In a blog that was posted at the beginning of 2017, the author reviewed some cool research that has been conducted surrounding the prediction of employee turnover. To elaborate, the researchers employed predictive analytics using machine learning to predict who would turn over and who would stay within a large organization.
They incorporated a wide variety of turnover drivers which included:
Number of Projects
Average Monthly Hours
Tenure with the Organization
Overview of Findings
They found that certain drivers were more likely to lead to turnover than others. Specifically, the researchers found that working in a certain department (what they called job function) was related to turnover, as certain departments had higher satisfaction and lower turnover than others.
Additionally, individuals who had low and high scores on their last performance evaluation were more likely to leave than those who scored closer to the middle of the distribution. Therefore, both the top and bottom performers are leaving at a higher rate than average performers.
Interestingly, the same pattern of results was observed for employee satisfaction: those who fell toward the low and high ends of satisfaction had higher turnover rates than those toward the middle.
Finally, not surprisingly those with the highest salaries were most likely to stay while those with lowest salaries were most likely to leave. Overall, the researchers concluded that satisfaction, the number of projects, and time with the company were the most robust predictors of turnover.
By combining these various drivers through machine learning the researchers were able to identify stayers and leavers with great accuracy. This is pretty impressive! However, simply identifying drivers does not fix the turnover issue, and the researchers left out a couple of core sets of turnover drivers.
3 Drivers of Turnover
At Select International, we categorize drivers of turnover into three primary buckets: Individual, Internal, and External. Each of these is described below:
Individual Drivers – traits employees possess that make them more or less likely to turnover. Examples include impulsivity, adaptability, and fit with the organization and/or work environment.
Internal Drivers – factors within the organization that lead to thoughts of quitting and turnover behaviors. This set of drivers is the largest and includes pay, benefits, opportunities for growth and promotion, leadership, coworkers, job design, and culture.
External Drivers – factors that are outside of the organization, but still impact the organization’s turnover rate. Examples include the unemployment rate, competition, reputation of the organization, available applicant pool, and location.Related: Record Low Unemployment, How will this Affect Your Hiring Strategy?
Consider All Types of Drivers When Evaluating Turnover
The researchers who conducted the previously discussed study focused on Internal drivers of turnover. While this category contains the majority of potential turnover drivers, focusing on only one source of turnover drivers limits the potential prediction of the model. Even more impressive results may have been obtained had the researchers investigated Individual and External drivers of turnover, as well.
For example, perhaps this particular organization is losing their high-quality talent because these individuals are simply not a great fit with the organization or culture. Or, perhaps this organization is constantly improving through change and certain individuals are not as adaptable, and it manifests in dissatisfaction. Alternatively, given that even the high performers have elevated turnover rates, it is likely that this organization is not an employer of choice, and as such, top talent is leaving for organizations that are employers of choice. To really understand what is driving turnover within an organization you need to examine turnover from all angles.
More on being an Employer of Choice
Data on the missing Individual drivers could be obtained through assessment scores. Individuals would complete an assessment (or set of assessments) that measures certain traits of individuals, along with how well their values and goals align with those of the organization. Combining the results of these measures can provide an overall picture of the individual’s personality and potential fit with the job or organization.
Data on the missing External drivers could be found by conducting a market analysis to see how this organization stacks up against their competition. You can use resources like Glassdoor to explore competing organizations’ salary and benefit packages and to obtain insights into their culture and leadership, for example. Incorporating some of these individual and external drivers into the algorithm along with the internal drivers already investigated would likely significantly contribute to the prediction of turnover. Once all the primary drivers are identified, resources can be dedicated to targeting those primary drivers, and as many of them as possible. This is when you will start to have a real impact on turnover.
In summary, focusing on only one set of turnover drivers likely will not allow you to reach your ultimate turnover goals. However, taking a detailed investigation into all three sources of turnover will have the most profound impact, and we know from our experience that single digit turnover is possible with the right investigation and intervention.