The oilfield services industry has an annual turnover rate of about 35 percent among its largely blue-collar workforce. Executives have long been resigned to this fact because they didn’t know what caused it, how much it cost or what to do about it.
In 2007, soon after Ray Lieber was hired as the HR vice president at Superior Energy Services Inc. in New Orleans, he calculated revenue lost for each job type that quit. That got managers’ attention: The business model for the oilfield services and equipment provider is based on billable hours, so turnover among its 4,300 employees equals lost revenue.
"The moment we lose an operator, revenue starts to drop," Lieber says. "We’ve always known it but never quantified it. Now we quantify it."
He also calculated turnover by job type. According to conventional wisdom, turnover was mostly among semi-skilled workers, but he found that nearly half of the people who quit were skilled operators or supervisors with a higher impact on revenue. Having made his case, Lieber sold some of the business unit managers on the next step: Using predictive modeling, the process of creating a statistical model for predicting the probability of an outcome, he identified action with the best chance of stemming turnover—more supervisor training, especially one-to-one coaching skills.
"Turnover is down significantly in two of the three business units where we conducted the analysis and put in new processes and practices," Lieber says. "We were at 34 percent in one, and we’re going to hit 26 percent to 28 percent this year. Even an improvement of 3 to 4 percentage points will show up in the profit and loss statement."
Most HR departments have spent years acquiring gigabytes of data about their employees and installing technology to store and organize that information. But most are still only using the data for transactional purposes—to more efficiently process payroll and administer benefits. They’re not using these rich veins of data to make better human capital decisions.
Lieber and a growing number of HR professionals are using quantitative methods and various business intelligence (BI) tools to analyze the data from their HR databases, corporate financial statements, employee surveys and other sources to make fact-based human capital management (HCM) decisions that impact the bottom line. These pioneers are forging a culture of inquiry and quantification in HR.
Facing the Three-Faceted Obstacle
While many organizations have HR information systems (HRIS) in place, few HR executives capitalize on the data and the technology in ways that can drive business performance—and make themselves true strategic partners. Most business leaders, including HR executives, still do not make people decisions with the same rigor as they do decisions about customers, products, supply chains and business strategies. When it comes to workforce decisions, they too often rely on intuition and conventional wisdom. In contrast, workforce analytics is fact-based decision-making—the use of data, metrics, statistics and scientific methods, with the help of technology, to gauge the impact of HCM practices on business goals.
"Our HR group has been on a transformational journey identifying how we can bring more data-based, fact-based information, not just intuition or gut, to the table to make better decisions about our people and the health of our organization," says Sherry Holtz, director of HR effectiveness for Thrivent Financial for Lutherans in Minneapolis, a financial services organization with approximately 3,000 employees.
Adopters typically first analyze low-hanging fruit—turnover is a common starting point. But only imagination limits use of workforce analytics; applications may include analyses of skills gaps, benefits and compensation packages, succession bench strength, and many other pain points, all aimed at actionable measures. Verifying or busting institutional myths has become commonplace.
"These opportunities excite me and many of my colleagues," says R.J. Milnor, who was recently hired as the first director of HR analytics at Citrix Systems Inc., a Fort Lauderdale, Fla., software developer. "Most companies are flying blind in workforce decisions. It is unacceptable for finance or other functions to do this."
In a recent study, based on interviews with 417 executives, about two-thirds of whom were in HR, researchers at the Aberdeen Group Inc. in Boston found that among companies they deemed best in class—the top 20 percent based on employee retention, performance and engagement metrics—only 28 percent use workforce analytics. Less than half of best-in-class employers had HCM metrics agreed to by key stakeholders.
At least three obstacles stand in the way of adopting workforce analytics.
Technology is not at the top of the list, but the tools still have some maturing to do. Nonetheless, many HRIS vendors have added BI capabilities, and various vendors offer point solutions. Some HR practitioners cited in this article use features of Excel spreadsheets and recommend against buying technology until you know what else you want to do.
A second, bigger issue: Despite massive long-term efforts to install technology for collecting and organizing data, HR departments still often:
Receive inaccurate data.
Get denied data from officials in other departments.
Drown in useless data.
Have data in different systems, such as an HRIS, a recruiting system and a talent management system, that may or may not "talk" to each other.
By persevering, these problems can be overcome.
Developing a Culture of Inquiry
A third challenge to implementing workforce analytics: a three-way, chicken-and-egg problem of HR culture, quantitative skills and business executives’ buy-in. Transforming to a culture of inquiry doesn’t happen without analytical skills, but skills don’t get acquired without fostering a culture that values them. Neither culture nor skills will evolve without buy-in from business executives, who won’t support workforce analytics without first seeing the benefits. HR practitioners must work simultaneously on all three problems.
