Rise of the Quants
With the rise of big data, more HR leaders are looking to staff their departments with highly analytical people who can help them translate the data into business insight.
By Andrew R. McIlvaine
John Hausknecht knows a thing or two about the growing appetite among employers for analytically talented HR staff.
"For the last few years, we've seen a huge surge in demand for workforce-analytics roles -- recruiters coming in here and searching for individuals who will exclusively do analytics work in HR," says the professor of industrial/organizational psychology at Cornell University's School of Industrial and Labor Relations.
Indeed, the demand has become so great that the I/O department added a course in analytics to its undergraduate program in 2010, and its enrollment has been at overcapacity every semester since then, he says. Next year, the department plans to add special courses in statistics for HR undergrads who want to specialize in analytics, as well as a "capstone" project in which they will work in teams analyzing workforce data at a participating company, says Hausknecht. "They'll work with the company to analyze the data, report on their findings and issue recommendations based on what they've observed," he says.
Other universities with I/O programs are making similar moves.
"The University of Wisconsin is making a big push into big data and analytics for HR," Hausknecht says. "People from other universities have sent me syllabi to review for courses they presumably plan to offer in HR and analytics."
What's behind this surge in interest among companies?
"I think, from a human capital standpoint, they are turning to analytics to try and understand how to beat the competition in attracting and retaining the best people," says Hausknecht. "For some companies, it's sort of a culture shift to be more data-driven."
Many of the nation's most-respected large companies, including Google and IBM, are centralizing the analytics functions in their organizations, creating teams of people to focus solely on enterprise data, rather than -- as has traditionally been the case -- having it simply be an add-on to an HR job, he says.
"Businesses are increasingly asking HR to stop reporting on what has happened, and instead tell them what they should do -- who to hire, what makes people quit the organization, who is more likely to be absent and how to control that, who has leadership potential," says Ranjan Dutta, director of PwC's Saratoga Institute.
Naturally, this is driving ever-increased demand for so-called "quants," people with numerical and analytical skills who've traditionally made their homes in areas such as finance and marketing.
"HR is emerging as an attractive place for quants," says Thomas Davenport, author of the best-selling book Competing on Analytics and co-author of a new book, Keeping Up with the Quants.
Of course, not all companies have the wherewithal to lure the sought-after grads from elite programs such as Cornell's. And having quantitative skills isn't enough -- to be effective, experts say, these staffers must have a good understanding of the business and be able to translate the results of their work into something that's usable by business leaders. The alternative could be wasted money and talent.
"You can have the best statisticians in the world, but if they can't tell a story around the data, it won't help much," says Amy Armitage, director of member research programs for Seattle-based consultancy Institute for Corporate Productivity.
From his perch within New York-based Mercer's workforce sciences group, which he co-founded and leads, Haig Nalbantian sees more and more companies taking the plunge into workforce analytics.
"Almost all of our work, up until about five years ago, was work that clients outsourced to us," he says. "The big difference nowadays is that organizations have finally decided they can't drive blind in this era when talent is such an important factor in determining the competitiveness of a company."
Today, many of those same companies are using in-house resources to perform this work and they are, by and large, placing these resources within HR.
"The value of tapping all this data you're sitting on -- you've spent so much on these enterprise systems and you have this amazing data that, if properly analyzed with the right lens, can be extremely valuable in helping you figure out where to direct your resources," says Nalbantian.
It can also mean dispelling erroneous assumptions about employee behavior, he adds.
For example, a large bank Nalbantian worked with looked at its data and found a strong relationship between the turnover rates of supervisors and their direct reports. If the supervisors left, the data suggested, there was a strong probability that their direct reports would also leave the organization.
A number of theories were proposed to explain this apparent linkage, chief among them that the employees were inclined to leave due to the departure of a supervisor who had served as their guide and mentor. Another theory held that every action within an organization sends messages about peoples' perceptions of their environment. In other words, a supervisor leaving might signal to her direct reports that better opportunities lay outside the organization.
How best to prove which theory was correct?
