This Sunday is Unicorn Backpack giveaway day at the Oakland A's game. Given the current mythology about good data scientists a/k/a unicorns, Billy Beane of baseball analytics fame (G.M. of the Athletics) comes to mind.
Unicorn verification process. I'm not minimizing the difficulty of policy research, analytics, data science, and other efforts to find meaningful patterns in data. But communication skills and business savvy dramatically influence people's ability to succeed. As part of an engagement or hiring process, I suggest asking a potential unicorn these questions:
1) What evidence have you worked with that can potentially improve outcomes? Where might it be applicable? 2) How do you translate a complex analysis into plain English for executive decision makers? 3) What visuals are most effective for connecting findings to important business objectives?
Can you talk the talk and walk the walk? While Mr. Beane brilliantly recognized the value of OBP and other underappreciated baseball stats, that's not what made him a unicorn. His ability to explain his findings and advocate for nonobvious, risky, high-stakes management decisions – and to later demonstrate a payoff from those decisions – is what made him a unicorn.
A colleague of mine worked at a successful, publicly traded telecom company. As a PhD economist, he managed a group of 25 economists. And he says the reason he led the team, and did most of the interacting with senior executives, was that he could explain their economic modeling in business terms appropriate for the audience.
Connect to what matters. Accenture’s extensive research of analytics ROI has found that “most organizations measure too many things that don’t matter, and don’t put sufficient focus on those things that do, establishing a large set of metrics, but often lacking a causal mapping of the key drivers of their business.”
It's a common theme: Translate geek to English. SAP’s chief data scientist, David Ginsberg, says a key player on his big-data team is someone “who can translate PhD to English. Those are the hardest people to find”. Kerem Tomak, who manages 35 retail analysts, explained to Information Week that “A common weakness with data analytics candidates is they’re happy with just getting the answer, but don’t communicate it”. "The inability to communicate with business decision-makers is not just a negative, it's a roadblock," says Jeanne Harris, global managing director of IT research at Accenture and author of two books on analytics.
Will Mr. Beane be wearing a unicorn backpack at the game on Sunday? I sure hope so.