"Culturally, we have not been data-driven. That has not been the expectation or the skill level," says Conee Biggs, director of global talent management and planning for Medtronic Inc., a Minneapolis-based medical technology company with 38,000 employees. Medtronic’s HR team has begun a transformation. "We have HR folks who look at themselves as business folks first, but we need more HR folks with stronger business acumen."
Valero Energy Corp., based in San Antonio, addressed the need to increase business acumen by paying for HR managers to earn general management certificates in an executive MBA program. "Our head of HR was behind this. Consequently, he ended up with HR leadership that could discuss finance, business value chains and other business issues," says Dan Hilbert, who spent five years as global talent leader at Valero and is now founder and CEO of Orca Eyes Inc. in New Braunfels, Texas, a developer of HCM software, including analytics software.
In data-driven corporations, HR managers are often more comfortable with fact-based decision-making. "I work for a guy who is very results-oriented," says Carl Willis, SPHR, HR vice president for ATK Armament Systems, a Salt Lake City-based business group of Alliant Techsystems Inc., which employs 7,000 workers in 13 plants.
"We use balanced scorecards," says Willis, a 30-year HR veteran with a business degree. "I was told when I came here that ATK was metrics-driven like you’ve never seen. And that’s true. However, we are still struggling with HR metrics, but struggling less, and just getting started with advanced workforce analytics."
Once a culture of inquiry takes root, it can grow like a weed. Wawa Inc. in Wawa, Pa., a food service and convenience company with 578 stores in five mid-Atlantic states, has used workforce analytics for several years. "We have a strong culture of inquiry here," says Ed Iames, senior director of HR. "The more data we produce and send to our business partners, the more questions we get, and the more they want. They become very engaged with what we are doing."
Using scientific methods to test long-held conventional and institution wisdom is one advantage of workforce analytics. Users call it myth busting. Here are some examples.
Experience Didn’t Matter
Managers at Thrivent Financial for Lutherans in Minneapolis thought turnover among new hires within their first year was inversely proportional to the previous experience they had in their disciplines. For instance, customer service employees were less likely to leave in their first year if they had worked in that function elsewhere. "We found the exact opposite," says Sherry Holtz, director of HR effectiveness. "Now we’re looking at why that is. A lot of times you get one answer that leads to the next question."
Predicting Long-Term Success
With many senior actuaries preparing to retire, Metropolitan Life Insurance Co. in New York wanted to know if anything predicts long-term success so it could devise career paths to move junior workers into senior positions. Nick Schaffzin, vice president of business architecture, says the institutional belief was that there was no standard profile for success. "We thought we wouldn’t find a pattern, and we did," he says. "We found employees with different job experiences in the company had better chances of long-term success."
Puncturing the Retirement Bubble
The University of Southern California, Los Angeles, challenged conventional wisdom that its aging workforce would start to retire in droves. It found two significant factors, says Rachel Levy, a USC compensation analyst. First, its non-tenured staff members as a group are too young to start retiring en masse. Second, tenured faculty are much older and approaching retirement age, but they’re not required to retire and most aren’t about to do so. Most tenured professors work into their 70s or longer. So, the fact they’re aging is not a huge issue, she says.
Absolving the Hourly Wage Rate
Managers at Wawa Inc, a Pennsylvania food service and convenience company, suspected that the hourly wage rate was the determining factor in turnover among clerks. Instead, Wawa researchers found that the biggest predictor of turnover was the total number of hours worked in a week—30 being the threshold. Workers are classified as full-time employees when they work 30 hours or more per week. The company moved from a full- and part-time mix of 30 percent and 70 percent, respectively, to a 50-50 split. "As a result, we’ve reduced the in-store turnover rate by 60 percent in the past four years," says Ed Iames, senior director of HR. "Nothing beats an anecdote like firm data."
HR Skills Issues
Analytics skills do not come naturally to HR professionals. In a recent survey of 298 executives in HR, IT and finance, Mark Smith, CEO of Ventana Research Inc. in Pleasanton, Calif., found that "Only half of respondents with HR titles said that they have the requisite skills to develop the analytics and metrics needed to provide workforce information to the organization."
Each HR practitioner interviewed for this article grapples with the problem: As with most technologies, if an HR department simply rolls out self-service software tools to managers, those managers are unlikely to be willing or able to use them.
Medtronic adopted a self-service software tool for nearly 500 HR users. About 10 percent of them use the tool, and about half of those are heavy users, Biggs says. Her team is working on an analytics road map with the goal of improving skills and setting the minimum capability an HR person should have.
