🚇 On this page:
About the risk of exit insight and how it’s calculated.
Related pages:
For a general overview on how People Analytics generates insights, see Insights 101
Overview
Turnover is a normal part of any organization. However, when high-performing employees or those in critical roles leave, it can significantly impact the business.
Our Exit Risk Insight feature helps you identify potential turnover before it becomes a problem by using signals from your employee engagement surveys and connecting to actual turnover data in your HRIS- thereby allowing leaders to intervene proactively.
How it Works
For customers already leveraging our Engage and People Analytics module, Exit Risk Insight seamlessly integrates with your existing data.
🧠 Science Behind
We've found across our employee data lake of over 4 million employees and approximately 6,000 companies that a simple question within your Engage surveys—"I intend to work at this company for more than 2 years" (which we call "Intent to Stay")—is a powerful predictor of actual employee attrition. We’ve consistently found a strong relationship between responses to this question and whether someone ultimately leaves the company.
Key Features
At-Risk Employee Headline: See how many employees in your company are at risk of leaving based on their "Intent to Stay" responses.
Response Distribution: View a breakdown of positive, neutral, and negative responses to the "Intent to Stay" question from your most recent survey.
Historical Comparison & Odds Ratio: Compare "Intent to Stay" responses between employees who left and those who stayed. This includes an odds ratio calculation to show if this question is a leading indicator of turnover in your organization.
Demographic Exploration: A table that allows you to explore different demographic groups and their current risk of exit based on the latest survey results.
Risk of Exit Insight Basics
Prerequisites
To unlock the Exit Risk Insight, certain data and connectivity requirements must be met:
Active Engage Surveys with "Intent to Stay" Question: You must be actively conducting employee engagement surveys through our Engage platform, and these surveys need to include the specific question: "I intend to work at this company for more than 2 years." This question forms the core data for our risk analysis.
For the Risk of Exit Group Table: This feature will be available if you have a recent survey that includes a sufficient number of responses to the "Intent to Stay" question.
For the Company-Wide Exit Risk Insight (including the Odds Ratio): This comprehensive insight becomes available only if:
Your HRIS source is connected to the People Analytics platform, enabling the capture of actual turnover data.
Sufficient historical turnover events have been accumulated within your People Analytics data.
A sufficient number of employees who subsequently left had also responded to the "Intent to Stay" question on a previous survey, allowing us to establish a predictive link.
How we calculate odds ratio for the company overall
The Odds Ratio helps us understand the relationship between how employees feel about staying and whether they actually leave the company. It's a powerful indicator of whether the "Intent to Stay" question is a leading predictor of turnover in your organization.
Here's how we calculate it
Identify Key Groups: We focus on two specific groups of employees based on their response to the "I intend to work at this company for more than 2 years" question from prior surveys:
Negative Sentiment Group: Employees who indicated they do not intend to work at the company for more than 2 years.
Positive Sentiment Group: Employees who indicated they do intend to work at the company for more than 2 years.
Note: Neutral responses are excluded from this specific odds ratio calculation to provide a clearer contrast.
Calculate "Odds of Leaving" for Each Group: For each group (negative and positive sentiment), we determine their "odds of leaving." This is calculated by dividing the number of employees from that group who actually left the company by the number of employees from that group who stayed.
Example: If 10 employees with negative sentiment left and 90 stayed, their odds of leaving are 10÷90=0.11.
Compute the Odds Ratio: The Company-Wide Odds Ratio is then calculated by comparing these two figures:
Odds Ratio=Odds of leaving (Negative Sentiment Group)/Odds of leaving (Positive Sentiment Group)
What the odds ratio means
Odds Ratio > 1.0: This indicates that employees with negative "Intent to Stay" sentiment are more likely to leave the company compared to those with positive sentiment. For example, an odds ratio of 2.0 means the odds of leaving are twice as high for the negative group.
Odds Ratio < 1.0: This indicates that employees with negative "Intent to Stay" sentiment are less likely to leave, or that the question is not a strong predictor of turnover in your organization.
Odds Ratio = 1.0: This suggests there is no significant difference in the odds of leaving between the two groups, meaning the "Intent to Stay" question is not a leading indicator of turnover for your company.
Important Considerations:
If either the negative or positive sentiment group is empty (i.e., no employees in that category responded), we default to a neutral odds ratio of 1.0.
We then classify the calculated odds ratio based on specific effect size thresholds to help you understand the strength of this predictive relationship for your organization.
How We Identify At-Risk Groups (Risk of Exit Group Table)
The Risk of Exit Group Table helps you pinpoint specific groups within your company (e.g., by department, tenure, location) that show a higher or lower risk of turnover compared to your overall company average.
Here's how we determine risk for each group:
Identify At-Risk Sentiment within the Group: For each specific group, we measure the proportion of employees who responded negatively (e.g., "Disagree" or "Strongly Disagree") to the "I intend to work at this company for more than 2 years" question in your recent surveys. This indicates a potential intention to leave.
Compare to the Company Average: We then compare this group's proportion of negative responses to the proportion of negative responses across your entire company (the overall average).
Statistical Validation for Significance: To ensure that any observed difference isn't just due to random chance, we perform a rigorous statistical analysis - the hypergeometric test. This analysis helps us confirm if a group's sentiment is significantly different from the company average.
What "Significant Difference" Means: If a difference is "significant," it means there's a high probability that this group genuinely has a higher or lower proportion of at-risk employees, and it's not just a random fluctuation.
Displaying the Risk Indicator: Based on the statistical analysis and the magnitude of the difference (using the Glass Delta measure for the effect size calculation), we display a clear indicator and label for each group in the table. These labels, defined by the following thresholds, help you quickly identify which groups warrant closer attention:
Access the Risk of Exit insight analysis
The risk of exit insight analysis can be accessed in two ways:
From the Stories page, within the Turnover & Retention chapter
2. From the Explore insight button on the risk of exit group table
Adding the Risk of Exit group table to boards
The risk of exit group table can be added to your board by selecting add to board and choosing the “insight” option.
Select Risk of exit as the topic and then select + Add to board.
Navigate to the demographic filter dropdown and select an option to display a list of groups, their headcount, exit risk and negative sentiment.
Select + Add to board for any group to add this group to your board, by using this feature you can pin and compare groups across different demographic sets.
Next Steps:
🚑 Data Health 101 (to ensure the data in your insights is valid)
📖Stories and Chapters 101 (for an overview of available metrics for your reports)
💬 Need further help? Just reply with "Ask a Person" in a Support Conversation to speak with a Product Support Specialist.