Rothausen and Henderson (2019) suggests that the cost of replacing and training new employees could be less than the cost effects of how current employees are treated and former employees were treated. Employees that remain with the organization, but strongly desire to leave because they are frustrated with how they are treated, could be more damaging as a factor than actual turnover. This leads to the need to seriously consider interviewing employees while they are still at the organization, before the need for an exit interview. Understanding what happened at various times before turnover occurs can allow HR to evaluate and change the work environment before intentions to lead become actual turnover behavior.
Predictive Modeling with "Stay" Interviews
This is where predictive modeling comes into play. Focusing predictions steps before turnover, rather than attempting to predict turnover alone, could provide organizations with a decision-making advantage through additional data in their turnover modeling. If these stay interviews provide data on how employees are treated as an antecedent to, or along with desires to leave the company, resources could be rearranged to provide intervention.
We've all heard the data on what it costs to replace an employee: $15,000, half a year salary, two times salary, and so on, depending on the data you're viewing. We can agree on one thing: the financial costs are significant as are the cost of other resources to replace and train employees. Slowly shifting resources to turnover intention interventions (conducting stay interviews and putting the data to use), could very well lighten the load for recruiting and all other functions related to hiring. In a fairly short time, depending on the prioritizing and effectiveness of such interventions, completing this shift could result in increased retention, increased performance and stronger culture. This can occur as a result of employees feeling heard and gain organizational commitment through the psychological force of reciprocity once something is done with the data they provide.
Many methods exist to predict turnover. Speer, Dutta, Chen, and Trussell (2019), provide some common variables for attrition modeling which include individual differences, job alternatives, organizational context factors, embeddedness indicators, direct withdraw indicators, and social characteristics.
These same variables can be used to predict turnover intentions (i.e. I really want to leave) for early intervention. In collecting this data, there are a few things of note:
1. Trust plays a large role in the quality of the data. The differences between confidentiality and anonymity must be understood and employees must be informed of which one is guaranteed, and how it will be guaranteed. Employees will want to know what is going to be done with the data and will need some assurance that the data will actually be used to better the workplace environment.
2. This also might require some courageous conversations; managers asked to identify potential flight risks might be hesitant to do so, if they believe they may be part of the problem. This can be a tough challenge but it must be properly managed. Also, gaining an understanding of what is working under the umbrella of a particular department might shed some light on why those people perform better and stay. Some stay either because of the department's alignment with the larger organization, or in spite of the environment in the larger organization.
Reducing turnover saves on hiring replacements and getting them up to speed. This is so much more than that. Rothausen and Henderson (2019) explore the accounted for costs that former employees have on company productivity and financial performance, and the morale of existing employees. Undoubtedly, some employees will leave your organization, even when they are treated well. They argue that "the result of the number of former employees a company has and how they feel when they leave, as well as how current employees feel for months or years before they leave” must be involved in the calculating the cost of turnover. They add that while companies can benefit not just from turnover by replacing poor performers, they also benefit from having people feel good about how they were treated for the duration, and what the company stands for.
• The same variables for attrition modeling can be used for predicting turnover intentions.
• The effects of poorly treated employees (both who stay and those who leave) can cost more than turnover itself.
• Slowly shifting resources from recruiting, to identifying and preventing turnover intention can pay off in a number of ways.
Rothausen, T., & Henderson, K. (2019). Two messages from the other side of the turnover coin: “Here to stay or go?” and “Should I stay or should I go?”. Industrial and Organizational Psychology, 12(3), 306-309. doi:10.1017/iop.2019.59
Speer, A., Dutta, S., Chen, M., & Trussell, G. (2019). Here to stay or go? Connecting turnover research to applied attrition modeling. Industrial and Organizational Psychology, 12(3), 277-301. doi:10.1017/iop.2019.22