Leave Risk Prediction
Covered in this guide
- Using the Leave Risk Prediction
- How we calculate leave risk for employees
- Using overall ratings to decide whether to take action
From the Overview page, you can click the "Leave risk prediction" tab, or click "Open report."
Using the Leave Risk Prediction
On the Leave Risk Prediction tab, the top chart gives you an overview of your entire organization, showing the distribution of your current employees that have been identified as High, Medium, and Low Risk. You can toggle to show or hide “unknowns.” This means Small Improvements won't show employees whose leave risk we can't predict.
Click on a segment in the chart to filter the table to show which employees belong to that Risk Group.
“Last Rating“ is the manager’s overall rating the employee received in the most recent Performance Review (if you use the “Overall Rating” question type). This can tell you if the at-risk employee is a high or low performer.
You can also see the leave risk prediction for each department:
How we calculate leave risk for employees
We use a proprietary machine-learning algorithm to calculate the "risk of leaving" score. The algorithm reviews a variety of data on performance, feedback, and turnover, and identifies patterns that it uses to predict future turnover. If we don't have enough of this data to predict the leave risk with a certain level of confidence, then we will not assign a risk level to that employee. As we receive more data, the algorithm improves itself and becomes even more accurate.
No algorithm can predict turnover with 100% confidence. This metric is designed to help your team understand which employees could be planning to leave so you can act in time to try to prevent regrettable turnover.
We calculate three categories of risk leave:
- Low Risk = between 0% and 49% Risk
- Medium Risk = between 50% and 74% Risk
- High Risk = between 75% and 100% Risk