Employers may have trouble finding data analysts with all the technical skills they want, but some are starting to zero in on general abilities that can lead to analytics success.
By Ed Burns
Few things are more crucial to helping businesses succeed on advanced analytics initiatives than having the right people doing the analysis work and managing projects. But in an environment in which analytics skills are at a premium, organizations are having to get smart about the kinds of people -- and abilities -- they look for.
Speaking at the 2015 MITX Data Summit in Cambridge, Mass., last week, Richard Grogan, vice president of measurement and analytics at Boston-based digital marketing and advertising firm AMP Agency, said data analytics is generally thought of as a technical or mathematical discipline. But increasingly, he added, having the right team involved in analytics efforts is the biggest difference between success and failure.
"When I think about the most important thing, technology is important, training is important, but finding the right people is most important," Grogan said at the conference, which focused on "The Art & Science of Data" and was put on by the Massachusetts Innovation & Technology Exchange, an industry association based in Boston.
But putting together a top-notch analytics team isn't easy when skilled people are hard to find -- or keep. The analytics skills shortage is widely acknowledged. In an often-cited 2011 report, consultancy McKinsey & Co. estimated that the U.S. could face a shortfall of 140,000 to 190,000 workers with analytical expertise by 2018. Partly for this reason, many employers are looking for specific personality traits more than technical skills when recruiting data scientists and other analytics professionals. While a person can learn to code in the R programming language or delve into software like SAS or SPSS, the inner traits that lead to analytics success can't be so easily taught.
More to analytics than building models
Leon Barsoumian, senior vice president of analytics and research at online marketing and ad services firm Havas Media's U.S. offices, said the primary traits he wants to see in candidates are general problem-solving abilities, curiosity and communication skills. According to Barsoumian, those elements reflect how data analysts are being asked to become more integrated members of marketing teams, which means they must be able to take on responsibilities beyond simply writing algorithms and creating analytical models.
"No longer are analysts in the back room doing their models," he said. "Now they're in front of clients trying to explain why their models deliver answers."
For Seb Maitra, executive vice president of analytics at Boston-based ad agency Hill Holliday, the traditional skills of technical ability, statistical knowledge and business expertise are still important. But the overarching quality he looks for in prospective job candidates is the ability to learn quickly.
That's a particularly important skill in today's fast-changing tech landscape, Maitra said. If he were to write up a job listing that included specific technology skills, it could be out of date by the time he brings in candidates for interviews. But the ability and desire to learn new things and to do so quickly are what enable data analysts to stay relevant, he said.
On the hunt for adaptive analysts
Adaptability is one of the top characteristics that Jasme Bantens, a senior partner and digital insights lead at marketing services firm Mindshare's U.S. office in New York, looks for in candidates. She said so much has changed in recent years in advertising itself, as well as the technology that's involved.
For example, Bantens said Mindshare used to rely heavily on media mix modeling, a statistical method for measuring the impact of ads in specific media channels and predicting future performance. But as online and mobile media have become more prevalent, the technique has lost some of its cachet. Now Mindshare is looking to do more predictive modeling and real-time analytics leveraging data from an increased number of digital channels.
Bantens also looks for people who can tell a story with data. There are so many data sources and tools available that can make the process of analyzing data relatively easy, she said. But all of the work is meaningless if an analytics team can't explain to internal business users or corporate clients why the findings are significant.
"How you tell that story will change your client's perspective," Bantens said. "Bringing that data to life and explaining it in a way that resonates with them is big."