Every company’s online application form includes a selection box asking candidates how they found out about the job—from LinkedIn, the company’s website, an employee referral, and so on. Part of the reason for the question is so human resources departments can gather data and optimize the best way to source candidates. At least that’s the idea.
But it doesn’t always work out that way. Candidates overwhelmingly select the first option in the drop-down menu regardless of what it is, says L. J. Brock, chief people officer at the digital currency exchange Coinbase. Brock conducted an experiment at his previous job where he rotated the first choice in the selection box every week—one week it would be Google, the next an email job list, etc.—and every week the first choice ended up being the biggest referral source.
The simple conclusion is that candidates weren’t taking the choice seriously. They were just clicking a box to move on in the application process. But for Brock, the results underscored one of the biggest issues that human resources leaders face in using data to generate insights. “If you don’t have the right systems and processes in place to capture the right data the right way every time, whatever insights you have won’t be real,” he says.
To be sure, the shift toward data-driven decision making in HR, underpinned by algorithms and machine learning, is uncharted territory for most organizations. Organizations collect education, experience, and performance data for employees, of course. But they have little to show in terms of how HR can help drive business strategy from that data.
“The problem lies in the fact that HR functions have tons of data but no real expertise in how to manage and analyze it to create actionable outcomes,” says Deepali Vyas, senior client partner and global co- head of Korn Ferry’s fintech practice.
Vyas says that increasingly, organizations—particularly those in the Fortune 500—are trying to solve the problem by creating an “HR data scientist” role. In its ideal form, the HR data scientist combines employee performance data with personal, environmental, social, and other external facts to create strategies to improve employee experience, cross-company collaboration, productivity, and employee well-being.
Moving past simply analyzing survey results is the next data frontier for HR leaders, says Brock. A dedicated HR data scientist can help generate insights to “direct the actions of leaders and employees,” he says.
Imagine a head of sales or marketing going to the CFO with a new capital expenditure project that will cost millions of dollars. Undoubtedly, the request would be met with a barrage of questions, among them what the money is being used for, how it will be divided by region, and of course what the expected return on investment is. Now imagine the response to those queries being “I don’t know.”
That would never happen, because the head of sales or marketing would never make such a request without working up a full forensic data analysis of the investment. But HR functions frequently lack the same kind of analytical sophistication when it comes to their initiatives, even though talent costs can range from 15% to 40% of an organization’s overall capital expenditures.
“HR hasn’t been speaking the same language as other business units,” says Jeremy Welland, global head of data and analytics at PayPal.
These days, however, an overwhelming number of critical business decisions revolve around finding top talent, developing a healthy corporate culture, and focusing on various people-related issues across the talent life cycle. HR leaders must decide and execute on talent acquisition, leadership development, human resources technology platforms, compensation, performance management, culture, diversity and inclusion, learning and development, succession planning, and employee relations, among a host of other requirements.
Dan Kaplan, a senior client partner with Korn Ferry’s CHRO practice, says HR leaders must start leveraging data and analytics to show the payoff of talent investments in ways the CEO and board can understand. “They need to back up their work,” so to speak.
Kaplan also says HR leaders need to think bigger—in terms of both initiatives and scale. On the initiatives side, HR leaders need to show how they are helping achieve strategic business goals across the organization. On the scale side, processes need to be designed with AI and automation in mind so that the data they collect can deliver on those goals in real time.
This is where an HR data scientist comes into play. Having a data leader in place can speed up the time it takes to build systems, and that person can help decipher links between data points and their business value, which might not be seen or understood by traditional HR leaders. For instance, the data may shed light on career pathway patterns, or on a set of specific reasons that employees are voluntarily leaving the organization. But if there is no data set to analyze a particular thesis, who will know? “Success means thinking big enough to link analytics to outcomes leaders care about achieving,” says Welland.
For many HR leaders, AI is used primarily in talent acquisition, as this is the area where companies see significant, measurable, and immediate results. AI can reduce time to hire, increase productivity for recruiters, and deliver an enhanced candidate experience that is seamless, simple, and intuitive.
But for all its value, AI also has significant drawbacks for HR leaders. It can be biased, for one. It also can’t account for intangible traits such as communication and collaboration skills that are hallmarks of high-potential talent. “AI-based recruiting tools are not bespoke,” says Vyas. “They can’t tell HR leaders what will make someone successful in their company.”
By contrast, Vyas says a dedicated HR data scientist can add diversity of thought to talent recruiting and retention. They can reverse engineer what makes an organization’s best employees successful by putting a sharper filter on how skills, experiences, personality traits, organizational culture, and other factors combine—and how that leads to success or failure.
Put another way, an HR data scientist can help HR leaders better identify what factors lead particular sets of talent to sustained high performance, which is particularly valuable for roles where success is more subjective and the determination could vary depending on who is doing the evaluating. Rather than hiring for a strict set of capabilities, an HR data scientist can also help identify where skill gaps exist within an organization and thereby open up recruiting or internal development to a broader set of talent to fill them.
That’s why, though he’s only been in the role for less than a year, Brock says he has big plans is to develop a “forward-deployed data team” for Coinbase’s HR operation. “We want data at the front of HR, not in the background.”