"This is data that used to take days or weeks to assemble," says the affable 40-year-old public servant. Another click of the mouse and he’s drilling down into the data by county and then by job position. Seconds later, he’s able to slice through the data to view individual workers. "We’re able to determine how pay scales vary in different states and uncover patterns that influence them, such as education, training, gender, and race. This is like lining up a Rubik’s Cube."

    The high-tech operation could never be mistaken for grandpa’s federal government. The Census Bureau--along with a growing number of companies such as Interwoven, Royal Caribbean Cruises, Merck, and Northrop Grumman Information Technology--is turning to sophisticated workforce analytics applications to mine data and gain insight into business processes and labor conditions. These applications can spot hidden patterns and problems and help an organization make more effective business decisions surrounding programs, policies, and pay. Yet, while workforce analytics uses data-mining and reporting tools that weren’t available just a few years ago, the system isn’t problem free.

    Unless an organization taps into the right data and ensures that it is accurate--and then puts the data into action--workforce analytics can become a trip down the rabbit hole. Ron Hanscome, senior program director at META Group in Plymouth, Minnesota, is one who can see the distinct advantages. "It turns the intangible into tangible," he says. META reports that about 60 percent of Global 2000 companies will use some form of workforce analytics by the end of 2003, mostly to improve decision-making on compensation, recruiting, retention, and performance issues. David Ulrich, a business professor at the University of Michigan, says that the technology is bringing healthy change. "Organizations are beginning to recognize that the softer side of the business can be measured and that it’s a key factor in achieving success," he says. Betty Silver, human capital strategist at SAS Institute Inc. in Cary, North Carolina, adds, "Workforce analytics allows organizations to examine a multitude of interrelated factors. It can drill into places that previously couldn’t be reached."

    Although workforce analytics is rooted in human resources, the ability to deliver detailed information about workforce issues and trends--ranging from where to locate a factory for the best labor pool to how a VP of sales should build incentives--has real applications throughout the organization. Done properly, workforce analytics can help an enterprise sail beyond the flat earth of spreadsheets and simple reporting and discover a brave new world.

    Simply put, workforce analytics puts line managers on the front lines of decision-making. A VP of operations can tweak compensation, rewards, and responsibilities to keep top performers engaged. If he or she adds a bonus, it’s possible to measure the corresponding results before and after. A finance executive can slice and dice data about labor trends and costs to create a favorable pay structure at a new factory. At the same time, a director of human resources can identify which benefits employees use, which are frivolous, and how to maximize corporate resources.

    META Group estimates that the initial cost of workforce analytics starts in the low six figures and averages about $250,000 per 10,000 employees, plus ongoing annual licensing and maintenance fees. In some cases, that can translate into an ROI ranging from a few months to a year or more. And a well-designed system can integrate with a human resources scorecard and incorporate industry metrics to provide a more complete picture of cause-and-effect relationships. That could mean taking industry averages for the cost of a particular benefit and overlaying them with internal costs, or examining how incentives at best-practice companies affect productivity and comparing it to internal practices.

    At the Census Bureau, the genesis of workforce analytics is rooted in such knowledge. One day, while Pejsa was sitting with his boss discussing human resources systems, the two realized how inefficient a homegrown application was for managing the 10,000 employees spread across 12 regional offices. "The system was inflexible, and we had questions about the data quality," he says. What’s more, the system had a tendency to crash and couldn’t accommodate all of the data residing in various databases and datamarts. So, they began to look at the various analytics and reporting tools on the market.

    Eventually, the Census Bureau chose a workforce analytics application from SAS. The selling points, Pejsa says, were the simplicity of the Web interface, built-in templates, and the flexibility to add and create reports as needed. Moreover, Human Capital Management allowed the Census Bureau to plug in a diverse array of formats, including Microsoft Access and Excel.

    After a six-month implementation, the Census Bureau went live with the system in the spring of 2002. In September, it moved to a Web-based version of Human Capital Management, and put the application on the desktops of 25 analysts within human resources (Pejsa hopes to expand the use of the system to 200 to 300 managers within the next year or so). Finally, the government agency was ready to put its human resources data to work in a big way. "We had access to data and information that would significantly change the way we make decisions," Pejsa says.

    That may be so. But like any other emerging technology, workforce analytics can present a hornet’s nest of challenges, frustrations, and problems. "Workforce analytics is only as good as the underlying data and how it’s used," META Group’s Hanscome cautions. One of the biggest problems is identifying business objectives and mapping data so that it can flow seamlessly across departments, divisions, and business units. Adds Marcia Barkley, president of MBarkley Consulting, a Sacramento, California, firm: "Building the infrastructure for workforce analytics can prove intimidating. Many organizations have purchased applications and wound up with them sitting on the shelf or sitting idle."

    There’s also the task of changing a culture that once allowed only the top executives to view data but now pushes decision-making further down into the organization. "If line managers and others don’t understand the value of a workforce analytics application and how to collaborate and share data, then the system isn’t going to produce anything close to the desired results," Barkley explains.

    At the Census Bureau, the learning curve has been steeper than expected, Pejsa admits. When the agency first turned to workforce analytics a couple of years ago, it had plans to use an assortment of features and tools that have since been placed on the back burner. What’s more, it wanted to introduce the application to dozens of senior-level managers, but most were woefully unprepared for the hands-on approach. "Our goals were too ambitious," he says.

    Another giant headache was revamping the agency’s organizational structure to fit the application. Over the years, the Census Bureau had assumed a decidedly horizontal structure, and the SAS software required a distinctly hierarchical model to function effectively. Pejsa spent more than a month working with his boss and staff analysts to map all the desired tools, capabilities, and permission rights to the appropriate users.

    Finally, the process of connecting all the various systems proved more than a bit troublesome. At first, other applications and systems produced hiccups when the Census Bureau attempted to switch on the workforce analytics tools. And in some cases, servers crashed and nerves frayed when the agency attempted to plug in back-end applications and stream data across the organization.

    The Census Bureau’s encounter with workforce analytics has been more successful than most. Only about 10 percent of organizations are now using it in any significant way, and another 10 to 15 percent are dabbling in it, says META Group’s Hanscome. Part of the problem, he says, is that some vendors’ applications aren’t ready for prime time. They simply cannot cull and organize data in ways that provide maximum insight. Another factor is that tight IT budgets and a heavy emphasis on other projects, including customer-relationship management and supply-chain management, have diverted funding and attention away from the somewhat nebulous world of workforce analytics.

    Yet, for the Census Bureau, which spent less than $100,000 to get the project off the ground, the payoff has been substantial. Despite occasional glitches and breakdowns, the agency is evolving into a smarter, more flexible organization that’s finally able to measure the cost and value of its workforce. And as the agency has plowed through various problems and challenges, it has found a clear path to success, Pejsa says. "We now have the tools to make strategic decisions faster and better.