- Businesses need specialists who are handy with advanced analytics but have the domain expertise to communicate with business teams
- Because the U.S. labor force lags behind in data skills, companies are developing analytics talent internally
- Data translators helped rescue one company from bankruptcy
As an analytics manager at a major consulting firm, Jordan Levine noticed a pattern several years ago with client companies—in the pharmaceutical, packaged goods, and other sectors—that were launching new digital services using advanced tools like machine learning.
Technical teams and business units often had trouble communicating. New digital tools promised big benefits, but the solutions were so complex that business teams didn’t grasp their potential. “Business units would ask for a product,” says Levine, “but after a few months the product that arrived didn’t necessarily solve their problem.”
The source of the friction wasn’t the technology, but the increasingly complex language used to communicate about it—data analytics. Not even a decade ago, companies relied on conventional roles like product managers to bridge the business-technical divide. They could do most kinds of statistical analysis in spreadsheets and present them in PowerPoint.
Today, as companies look to deepen their expertise with machine learning and other advanced technologies, they need new types of specialists who are familiar with data science tools like R and Python to understand statistical and algorithmic modeling, and can also explain—and sometimes sell—the outputs to product and business teams.
Enter the analytics translator—a role that many companies are now hiring for and, more often, recruiting and training for in-house. U.S. companies alone are expected to hire two million to four million people for these roles between now and 2026.
“Analytics leaders realize that translators are a critical new support function to business,” says Levine, who is now a partner at Dynamic Ideas, a firm that develops analytics training materials for executives.
Filling the data skills gap
Data translators are in big demand but in short supply in the labor force, in part because the skill set isn’t a native one. That’s especially true in the United States, which ranked 21st in a study by the National Center for Education Statistics that assessed the data-based problem-solving skills of adults in 23 developed countries. That’s one reason why many companies and philanthropic organizations are backing digital literacy programs to help address skill shortages.
At companies where the need for analytics translators is urgent, ad-hoc training programs can have both immediate and far-reaching influence. Back in 2015, for example, Levine started a first-of-its-kind training academy at the consulting firm. Several colleagues and trainees in that program are now putting those skills to work in leadership roles at other companies.
In 2017, for example, Michael Joyce—who ran the training program after Levine—joined JOANN, an Ohio-based company that sells fabrics and craft supplies at more than 850 U.S. retail stores, as an analytics manager.
Today, Joyce is JOANN’s SVP of supply chain optimization and leads a team that relies heavily on data translators to tackle questions like how to distribute products across stores—or in some cases, how to determine when advanced modeling techniques are overkill.
For example, the team recently discussed building a predictive analytics tool that could forecast demand for core products. But they realized the existing model was sufficient because many of JOANN’s products are slow-moving. As a result, there wasn’t enough business value to warrant building the new tool. Analytics translators helped make the case to business units.
One student of Levine’s training program, Tim Eisenmann, has led a similar revolution in analytics at North Carolina–based American Tire Distributors since joining the company in 2017.
Eisenmann, now ATD’s chief analytics officer, created an internal analytics center of excellence. That team’s work helped pull the company out of Chapter 11 bankruptcy in 2018, according to a Charlotte Business Journal report. As Levine noted in a LinkedIn post about his former colleague, the team built a tool to help ATD’s sales associates recommend tire brands that would make retailers more profitable. Data translators proved to be key in explaining the data models.
New career paths
Other types of analytics translators are finding their way into these roles on their own. Two years ago, while completing his master’s degree in organizational psychology, Brad Clum took an internship in HR analytics at Intuit that set him on a new career path.
Today, the 33-year-old is an analytics translator at Genentech. He works with both business and technical teams to create internal HR tools backed by advanced analytics. One of his recent projects was a calendar app that lets employees set goals—for example, to improve productivity by securing more time for focused work—and recommends scheduling changes that data suggests will help them achieve those goals.
To identify these needs, Clum surveys and interviews end users (including execs) using an Agile-based process. “I take the feedback, translate it into user stories, and then prioritize them from top to bottom,” he says. Clum then typically meets with data scientists and developers to discuss requirements and options.
When a specific tool is ready for testing, Clum monitors its usage and communicates its functions and benefits to users, making sure they take full advantage. Like most analytics translators, Clum doesn’t specialize in coding, but he has enough knowledge and experience selling analytics products to comfortably discuss which statistical models are appropriate for a given problem.
The common thread in each of these cases is a business-centered approach to navigating new technologies. Over time, says Levine, data translators will become ubiquitous in the enterprise. After all, as business leaders become increasingly surrounded by digital tools, they will have an ever-growing need to know how to use them.
|Fluency with modern analytics||Analytics translators don’t do a lot of coding themselves, but need to be familiar with data science tools like R and Python and which types of statistical modeling best apply to specific business problems.|
|Strong industry knowledge||Solid domain expertise allows translators to understand the business value of machine learning and other advanced technologies.|
|Great communication skills||Translators are often called upon to present analytics findings and outputs in an understandable format to business teams.|
|Project management||Translators need to manage an analytics project from pitch to production and adoption, a process of complex dependencies with multiple teams.|