Federal agencies, long accustomed to the stability of a personnel system that hired people in their 20s and kept them until their retirements, are now facing staffing upheaval on dual fronts. Agencies are in the midst of record-breaking attrition from Baby Boomer retirements. At the same time, they are under White House orders to make major staffing changes — mostly reductions, but also some significant increases — while changing long-standing processes and systems.
On the surface, it may seem like mass retirements are coming at the right time as most agencies look to downsize. But, as leaders know, successful workforce management is not just a numbers game of indiscriminate downsizing. Agencies must have the right people in the right jobs at the right time. That can only come from insights such as who is leaving and why, how robust is the applicant pool for replacing them, and what will cause the best employees to stay.
Human resource managers have the information to guide these changes. The problem is, much of it is buried in data stores too massive for people to glean valuable information. That’s where data analytics comes in.
Using data more effectively
Many large organizations are finding that data analytics is the answer to workforce management. The ability of customized software to sift through thousands of terabytes and pull information to solve specific problems makes data a strategic asset, rather than a storage burden.
Data sources — such as an agency’s HR system, time and attendance systems, employee satisfaction surveys, and performance or financial systems — can be integrated and used as part of a centralized analytics dashboard. Then, algorithms are created to pull as much or as little data as the agency needs to address the challenge at hand. HR managers may want to pull analytics from exit surveys — is there a pattern of complaints in one department or against one manager? Or, they may study the trends of when people take leave — will they be short-staffed during certain times of the year?
For agencies facing large staff reduction requirements, it would be helpful to predict which employees are interested in leaving and which departments they work in. Age is one way to predict attrition; others are annual reviews, job classifications and satisfaction surveys.
There are dozens of variables that can be used in workforce management analytics. They include occupational categories, employee grade and step, salary locality adjustment, employee duty location, employee start and end dates, work schedule type, awards (number, amount, frequency) and tenure … to name a few.
With this information, agencies can gain insights into whether employees with high-valued skills, such as cyber, are in danger of leaving. They might also learn if there is a certain location that is at risk of a sudden flood of retirements. How many people may retire in the next six, 12 or 24 months?
For agencies under orders to back fill or grow quickly, HR managers should understand the demands of the agency and the supply of available talent to fill the positions. Using data analytics, the agency can quickly pull information from thousands of resumes. For example, they can create an algorithm that flags all the resumes that meet eight out of 10 skills requirements for open positions and put those resumes at the top of the stack.
The benefits of intelligent data
As agencies gather this workforce intelligence, they can then take actions to manage outcomes. Retention is a major issue. The cost to an organization to replace an employee is six to nine months of that employee’s salary, according to the Society for Human Resource Management. To improve retention, agencies may offer incentives, such as cash awards, for workers to stay. Again, analytics can provide insights by showing past experiences of how many stayed previously after receiving a cash award of a certain amount. Will people in hard-to-fill positions stay for a $1,000 bonus? Maybe $800 would be enough to retain them? Or, perhaps giving them a telework day (virtually free to the taxpayer) would make them more likely to stay.
Analytics also can correct popular misconceptions. When the military’s Special Operations Command leaders wanted to better understand when most people dropped out of training and why, most believed physical skills (i.e., number of push-ups, obstacle course time, etc.) were the best predictor of who would drop out. However, the data showed it was the soft skills (leadership, team skills, etc.) that were better predictors. The analytics also showed a surprising finding that recruits from Alaska were more likely to drop out than their peers. The analytics showed it was because the training was done in North Carolina in August — a climate too hot for most Alaskans to adapt to. This is just one example of a workforce-related program that is experiencing a true benefit from analytics.
Data analytics is both art and science — from preparing the data to creating algorithms to analyzing the information. When done effectively, analytics can save time and money, improve efficiency and allow for better decision-making. Most of all, it can help agencies manage and predict the major changes that lie ahead for the federal workforce.
Jon Lemon is a strategic solutions specialist in the government unit of SAS, a provider of advanced analytics. Prior to SAS, Mr. Lemon was a budget director at the Department of Homeland Security.