Nalbantian's team created a "simple statistical test" that looked at what happened when a supervisor transferred to a different part of the organization instead of leaving. When this factor was added to the model, he says, "lo and behold, the relationship between supervisor turnover and employee turnover disappeared," he says. "It was much more a matter of signaling that better opportunities lay outside the organization that triggered the direct reports to leave, rather than the personal relationship between them and their supervisor."
In another example, Nalbantian's team was able to help a large professional-services firm stop hemorrhaging newly hired female partners. By carefully analyzing the firm's performance-review data, they found that the company -- which had recently switched to a forced-rankings model -- was inadvertently chasing the new hires away via its evaluation system. Because the female partners were new to the teams, Nalbantian found, they were much more likely to receive lower ratings than employees with longer tenure. Retention rates improved when the firm exempted the female partners from performance reviews during their first year, he says.
Modeling data in this fashion can reveal more about an organization than traditional measurements, says Nalbantian.
"It used to be all about benchmarking," he says. "But our research has concluded that context is everything in this area -- what works in one environment may not work in another. Companies are starting to get that now."
At Collegedale, Tenn.-based McKee Foods, the maker of Little Debbie's snack cakes, the 5,000-employee company has focused on building an HR-analytics team from scratch, says Mark Newsome, the company's senior corporate HR manager.
This approach is part of the HR department's shift from measuring things that may not be so important to the business -- traditionally, HR measured turnover, but McKee Foods has been experiencing very low turnover -- to actual trends that can help the company plan its strategy.
"We're starting to spend time creating measurements that look at the turnover among key positions or among high performers," he says. "Instead of measuring time-to-fill, we're trying to measure quality of hire."
Chuck Kleinknecht, human resources director for Bethesda, Md.-headquartered Lockheed Martin, says the HR department at the aerospace-and-defense giant is also very interested in analytics.
Kleinknecht attends meetings of Lockheed Martin's Enterprise Leadership Council, sitting alongside counterparts from engineering and finance to present data analyses their respective departments have conducted during the preceding months.
In HR's case, the analysis looks at trends in areas such as attrition, staffing and engagement to help the company's leaders anticipate what the future will look like in terms of human capital, he says.
"We are, like many other companies, moving beyond reporting on what has happened and instead, through reviews of our big data, looking at ways we can continue adding value, highlighting concerns and uncovering any trends going on as we predict what we'll need in the future," he says.
Lockheed Martin is -- along with other high-tech companies -- competing to hire lots of engineers and scientists and wants to make itself a sought-after destination for such talent, says Kleinknecht. HR also creates "talent indexes," based in part on data extracted from individual development plans, to help the company's business units determine whether they have what they'll need to stay competitive. LM's efforts in this area have won acclaim: Last year, the company's Missiles and Fire Control division was honored with a Malcolm Baldrige National Quality Award for, in part, its use of HR analytics for workforce planning, he says.
At Regeneron Pharmaceuticals, a 2,000-employee bio-pharmacuetical firm based in Tarrytown, N.Y., the culture is "highly quantitative," says Ross Grossman, vice president of human resources.
"You can't be credible here if you only talk about the soft stuff," he says.
Grossman is leading an effort to "get more analytical talent into HR," and not just so the function will mesh more tightly with the organization. He's focused on helping Regeneron find and hold on to sought-after scientific talent so the company won't have to worry about churn as it develops new drugs to fill its pipeline.
"We do a big engagement survey every year and we tear the findings apart when they come in, looking for things that will make good people want to stay here, because our business model requires really engaged people," says Grossman. The company is looking to increase its bench strength in this area by luring graduates from universities that have strong programs in big data and analytics -- including Columbia University, from which he hired a graduate "who's great at looking at large amounts of data and coming to just the right conclusions."
At the moment, Grossman and his team are carefully examining the latest survey data and have already uncovered some "concerning" anomalies: Although the results from questions that measured engagement and emotional attachment to the company "got extremely good results," a question that measured intent to stay with the company did not score as well. So the team is delving into data "two or three levels down" and is convening focus groups to determine what steps to take to boost those numbers.
Where to Find Them?
Skilled HR analytics people are much more than statisticians, says Saratoga Institute's Dutta.
