Big Data Talent War: 7 Ways To Win

What makes a great data scientist? Technical chops are essential, of course, including the ability to devise algorithmic solutions to solve complex business problems. But the qualified applicant must also be able to effectively communicate these data-driven insights to a less-technical folk, including the vast majority of business users.

In a phone interview with InformationWeek, Dr. Olly Downs, senior VP of data sciences at Globys, a cloud-based marketing analytics firm, said that finding this balance is often tricky.

"I don't feel like we've solved it, but it's a real challenge point for the industry: How to present robust, scientific results and convey that kind of robustness and integrity, but at the same time, make that information accessible to a broad audience," Downs said.

Globys runs highly contextualized marketing campaigns for mobile carriers via SMS, outbound telemarketing, electronic direct mail and other channels.

[ Big data was the clear winner in the presidential election. Learn Nate Silver's Big Data Lessons For The Enterprise. ]

Before joining Globys to manage a team of six data scientists, Downs held chief scientist positions at Atigeo, a big data analytics company, and Mindset Media, an ad-targeting firm acquired by Meebo in Feburary 2011. He's also served as director of research at Pelago (bought by Groupon in April 2011) and was principal scientist at INRIX, a traffic information services provider spun off from Microsoft Research.

There's a well-publicized shortage of data scientists. An oft-quoted 2011 study by the McKinsey Global Institute, warns the U.S. could face a shortage of up to 190,000 data scientists by 2018.

Why the shortage? One reason is that the job requires knowledge of a wide array of technical disciplines, including analytics, computer science, modeling and statistics. But Downs believes there's more to it than that.

"It is a challenge to find qualified people, but the challenge is shifting a little bit," he said. "It used to be that people didn't know what this was. There were people with this expertise -- the right expertise -- but in very disparate areas of science."

That's changing, however.

"What we experience today, actually, is there are lots of people who want to be a data scientist," he said. "They've perhaps dabbled in it and applied common, nicely packaged machine-learning algorithms to data to answer a specific problem."

However, these people may not have a great deal of experience in identifying business challenges and developing the right algorithmic approaches to solve these problems, Downs added.

And that's essentially what Downs and his Globys team of six data scientists do: Find data-driven insights that help marketers, including those who aren't mathematically or statistically inclined, apply abstract analytical concepts to real-world problems in the business world.

Downs believes data scientists need more tools to uncover actionable insights.

"There's still a very big emphasis on data management and less emphasis -- because it's much harder -- on discovering useful information, and I think that will come next," he said.

Data scientists aren't magicians, of course, and it's easy to overstate their data-manipulating wizardry. In a June 2012 blog post, Gartner research director Svetlana Sicular poked fun at the business world's sudden fascination with the data scientist:

"Companies are desperately seeking mysterious creatures -- data scientists. Some people claim to have seen them in LinkedIn and Target. Perhaps, those were encounters with data scientists from LinkedIn that shop at Target? Or Target data scientists who search on LinkedIn for pregnant teens? Either way, the companies are desperate (except for LinkedIn and Target). But they are seeking anyway. Why? Because nowadays, everyone wants to compete in the new, data-driven economy, where Google and Amazon have already figured out 'data alchemy' -- turning data into gold," wrote Sicular.

Sure, data scientists may not solve all the world's data-related problems, but they can help glean a little extra meaning from those petabytes of information that keep rolling in.

In-memory analytics offers subsecond response times and hundreds of thousands of transactions per second. Now falling costs put it in reach of more enterprises. Also in the Analytics Speed Demon special issue of InformationWeek: Louisiana State University hopes to align business and IT more closely through a master's program focused on analytics. (Free registration required.)