What makes data science different?

Reflections on the 2017 MIT CDO/IQ Symposium

This was one of my favorite moments at the MIT Chief Data Officer and Information Quality Symposium this summer. The organizer, Dr. Richard Wang, called all of these people up to the front and said pointedly to the room: “Now, these are vendors. But they’re also my friends. So be nice to them.”

There was laughter. But you know what? Everyone was. Nice. Really nice. Vendors, CDOs, CIOs, and other panel participants readily shared their own experiences, and just as readily admitted their weaknesses. It was refreshing to hear people say, “You know what, I’ve never really done that. You should talk to that person over there—they have a lot of expertise in that.” Even when “that person over there” was a vendor from a different company.

The same collaborative atmosphere was present at the Open Data Science Conference East earlier in the summer. Vendors, old hands, newbies, and fellow travelers all together in a friendly, collaborative environment sharing, critiquing, and suggesting.

Perhaps I should not be so impressed by this level of cooperation. But to someone who has seen their share of professional conferences and vendor expos—not to mention vendor selection committees—the atmosphere was striking. Turf battles? Defensive hedges about capabilities? Didn’t see it. Instead, there was a pervading sense of everyone being on the same side. One person’s success didn’t diminish another’s.

I’m making a big deal about this because folks coming up in data science now may take it for granted. The collaborative values of open source and open data seem natural to those who pursue it academically. But to many corporations, they are still alien.

There are two important lessons that data professionals, executives, and thought leaders should take from this.

First, be sensitive to the impact of your company’s culture on your internal clients. We know that data takes on new power when it is shared, aggregated, and shared some more. But it’s hard for a Chief Data Officer to promote a culture of openness, collaboration, and collective business benefits if other leaders are pitting directors or managers against each other for project funds or operating budget. A great CDO is defined by the ability to persuade others that data serves their best interests as well as the company’s. A prerequisite for this is listening and understanding what those interests really are.

Second, leverage the values of collaboration in meeting your own business needs. Is it really necessary to pit vendors against each other? Can we bring some flexibility to the way we analyze our own business needs, so that we can invite collaboration and get strengthen complementary benefits? (I’ve written more about this in a previous post, “Core, Collaborate, Delegate.”)

And while we’re at it, let’s all be nice to each other. How about it?