Predicting Employee Attrition
Explore & understand reasons for employee resigning. Also model and quantified potential measures to manage voluntary attrition
Customer: Global Consulting Company
Challenge: Highly specialized human capital leads to high (opportunity) cost of attrition
Objective: Identify actionable recommendations at HR policy and leadership level
Solution: Mash up data from different HR systems with ERP data and run machine learning models. Result can be modelled and simulated in Peacock app including cost of attrition
What were the key challenges?
Depending on the role of the employee, the opportunity cost of an employee leaving its ranges from 33% to 250 of their annual salary.
While other risk are more easily quantified and therefore most closely managed, HR is usually struggeling to quantifiy the cost of attrition as well as quantifiying the impact of concrete measures
If drivers for attrition can be identified and counter actions can be developed, decision-makers in HR could, often for the first time, show the ROI of certain changes.
How does our solution differ?
Within the app, we allow for visual exploration of the variable to develop a first feel for the data
Our application then runs different machine learning models to determine beste model quality to predict attrition. Analytical user are able to set different parameter and check model accuracy.
We present the results for business user in different formats and configuration options. This allows the user to explore the results according to his/her interest and specific question.
“it was a great experience developing this Advanced Analytics Use Case together with Predict42. The team has developed a model that will enable us to reduce involuntary staff turnover. The outcomes were partly expected and partly surprising. I think this mix is what makes an Analytics project great.”