Why Hire Expensive Data Scientists?

Why Hire Expensive Data Scientists?

Data science 101: With affordable analytics services popping up, why hire costly manpower? Here’s a quick insight.

Dubbed the “sexiest job of the 21st century,” data scientists are in-demand, which of course means they are also expensive, with an average salary of $119,000, according to Robert Half Technology.

A shrewd executive might also observe more and more machine learning services becoming available. So why pay for a premium to construct an internal data science team, if you can just plug data into a service?

Great question. We asked Greta Roberts, CEO and co-founder of Talent Analytics (recently acquired by Hire Smarter, Inc.), to explain the real role of data scientists and offer concise guidance on structuring a data science team.

Connected Futures: How do you explain the data scientist job to a CEO?

Greta Roberts: There’s a lot of talk about how data is the new currency for organizations. You need to have people that mine and extract value from the data that you’re acquiring inside your business.

Data scientists are the miners who are helping your company be smarter, faster, and more competitive.

Why does a company need a data scientist when there are easy-to-use, cloud-based analytics solutions?

A common misconception is that dashboards deliver answers. They don’t. In reality, analytics solutions only account for 25 percent of the total effort.

A quarter of your effort goes into framing the problem you’re trying to solve with data, and identifying what specifically needs to be optimized and predicted, if you’re doing predictive work.

Fifty percent of your effort goes into getting the data laid out correctly.

Only after that does the final quarter of the work go into something like Watson or other analytics solutions.

That final quarter doesn’t matter if the other three-quarters of the effort isn’t there. Those three-quarters are what the data scientists do.

What’s the payoff for that investment?

CxOs need to define their value by understanding what types of projects data scientists will be working on. Data scientists need to be working on projects that affect either revenue or expenses at your organization.

For example, maybe you look at what employee attrition costs the business. There’s a retail operation that has close to 100 percent turnover of sales reps in their stores every year. They lose nearly $14 million annually just in turnover. If you can use data science to figure out how to reduce that to even 85 percent, that’s massive.

Focus on areas in which you have a lot of business problems and ask which ones could be addressed with an analytics approach. In solving that those problems or capitalizing on those opportunities, the benefits become obvious.

How do companies best use these employees? Is there a typical or best organizational structure to capitalize on top business priorities?

People talk a lot about a center of excellence. Compare it to how IT is often organized: You might have a core set of IT folks that all work together, but then you may also have some that work just for sales or marketing. In this sense, some data scientists may be embedded in the line of business and also have a dotted line reporting to that center of excellence. This helps them get to know the business really well and understand some of their problems.

It’s not just about who these people report to, though. Because data scientists typically pick up a project after it’s framed by someone else, you need to have business leaders and managers in place who know how to find projects that lend themselves to analytics work.

Did you like this article?

3
0

Smart Data