line

“The last mile,” as Berinato describes it, is the communication of insights in all of the information you’ve amassed.  Getting there requires a data science team that blends the necessary talents to get you across the finish line.

Comments


Join the Front and Centered Community

It's where life and leadership intersect. Get once-weekly updates.

Subscribe:

Email Address *

I understand SEI may send future emails to me, even if I opted-out before, and that I can opt-out again later.

Front and Centered team