Metropolitan Life Insurance Co. in New York, with 54,500 employees worldwide, rolled out software to HR managers, and the reception was half-hearted, says Nick Schaffzin, vice president of business architecture, whose responsibilities include HR systems, metrics and analytics. "Don’t expect the competency in your entire community. Metrics self-service is not the way to go. We’re now building a center of excellence."
Building analytics around a center of excellence (CoE) is a proven strategy. In a survey of more than 200 HR executives worldwide by the Infohrm Group Inc., an Australian provider of workforce planning, reporting and analytics, the 41 percent of respondents who said they had a CoE reported more comfort with and superior results from workforce analytics than others.
There’s no standard blueprint for a CoE. It could simply be a group of expert and frequent users. MetLife’s self-service tool helps Schaffzin identify power users, and he’s now consolidated about six into his CoE.
A CoE needn’t be highly structured. Wawa has an ad hoc group of about eight team members from different functions. Only one employee works full time on analytics, Iames says.
A CoE needn’t be large. Thrivent’s CoE has two staffers, including Holtz, who also has other responsibilities. She has a computer science background and a master’s degree in industrial relations with some statistical training. The full-time person is a certified public accountant who moved from accounting to compensation.
It isn’t necessary—and it’s probably impossible—to rely entirely on HR people. At McLean, Va.-based Freddie Mac, the CoE has grown to five employees, with only one person—the leader—whose primary experience was in human resources. Mike Moss, director of workforce analytics, has a doctorate in organizational psychology, with research methods and statistics training. The CoE includes a technical person who works with the HR information system; the others have skills in statistics, databases and analysis. Some of them had previous HR experience, but their main competence was in data and analysis.
All these practitioners urge doing whatever it takes to acquire employees with analytic skills.
When he was global talent lead at Valero, Hilbert ran the companywide intern program: "I cherry-picked from IT, stats and economics interns" for workforce analytics.
In six years at ATK, Willis has replaced four of five direct reports, each time hiring people with experience in HR process improvement or metrics-based HR.
The third aspect of a workforce analytics regimen is the business executive. Executives understand the value of data analysis for every other function, but not all have thought much about workforce analytics. Without their buy-in, business leaders won’t use the analyses and may block access to business unit databases that HR professionals need to access—customer service or compensation data, for example.
In surveys and interviews, professor John Boudreau of the Center for Effective Organizations at the Marshall School of Business, University of Southern California, Los Angeles, has found that most business leaders’ expectations of HR are traditional: "They’re stuck in the HR service paradigm," he says. "They don’t know what is possible from HCM data analysis. I’m not sure the HR profession understands part of the challenge is to help management understand what is possible with analytics."
Some do: At Valero, for virtually no cost, Hilbert used predictive modeling functions in an Excel spreadsheet to show where the company could expect talent gaps. "When I showed this to the business leaders, it was instantaneously transformational and we were invited to their meetings ever after."
Thrivent’s Holtz learned a similar lesson. "In the beginning, when you ask your business leaders what they want, they may have no idea," she says. "They may not be able to envision what human capital analytics can do for them."
While Holtz doesn’t have a mandate from her CEO, "HR works with his direct reports, and progress has been made."
Randy Stevens recently became head of talent acquisition at Zurich North America (Zurich NA), a Schaumburg, Ill.-based subsidiary of Zurich Financial Services Ltd. in Zurich, Switzerland. With 35,000 employees, Farmers Group Inc. in Los Angeles constitutes Zurich NA’s largest business unit. Stevens recommends starting with a pilot: "I spent some time on how we best demonstrate real ROI [return on investment] to the business. How can we, in terms they understand, evaluate whether our proposed changes will deliver what they need. If you don’t do that, no one wants to waste the time."
Similarly, Freddie Mac has great support from HR executives, but widespread adoption of analytics still takes time, Moss says. He urges HR professionals to be patient and to learn to use data and analysis to tell stories that have meaning for executives. "Some people are still uneasy interpreting numbers. Incrementally, we are moving in the right direction."
Top executives at the German chemical giant BASF SE support workforce analytics. "But, it is like any other organizational change challenge. How ready, willing and able is your HR function? And how ready, willing and able is your business leadership to embrace HR doing business that way?" asks Judy Zagorski, vice president of HR development and strategy for BASF Corp.’s North American headquarters in Florhan Park, N.J. "You have to pick your entry point. What is causing the most pain? Where can you create value through insight and solutions?"
Persistence pays off. At Superior, Lieber identified four labor-intensive business units that would benefit most from his turnover analyses, but he needed access to data only their leaders controlled. He initially won support from two unit leaders—now his two biggest cheerleaders—a third more recently, and he’s working with a fourth. "The sales cycle to break down their resistance has been six to nine months," he says.
The author is contributing editor for technology at HR Magazine.