"You need someone who can actually make sense of this data and convert it into business insight, and this tends to trip organizations up because they can hire a very skilled statistician, but the models they build aren't going to be of much use unless they're translated into things that make sense to the business, and that's a rare skill," he says.
Dutta has seen companies build centers of excellence staffed with graduates of industrial/organizational psychology doctoral programs remain underused because no one knew what to do with the results generated by these staffers, who tended to have little or no actual business experience.
"These folks built a very nice model that looked at historical data, at all the characteristics prized by the organization, but when they presented the results, no one knew what to do with them," he says.
Which is why Dutta believes HR should look to experts in the finance and marketing departments and lure them over to HR.
"The answer is not to hire three I/O Ph.D.s and [say] 'done,' the answer is to maybe hire someone from a different function who's done this before, who can look at these deep models and translate them into real business insight," he says.
Because finance and marketing departments have been doing this sort of work much longer than HR has, they are much further along this continuum, he says. Encouraging their analytics experts to move over to HR may be a relatively easy sell, he adds.
"These tend to be ambitious people and, if someone wants to go higher in today's organizations, what's most valued is a well-rounded person, someone who's worked in finance, marketing and HR, rather than in just one function," he says. "So present it as a developmental opportunity."
Some companies operate shared centers of expertise between different departments, but Dutta says the ideal situation is to have one exclusively for HR. "I think the jury is still out with respect to whether a shared center works better," he says. "The difference in maturity level between HR and other functions in predictive analytics is significant -- marketing and finance have bigger needs and a more mature outlook, whereas HR is just starting out."
Wes Wu, managing consultant with Knowledge Infusion in San Francisco, agrees that finding statisticians in the finance department and luring them over to human resources, rather than trying to build staff from the ground up, represents the most practical move.
"When you can get someone from finance who actually understands the organization, who's done this type of work and has maybe had to partner with HR in a previous life, that's a really quick and easy stepping stone to building this kind of capability within your HR department," he says.
An HR function that is strongly aligned with the business, in which HR business partners can serve as the interface between the business and the quants within HR in order to uncover correlations within data, is the ideal arrangement, says Wu.
"Anyone in the world can tell me, for example, what the five-year trend is for turnover among African-American females in the organization," he says. "What I want to understand is how can I correct for a downward trend -- is it correlated to a set of competencies, compensation elements, benefits or office locations? That's when the analysis gets interesting."
What's key is for HR leaders to ensure the work done by these "human capital scientists" is actually put to good use, says Nalbantian.
"There's nothing more disheartening to a data scientist than to have leaders who bring you in because they think they've got to do this but, when push comes to shove, will fall back on conviction [or] philosophy, or just claim bias about what works and what doesn't in the human capital management realm," he says.
In addition to staff in its various HR Centers of Expertise, Lockheed Martin has 40 employees in its enterprise HR-analytics operation who supply data to HR business partners, says Kleinknecht. The business partners, in turn, use the data to help the company's executives detect trends and spot potential areas of concern, he says.
At LM, HR staffers are encouraged to enhance their skills and knowledge of the business by rotating between different areas, with the ultimate goal being to help them become more valuable and engage with their clients, he says.
"We think it's important for folks to explore different areas and then decide which areas of HR make the best use of their talent," says Kleinknecht.
By working with colleges and universities that offer degrees in industrial and labor relations, he says, the company recruits HR staffers with a good educational foundation and then builds on that via the rotational assignments, he says.
These staffers are supplemented with more-experienced professionals, such as doctorates in industrial/organizational psychology, to identify trends and correlations between data sets, such as degrees of satisfaction with organizations' recognition and reward programs.
Grossman has hired a "a young guy who's really great at looking at large amounts of data -- I can put him in with our senior scientists and he's perfectly comfortable talking data with them," he says.
McKee Foods, which is in the process of acquiring the Drake snack-cakes business from bankrupt Hostess Brands Inc., has three employees in its HR department who -- in partnership with several IT counterparts -- spend much of their time on analytics.
Newsome recently sent the HR team working on business analytics to a training session on business modeling similar to the work the logistics department -- a vital function at McKee Foods, which distributes its baked goods nationwide -- is doing to ensure alignment between the two functions.
"This represents an unprecedented opportunity for HR," he says.