Data Science: it’s a Team Sport
BY CHUCK DENSINGER – COO & CO-FOUNDER
As consultants, we often have “bullpen” cubicle areas assigned to us at client sites. Recently, I happened to arrive at one site before anyone else from our team was there. I found it looking like this:
Initially I thought, “Wow, we left our space a mess,” but I quickly reassessed. There’s a pattern to that pileup of chairs on the left…it’s a residual of collaboration.
A team of typical Data Scientists might have left their chairs neatly at their desks, because that’s how typical Data Scientists work. It’s solo work: data munging, coding, analyzing, creating visuals, then printing or emailing final outputs to the recipient.
We’ve found, however, that solo science produces weak results. Instead, we encourage—nay, require—our Data Scientists to work in twos and threes. Further, they collaborate with marketing strategists, technologists and consulting generalists on our team as well. And, of course, with members of the client team.
Collaboration produces better results because much of data science is exploratory. Collaborators can spark ideas in one another, challenge limiting assumptions, and bring unique experience, perspective and talent to the problem. Sure, we still have divisions of labor, and people have to do work on their own some of the time. But the days of the lone Data Scientist are long gone and forgotten.
So keep bunching up those chairs, team! If anyone comments on the mess, I’ll welcome the conversation.
Data Science: it’s a Team Sport
BY CHUCK DENSINGER – COO & CO-FOUNDER
As consultants, we often have “bullpen” cubicle areas assigned to us at client sites. Recently, I happened to arrive at one site before anyone else from our team was there. I found it looking like this:
Initially I thought, “Wow, we left our space a mess,” but I quickly reassessed. There’s a pattern to that pileup of chairs on the left…it’s a residual of collaboration.
A team of typical Data Scientists might have left their chairs neatly at their desks, because that’s how typical Data Scientists work. It’s solo work: data munging, coding, analyzing, creating visuals, then printing or emailing final outputs to the recipient.
We’ve found, however, that solo science produces weak results. Instead, we encourage—nay, require—our Data Scientists to work in twos and threes. Further, they collaborate with marketing strategists, technologists and consulting generalists on our team as well. And, of course, with members of the client team.
Collaboration produces better results because much of data science is exploratory. Collaborators can spark ideas in one another, challenge limiting assumptions, and bring unique experience, perspective and talent to the problem. Sure, we still have divisions of labor, and people have to do work on their own some of the time. But the days of the lone Data Scientist are long gone and forgotten.
So keep bunching up those chairs, team! If anyone comments on the mess, I’ll welcome the conversation.
Tags: