360 Guide:

Why the future of IT jobs may give techies a pass

olly - Fotolia

Citizen data scientist trend compensates for lack of skills

Self-service analytics tools are fueling the citizen data scientist trend, helping companies compensate for a lack of analytics skills in the market and making analytics more accessible.

There's no doubt that data science has been a game-changer for some businesses. But data scientists, the magicians behind the curtain, remain elusive. So, rather than pursue the unattainable, some enterprises are opting to put the power of data science in the hands of business teams.

"I want [managers] to be able to get answers without needing to employ a data scientist," said Todd Knapp, owner of Envision Technology Advisors LLC, an IT consulting firm in Pawtucket, R.I.

This sentiment is becoming more common today. With the scarcity of data scientists, and the high price that comes from hiring one, many businesses are opting to simply let business users get insights for themselves. This citizen data scientist trend is becoming more prevalent thanks to emerging tools that automate large portions of the data science process.

For example, Envision uses a tool from Sisense that continually reviews data sets and alerts business users to interesting features in the data, such as outliers or statistically significant changes in metrics over time. This isn't so much about the more intensive data science activities, like machine learning, but they are tasks that typically involve some advanced math. By automating them, anyone can get insights that used to require that a data scientist be involved.

"With most data applications, you do have to be a data scientist," Knapp said. "You have to see a question and dig to get the answer."

But, he added, new analytics tools diminish the need for hands-on data science.

Data democratization breaks down silos

One of the advantages of this approach, Knapp said, is that it breaks down silos. There's no data team working on reports for business units, which might themselves be using their own tools and developing different reports. Instead, everyone has some responsibility for using data, and everyone uses the same tool.

But just because analytics is automated in citizen data scientist shops doesn't mean self-service is strictly a business initiative. Olga Polyakov, who runs a Tibco Spotfire self-service analytics deployment at the college preparatory school Colegio Nueva Granada in Bogota, Colombia, said she couldn't do it without IT's involvement.

"It's not a one-woman shop," said Polyakov, the school's assessment data analyst.

Polyakov uses Spotfire to produce reports tracking student progress on standardized tests and learning goals, enabling her to see how current scores compare to long-term trends.

Polyakov, whose background is primarily in education and research, said that while the analytics can get complicated, she can handle all of it. But the tech side of things is more challenging. She leans on IT to manage the software on the school's servers and to install updates, as needed.

Citizen data science doesn't come free

This is one of the big cautions to the whole citizen data scientist movement. While it doesn't take a Ph.D to extract deep insights from data these days, it does require some technical know-how to run on the back end.

Another major caveat, said Gartner analyst Jim Hare, is that self-service data science tools can only take a business so far.

"The limitation is that, if you're trying to do something that's cutting edge, these tools don't have it," he said. "But if you're trying to do something basic, like a linear regression, you can do it."

For this reason, Hare doesn't see self-service analytics tools replacing data scientists anytime soon. Enterprises getting the most bang for their buck are using these tools to augment data scientists, allowing them to extend data science capabilities throughout the organization.

"Businesses are looking for these easier to use citizen data science platforms," Hare said. "And they can take people who have some skills with stats, and then use [them] in operational use cases."

Next Steps

Hiring data scientists is still challenging due to a lack of skills

Analytics executives share advice on hiring data scientists

Can't find a good data scientist? Consider a physicist

Dig Deeper on Advanced analytics software

360 Guide: Why the future of IT jobs may give techies a pass

Article3 of 8

Up Next