Machine learning might be able to predict which employees within an organisation are readying themselves to leave the company for whatever reason. Research published in the International Journal of Data Science, explains how employee turnover costs organisations billions of dollars annually. Finding ways to improve employee retention might be guided effectively if there were a way to spot the trends in employee intentions ahead of their making any decision to move to a new position within another organisation, for example.
Owen Hall of the Graziadio School of Business at Pepperdine University in Malibu, California, USA, points out that, as one might expect, engagement, job satisfaction, experience, and compensation are four of the most obvious factors that point to an employee’s decision to leave when any combination of those factors fails to align with that person’s aspirations and expectations with regard to their career and prospects.
Employee retention is a perennial issue for those working in human resource management. This has become even more acute during the COVID-19 where normal life and work practices have been changed beyond recognition in many areas of employment. Increased competition, more customer demands, and intensified recruiting and onboarding challenges, have never been of greater concern, it might be said.
Within HR, the understanding of employee turnover has generally been done in retrospect, perhaps long after specific employees have already moved on. A proactive stance is needed, which is where Hall suggests machine learning might be able to assist. “Machine learning can be used to both identify employees that are planning to leave and design specific implementation amelioration strategies,” writes Hall.
Machine learning can do this with much less bias than might be experienced with human assessment of the situation as it unfolds in terms of employee intentions. Of course, engaging senior leadership is then required to mitigate against the loss of experienced and useful employees to opportunities elsewhere. Hall explains that “The results of a machine learning analysis featuring extreme gradient boost trees and neural nets of a representative employee database yielded classification accuracy levels on the order of 90%.”
Hall, O.P. (2021) ‘Managing employee turnover: machine learning to the rescue’, Int. J. Data Science, Vol. 6, No. 1, pp.57–82.