High rates of employee turnover pose a danger to any business and cost employers time and money to overcome. Employees leave their jobs for a variety of reasons. What if there was a method to determine exactly what your employees need to feel job satisfaction?
Enter predictive analytics, a branch of mathematics devoted to crunching behavioral and environmental data to uncover insights and trends. In the past, reasons for employee attrition have been compiled based on anecdotal evidence, but
Matthew Stevenson, partner and co-leader of
Mercer’s Workforce Strategy & Analytics group, says that’s a troublesome approach. “We find that statistical data is much more effective and actionable than perceptional analysis, because there's all sorts of bias in asking people questions,” he says.
Here are three ways you can take advantage of existing data to mitigate attrition and achieve employee engagement.
Make Decisions about Employee Engagement Based on Data, Not Speculation
Anecdotal data can result in assigning blame for attrition to one individual factor. For example, many employees cite their manager as their reason for leaving a job. But as Stevenson explains, “management often reflects systemic policies, so anecdotal evidence makes teasing out the difference between an individual and the business as a whole very difficult.”
Dr. Paul Fairlie, work psychologist and CEO of
Heliosophy--a consulting service that applies psychology to solve issues with employee dynamics--agrees. “There’s a lot of folk wisdom around why people leave employers, and it's not very evidence-based,” he says. “But analytics can counter human biases. Analytics go beyond opinions and hearsay when identifying real turnover drivers.”
Provide Better Data to Fuel Retention
Until recently, the data being produced by business analytics focused on just that — business. It’s important to remember that employees are human, and subject to human behaviors. Fairlie notes that behavioral analysis has not traditionally been applied to business, but that it constitutes a missing link in the data. “Businesses need to merge variables from different sources to find combinations that predict past turnover among high-performers,” he says.
However, Stevenson cautions that data isn’t the only key to understanding attrition. “Predictive analytics relies on data, but also on questions. Prediction does not equal causation,” he says. “It’s important to learn what types of questions we need to be asking about what the data can teach us.”
Help Create Organizational Change
It’s important to remember that the data derived from predictive analytics is intended to make changes company-wide, not just to individual behaviors. Employee engagement is often about organizational culture and systemic issues. Predictive analysis can help identify those bigger-picture issues that are driving employee attrition or engagement.
“Satisfaction, commitment, discretionary effort, stress, burnout — the list goes on,” Fairlie says. “Research is full of things known to drive attrition, and organizations are likely harboring many other drivers that will only be discovered with data mining.”
Stevenson notes that once the data is compiled, it must be applied to the organization as a whole. “Predictors should be used in an actionable way to make changes to the system, rather than changes to individuals,” he says.
