Bill Franks, Chief Analytics Officer of Teradata, talks about data artists vs data scientists and 5 core skills he looks for in an interview
Search Business Analytics, Nicole Laskowski, 3 Aug 2012
has a love-hate relationship with the term "data science."
"It's way over-hyped," said Franks, author of the recently released Taming the Big Data Title Wave and chief analytics officer for the data warehouse appliance vendor Teradata. "If you look at what people described as data scientists are doing today, they're doing what I've always done and what I've always looked for from great analytic professionals."
While they may be using a new set of tools on new types of data, such as social, an analytic professional's job description hasn't changed all that much: thought process, analytic goals, deriving value for the business, Franks said, it's all the same thing.
But he also embraces the new title as a label for the analytics professional he tends to seek out: Someone who has an analytical technical background as well as commitment, creativity, intuition, business savvy and presentation skills -- what Franks refers to as "softer skills."
You recently started using the term "data artist." Why?
I don't expect people to literally adopt that term as opposed to data scientist. It's more of me trying to provoke thought about what really makes a good data scientist. An analytic professional could be a data scientist, data modeler or data miner. And the conclusion I've come to is that the technical skills required for the job are important, but they're not what differentiates super successful analytical people from the run-of-the mill or the not-so-successful. Some of the traits are things that are often not specifically associated with hardcore analytics people.
How does an analytics professional pick up these "softer skills?"
It's sort of like athleticism. There are people who are athletic and those who are not. And those who are inherently athletic, you can put them on a basketball court or a soccer field, and they'll probably do well at any of them. ... You could take someone who just isn't athletic, train them all you want and they probably will never be that good. That's the key. A data artist is someone who has the technical skills and acumen required, but they also have intuition, which is hard to teach. You can help people leverage it, and understand the best ways to use it, but I don't know of a way to teach intuition.
For businesses looking to fill a position of a data scientist, how can they assess whether a candidate has these "softer skills?"
Franks: I can run down the five core areas I look at.
I'll listen to how they are describing the work they've done. I try to hear insights into how they deal with problems. Do they attack or gloss over the problems? Are they going the extra mile when they have the opportunity? ...
This is a big filter for me. I like to say maybe 15% to 20% of the people whose resumes pass my experience and skills test will pass creativity. I ask about those "oh gosh" moments in their careers [where] something horrible went wrong or they hit a huge barrier they might not have anticipated. And I listen to how they worked around that.
... Business savvy:
I ask why they made the decisions they did. Particularly for an analytics process, this is important because what I want to hear is not just some technical reasons. I also like to hear practical and business considerations