Numbers Game: Companies Utilize Data to Predict Workforce Needs

Employers are beginning to turn their gaze from the rearview mirror as they develop and analyze workforce metrics.

For the past decade, large employers have tried to get a better handle on basic data, including head count and turnover. Some have taken those initiatives a step or two further by calculating such metrics as the attrition rate of high-performers or the amount of revenue and profit per employee. But generally, workforce metrics have tended to focus on past activity, with limited attempts to peer into the future. Now, that’s slowly starting to change. With the help of software tools and expert statisticians, organizations are beginning to project their labor needs and challenges with greater precision. Already, predictive workforce analysis is standard operating procedure at some companies, such as weapons-maker Alliant Techsystems Inc., better known as ATK.

The company began a metrics push in 2008, aiming to improve the recruiting function of one division. Gradually, the effort morphed into a broader workforce planning project. The company, which makes aerospace, defense and commercial ammunition products, continues to work to more accurately anticipate its labor supply and demand.

Based on an analysis of past departures, ATK officials created a “flight-risk model” that calculates the probability of attrition for each employee. Last year, for example, ATK correctly projected unusually high turnover in a vital plant maintenance group. 

ATK will likely have more company soon. Jac Fitz-enz, head of the consulting firm Human Capital Source and a pioneer in workforce metrics, expects more organizations to take advantage of their human resources data to forecast what’s ahead. With better predictions of expected attrition among key performers, for example, firms can take action to try to prevent or prepare for the losses. For instance, employers can develop targeted recognition programs to try to retain flight risks or they can prepare for departures through better succession planning, training and recruiting campaigns. “Those who are into predictive metrics have a competitive advantage in the war for talent,” Fitz-enz says. “They know more about hiring and developing people for business opportunities, as well as how to contain attrition of high-potential talent.”

To be sure, the number of firms jumping into predictive workforce analytics remains small. In fact, only 20 to 25 percent of companies have a workforce metrics system of substance, Fitz-enz estimates.

A variety of obstacles have slowed the adoption of workforce data analysis. For one thing, companies have tended to prioritize data related to sales and production, which can more easily be tied to the bottom line, says Lois Melbourne, CEO of Aquire Inc., a software firm in Irving, Texas, that sells tools for workforce analysis and planning. Another hurdle is the plethora of HR systems many companies use, each with its own set of data. Multiple systems can complicate big-picture analysis because they can be difficult to integrate and may contain contradictory data. What’s more, Melbourne says, many people drawn to the HR field tend to be uncomfortable with number crunching. Still, the tide might be about to turn. Melbourne sees companies studying their workforce information more carefully and notices growing interest in prediction. Employers “know they need to get there,” she says.

Employers want to take a longer view of their labor needs now partly because of the shortsighted way many managed their manpower during the recent recession, says Jason Averbook, CEO of the consulting firm Knowledge Infusion in Minneapolis. Averbook estimates that 30 to 40 percent of the employees laid off by firms during the downturn were the wrong choices. “Organizations for the most part did a slash-and-burn of talent,” Averbook says. “Most HR leaders never want to do that again.”

Hunger for more data-based decision-making prompted a workforce metrics project at Nationwide Mutual Insurance Co. Launched in 2009, the effort has resulted in monthly “score cards” for managers that contain a number of metrics, such as head count, turnover of high performers and span of control—which shows the average number of direct reports per manager. The score cards also display organizational goals for each of the measurements and managers’ performance relative to those goals.

The reports generally amount to rearview-mirror workforce metrics. But the Columbus, Ohio-based insurer is shifting its gaze forward, as well. The company began its metrics with Excel spreadsheets but is upgrading to an analytics tool from software provider Oracle Corp., which will speed up creation of the score cards. “That frees us up to focus on that next step,” says Scott Nemeth, senior consultant for human capital analytics at Nationwide. “We will use historical data together with market intelligence and benchmarking data to provide better workforce forecasting.”

Software is crucial to a cornerstone of predictive analytics: mining the valuable data and insights that can be buried in a mound of information. Computers act as the drilling equipment to discover nuggets difficult to find by hand.

That computer gear, in turn, requires skilled operators who are well-versed in data analysis. ATK, for example, hired a statistician as part of its initiative, and Nationwide brought on Nemeth to beef up its analytical and planning expertise. Before arriving at Nationwide, Nemeth spent five years in General Motors Corp.’s corporate strategy and planning department and another five in GM’s global workforce planning unit.

But powerful software and data analysts alone won’t result in effective forecasting and planning. Companies need to set goals and ask the right questions. A cutting-edge computer system for analyzing data is no more than a metal detector, Averbook says. Companies must know where to sweep to find useful insights. “You are walking around on a golf course with a metal detector looking for needles,” he says. “In every business, there are different needles.”

Some HR experts worry that companies could focus too much on hard numbers and neglect the softer side of talent management. University of Michigan management professor Dave Ulrich says companies shouldn’t obsess over measurement to the exclusion of inspiring employees and helping them feel part of a team. “Turning work into a series of numbers and analytics may take away the emotion that brings meaning to our lives,” he says. “When people find meaning at work, they are more productive, which leads to companies being more profitable.”

Fitz-enz and other analytics advocates say organizations usually err in the other direction, making less-than-optimal decisions about talent because of a lack of concrete evidence. In any event, Fitz-enz says fears of overemphasizing metrics are overblown. HR professionals, he says, tend to blend hard calculations with a soft heart.

Workforce Management, March 2011, pgs. 20-21 — Subscribe Now!