|
|
Reducing Employee Turnover with Predictive Analytics

Reducing Employee Turnover with Predictive Analytics

Employee turnover banner

For companies, reducing employee turnover is not just about wellbeing – it also raises financial questions. When an employee leaves a company, it can cost around 33% of their salary. So, when turnover rates are high, costs skyrocket. This translates to mass recruitment if a large number of employees leave, or operational inefficiency when highly-paid employees leave. Either way,  reducing employee turnover is key for companies… and Predictive Analytics can help!

Large retailers, health and social care establishments are among the most affected by high employee turnover, impacting their costs significantly. In France, while the average turnover per company is around 15% (lower than the global average of 23%), it exceeds 35% in these sectors. 

So, players in these industries must improve employee retention to limit costs and serve their customers better. We recently helped a French leader in childcare to predict which employees would leave and why, six months in advance. To do so, we only used data they are legally required to collect. 

Harnessing Data that is already available to reduce employee turnover

Our customer provides childcare services out of hundreds of branches. They have a few thousand employees, mostly in low-qualified jobs. For years, they have been facing an employee turnover above 35%. This results in increased costs and a lack of operational efficiency. Finding replacements takes time, energy, and money. Meanwhile, remaining employees find themselves adapting to ever-changing working conditions and covering leavers’ tasks and responsibilities, thereby upending their usual ways of working. This can create frustration, unease, and tension, which, in turn, leads to even more departures.

Hence, our client was looking not only to anticipate possible departures a few months in advance, but also understand leavers’ motivations in order to try to retain them. We used existing HRIS data from the previous five years to build predictive models with ZGP:

  • Contract details
  • Salary information
  • Absenteeism data

The purpose was to build a binary classification model able to identify employees presenting a high resignation risk and potential reasons that would cause them to quit. Based on past data, the model we created was right three times out of four (a recall of 75%) when predicting who would leave the company within six months.

Adding Context to Employee Turnover

MyDataModels’ Decision Intelligence Platform has other perks for HR and management. As requested, it allows them to identify the reasons why employees might leave. This is the first step towards them understanding the underlying causes and taking informed decisions to retain their employees.

Furthermore, we were able to adapt models for each branch and each area. It allowed us to tailor the model to specific local conditions. It proved itself very useful in adapting HR policies accordingly. For example, the client used MyDataModels’ “what-if” features to try different scenarios and visualize the impact of their decisions in maximizing employee retention. They were able to test constrained scenarios: if we can’t change local management, what would be the best leverage to reduce employee turnover in this specific branch?

Here’s what our Decision Intelligence Platform for HR and management looks like:

QEHR-screenshot

Reduced costs for maximized operational efficiency

We identified with a high level of probability three quarters of the 35% of employees who eventually resigned within six months. Our customer is now using our Decision Intelligence Platform to adapt its HR policies and tackle its employee turnover problem. The first results should be visible by the end of the year.

In foresight, we can simulate the impact that MyDataModels’ Decision Intelligence Platform will have on our customer. The goal is to reduce the turnover rate by 25% on a population of 2,500 employees with an average gross salary of €30,000. Not only will this impact the financial performance of the company, but it will also limit avoidable turnover and help their employees gain in seniority.

MyDataModels impact on turnover costs

Employee turnover is a natural process in a company’s life cycle. People come and go, as they develop new aspirations and projects. However, when turnover rates climb too high, companies must act and understand the underlying causes. Decision Intelligence Platforms, such as MyDataModels’, can detect hidden patterns, identify reasons for leaving and help decision makers improve their employees’ wellbeing and engagement, while saving a lot of money in the process. If you can relate to that, feel free to reach out to our team!