I think the folks at Harvard need to go back to school. I understand why they believe that data scientist will be "the sexiest job of the 21st century", but what I don't understand is the lack of applying basic business school fundamentals to the claim and then taking the next step. I realize that I'm a little late to criticize an article that's nearly a year old and I hope that people will cut me a little slack given that it was just brought up again in these pages a few weeks ago (which was the first time I saw it).
Right around the same time, the HBR article was published, I wrote an article that also foretold of the rising power of Big Data professionals, but where HBR saw "ability to code" as the prime differentiator, I saw insight and artistry as the prime radiants.
The 5 Year Century
The HBR article details how "data scientists" are poised to uncover big value for big companies and used Jonathan Goldman of LinkedIn as an example. Goldman, through trial and error combined with the freedom given him by LinkedIn's then CEO, Reid Hoffman, created LinkedIn's highly successful "People You May Know" feature.
The article references how some VC firms are even going so far as to create specialized recruiting teams to channel Big Data talent into enterprises and startups. We all know that "Big Data" skill-sets are a rare commodity and that companies "really need people who can manage it and find insights in it" (where "it" is Big Data). Several times the authors talk about the ability of Data Scientists to write code, which the article calls "Data scientists' most basic, universal skill", and warns hiring managers not to "bother with any candidates who can't code". Strangely enough, however, there is a two sentence call out buried within a 5 page article that, when referring to the coding skill, says "this may be less true in 5 years time". There is no other content within the 5 pages that explains this caution or what to do about it.
How is it that the sexiest job of the 21st century has a shelf life of 5 years? One word -- Commoditization. Just like a brand, a product or a service, a skill needs "defensible differentiation" in order to be valuable over any significant period of time in the job market. What is really strange about the HBR article is this; the one skill they point to as the foundational skill to filter all candidates by (ability to code) has no form of defensible differentiation at all. In fact, the HBR article itself notes that the "Insight Data Fellows Program" can take post doctoral candidates from a wide range of science fields and turn them into data scientists in 6 weeks! Furthermore, for those claiming that "post doctoral" is equivalent to "defensible differentiation", take a look at the National Science Foundation's latest figures on the growing horde of unemployed PhD holders.
If you are thinking to yourself that this article is saying "there's no money in being a data scientist", you're missing the point. Of course there is money in being a data scientist; and I am absolutely sure that anyone who can master the technical skills will get a very nice job. The point I'm trying to make is this: The technical skills that the HBR article says are "critical" can be acquired by anyone with sufficient technical acumen over the entire world; i.e., Asia.
The Defensible Differentiators
Fear not, Big Data enthusiast! There is hope for a sustainable competitive advantage in the big data job market, you only have to look at the ground covered by your UX comrades and best selling author Dan Pink in his masterpiece -- Whole New Mind. The folks who write at HBR must be of two minds, given that several months later two guest bloggers wrote a little piece on how storytelling was the real job of Data Scientists. I concur with the authors here, but I would articulate it a little differently; The defensible differentiation strategy lies within the application of how cultural grounding can be leveraged to first identify hidden insights and then woven into a story.
The original article does indeed reference this, but it is completely minimized to an afterthought. When talking about finding the data scientists you need, the article lists 10 tips, with item number 8 advising the reader to "make sure a candidate can find a story in a data set and provide a coherent narrative about a key data insight". These two skills are what will, long after the 5 year expiration date, keep the title sexy while the code-centric masses are showing their age.
Silver is The Gold Standard
To prove my point, look no further than Nate Silver. Silver is now a household name because of his electoral predictions, but Silver first came into the public eye when he developed and sold PECOTA; the standard system for forecasting the performance and career development of Major League Baseball players. Silver has bumped back and forth from ESPN to the New York Times and then back to ESPN, not upon the strength of his coding prowess, but on the strength of his ability to see meaning within the same data that everyone else has access into.
Are Silver and his brethren scientists, artists or something in between? Data Archeologists (given that they sift through mountains of the past to find truth)? Data Prophets (given that they predict future events)? Maybe we should be looking somewhere else for a metaphor give that only in science fiction have we seen the likes of how the masses of data are revealing the future; maybe what we are seeing is, what Asimov was predicting in Foundation (quite possibly the greatest data-based, future telling, saga of all time). Is this the rise of the Psycho-Historian?
Image courtesy of Jordan Weissman (The Atlantic)