If we could reliably predict which employees want to leave, we might react more flexibly and thus save a lot of money. But how to do this? The best way would be somehow to automate the entire process. Let's learn to use our employee data better and computers will warn us of impending departures. The Wall Street Journal notes that this is how increasingly more companies are thinking about workforce planning.
Companies like Wal-Mart or Credit Suisse have already started testing algorithms for the detection of employees who may soon leave. They analyse data covering a wide range of factors with a potential impact on the employees' decision to leave. The data includes job tenures, geographic locations, results of performance evaluations, employee surveys and the like.
Analysis of this data reveals what motivates employees to stay or leave. However, it is not as simple as it might seem since there is no single metric that would clearly predict an employee is going to leave. It is necessary to seek the right combination of risk factors, create different models and test them over a longer period of time.
Timely promotions may prevent departures
During the past three years, the Credit Suisse bank has focused on the analysis of everything that happened with employees (promotions, salary increases, but also significant changes in personal lives), in an attempt to predict whether they would leave the following year. The result was the launch of an internal recruitment programme, whose objective was to activate employees' interest in applying for new jobs. Thanks to this programme, more than 300 employees, who would otherwise probably have left, were promoted.
The company wants to continue testing the predictive algorithm by focusing, for example, on the differences in the departures of men and women from different positions. William Wolf, Credit Suisse's global head of talent acquisition and development, told the Wall Street Journal that the reduction in unwanted employee turnover by one percentage point could save the bank $100 million annually.
Wal-Mart has decided to follow a similar path. The supermarket chain wants better to predict when employees may want to leave a particular position but not necessarily the entire company. It is trying to speed up the prediction of when individual employees will be promoted in order to ensure new people will enter the vacant lower positions quickly enough. The company promotes up to 170 000 employees a year.
As summarised by lpida Ormanidou, Wal-Mart's vice president of global people analytics: “If we can tell three months in advance [that a position is going to be open], we can start hiring and training people."
How do you use people analytics in your company?
-kk-