In This Issue:
Conflicting Advice: Do You or Do You Not Need a Data Scientist?
In an era of Big Data, data science has become one of the hottest occupations as more and more companies aspire to reap the benefits of being data driven.
Yet, a recent article in Entrepreneur (March 23, 2018) is “Think Your Company Needs a Data Scientist? You’re Probably Wrong” by Talia Borodin. However, a link to related articles in Entrepreneur takes readers first to a piece about managers not understanding how to hire data scientists, then there is a link to another related article titled “Want Big Data to Help Your Marketing Team? Hire a Data Scientist” by Brian Kardon (June 26, 2015). His article makes the case for why marketing departments should hire data scientists. These articles may seem to offer conflicting advice, with one saying you don’t need a data scientist, while the other tells you to hire them.
Still another link goes to a related article written by a general manager at IBM. He explains that even smaller organizations could benefit from Big Data, although the example in his article has 100 employees, small by IBM standards, but perhaps considered large to many entrepreneurs. However, this muddles a possibility that company size might explain why Borodin says you don’t need to hire a data scientist, while Kardon encourages hiring them.
Nonetheless, both the you don’t need one article and the hire them article do have something in common. Both articles discuss what it takes to successfully bring data science into an organization. Both point out that there are challenges. The hire them article warns that “the question becomes how to make sense of it all” and stresses the importance of having people who know the technology and have business acumen and who understand what data to collect, how to get insights from the data, and how to effectively communicate the findings. Also focusing on the challenges, the you don’t need a data scientist article points out the difficulty in hiring a lone data scientist, especially a junior one, when a company has no idea what they want to get from the data.
A major difference between the two articles lies in how they seem to perceive the challenges. It’s as if the glass is seen as half empty by the you don’t need one article and as half full by the hire them article. This very well may reflect the fact that the hire them article tends to be oriented toward companies with more extensive resources, and perhaps already with staffers used to analyzing market research data. My blog has discussed the value of integrating existing data capabilities, such as those from a traditional market research department, with the newer emphasis upon Big Data.
The you don’t need one article is more oriented toward companies that have done little, if any data analysis in the past, have fewer resources, and are literally starting from ground zero. For them, hiring a data scientist, especially a junior, makes much less sense. Furthermore, in her you don’t need one article, Borodin also points out that many of today’s sophisticated data tools require huge volumes of data, and companies without much data won’t really benefit from hiring a data scientist.
What can we learn from these articles? Successfully becoming data driven requires much more than just technical skills. It also calls for an understanding of the business that can lead to insights. As with any analysis, somebody has to understand the business well enough to work on problem definition. And, somebody has to present the results. Often, all of these diverse skills won’t be found in a single person.
In her article, Talia Borodin tells of the huge change since she started her career 15 years ago. She says, “I could never have envisioned the sexy rebranding of my work with the coining of the term data scientist, let alone the immense popularity it’s achieved in recent years”. Borodin's statement is something I relate to because in my early career in data far more than 15 years ago, I, too, could never have envisioned its rebranding and tremendously increased popularity. Back then, I was that era’s version of a data scientist, though the term data scientist hadn’t been coined yet. In my work, I saw the value of data analytics and I also saw its limitations. That fueled my lifelong quest researching business success and failure patterns, which went well beyond the numbers and encompassed a much broader understanding of business dynamics. And, that broader level of skill, that goes well beyond the mere technical, is needed in data science today.
That’s why successfully implementing data science takes much more than just hiring a single data scientist. And, that’s why Borodin discourages companies from hiring a data scientist if their business isn’t really ready for it. She does explain, however, that hiring data scientists can be beneficial for some companies—those that put real thought into what they are trying to accomplish. But, many companies are not ready to do this. It’s also why Kordon’s article, which encourages the hiring of data scientists, explains what companies must do when hiring them.
Finally, before concluding this article, I’d like to point out an example from the March 29, 2018 Wall Street Journal article “The Data Wonk Who Became a Coach” by Jared Diamond. The article, which is about Big Data in baseball, illustrates the tremendous effort organizations go through to become data driven. The article tells how the defending world series champion Houston “Astros sent Sig Mejdal, one of baseball’s most respected quantitative analysts, to spend the 2017 season with the Tri-City Valley Cats, a single A affiliate near the bottom of the minor league ladder.”
Although Mejdal joined the Astros as “director of decision sciences”, in the field “he coached first base, helped with batting practice and hit grounders to anybody who requested extra work. He endured the six-hour bus rides…cheap motels, and subsisting on late night drive throughs.” He “struggled to shake the feeling that as an analyst from the front office he didn’t belong in this world.” It helped that he had the support of the team manager. Next year, the Astros plan to have Mejdal work with all four minor league teams, instead of just the one with so many new recruits, where he started working in the field this year.
Granted, a team like the Houston Astros has far more resources and budget than many entrepreneurs who may make up the kinds of businesses Borodin says should not hire a data scientist. But, the Astros example illustrates how important it is for data science to be integrated into the organization, rather than just analyzing data in insolation. Along the lines of what Borodin explains, companies that aren’t ready to do what it takes, and think all they have to is hire a data scientist, probably shouldn’t hire one, especially a junior one. On the other hand, companies that make the effort to find ways to combine knowledge of the business with the technical skills of data science can reap benefits from becoming data driven.
La Grange Park, IL