Employee at Risk  in  Talenta Insights shows Predictive Attrition Risk that predicts the risk rate of Employee Attrition .  Employee attrition (Reduction in the number of employees) is a condition where the number of employees leaves the company gradually. This condition occurs due to resignation, age, health, etc. This prediction is valid for the next 1 month only.
E.g.: If you open the dashboard in December 2023, the system will calculate data from 1-31 November 2023 for predictions for January 2024.

Important
The factors used as a basis for analysis will differ from one company to another and change from time to time. Some of the main general factors used are tenure, average schedule, employment status, 1 yearly/6 monthly salary increases, leave, overtime, average real working hours, and division.

Following are the steps to access information for Employees at Risk.

1. Go to the insights.talenta.co page then  Sign in  to your account.
2. After entering the Talenta Insights page, click the menu "Employee At Risk" .
3. Then you will be taken to the Employee at Risk page .
4. If you want to set your Employee at risk dashboard display to display a certain time range and/or a certain branch/organization , use filter columns,  then click “Apply filter”.
   
5. At the top right of the page, you can see the following display.
   
   

   No.	Column/Button Name	Description
   1	Accuracy score	Accuracy indicator value (in percentage form) of how accurate the modeling used by Talenta is. The system calculates the total number of predictions who have resigned and those who have correctly resigned, as well as the total number of predictions who have not resigned and those who have correctly not resigned. The score displayed by each company varies depending on the number of employees and employees who use ESS (Employee Self Service) Talenta . The more employees and workers who use ESS Talenta, the more accurate the score is. You can read further explanations in the Important box below the table.
   2	View factors	Click to open an explanation view of key factors. The factors below are determined with InterpretML. With  InterpretML , the factors that most influence the prediction of employee resignation can be seen. Apart from that, this feature uses LightGBM to get employee resignation predictions. Here's the explanation: Working period Assess the total duration of active work, measured from the start date. Employees who stay longer are less likely to resign. Average schedule Calculate the average number of hours worked by an employee in the last month. Fewer hours means employees stay longer. Employment status Analyze employment status (internship, contract, permanent) with a resignation risk score. Those who remain are less likely to resign. 1 year's salary increment Measuring salary increases in a year. Higher increases correlate with a lower risk of resignation. Time off Track the total leave taken by an employee in a month. Employees who take more leave are more likely to resign. Overtime Track how often employees work overtime in a month. Less overtime means employees will work longer hours. 6 months' salary increment Measure salary increases in the last 6 months. Higher increases correlate with a lower risk of resignation. Average actual work Average actual working hours of employees in a month. The lower the actual working hours, the lower the chances of an employee resigning. Organization Analyze turnover rates by department to identify areas with high turnover tendencies. Lower turnover often indicates effective management practices. 2 years' salary increment Measure salary increases in the last 2 years. Higher increases correlate with a lower risk of resignation. Job levels Review various job levels from low to high. Employees with higher levels of employment are less likely to resign. Age Analyze resignation trends across various age groups. Older employees generally have lower attrition rates. Late in Track total days of employee tardiness per month. Frequent lateness may indicate a higher risk of resignation. Tax status Analyze employee dependents for PTKP (Non-Taxable Income) tax purposes. Employees with more dependents often stay on the job longer. Average early out Track the average time employees leave early each month and the number of days it occurs. Frequently leaving early may indicate a higher risk of resignation. Reprimand Calculate the number of warning letters an employee receives in a month. Employees who receive fewer reprimands are less likely to resign. Click the  “cross” icon  to close the pop-up.
   3	Help	Click to open the following dropdown. - Guidebook: Click to open this Employee at risk guide article. - Quick tour: Click to get a quick guide tour directly on Talenta. - Share feedback: Click to open a pop-up that you can fill in with criticism and suggestions for this feature.

Important
Factors that Influence Accuracy in Score Accuracy:
- Accuracy varies between companies, depending on the number of employees and use of ESS Talenta.
- The more employees and users of ESS Talenta, the more accurate the predictions.
To evaluate the effectiveness of the prediction model, we use Precision and Recall.
- Precision: This metric assesses how accurate the predictions made by the model are. In the context of this dashboard, precision answers the question: "Of all the employees the prediction model said would resign, how many actually resigned?"
High precision means that when the model predicts an employee will resign, there is a high probability that the prediction is correct.
Example: If the model predicts that 100 employees will resign and 80 of them actually resign, then the precision is 80%.
- Recall: This metric assesses the model's ability to identify all relevant cases. In the context of this dashboard, recall answers the question: "Of all the employees who actually resigned, how many did the model successfully predict that they would resign?" High recall means the model is effective in identifying the majority of employees who actually resign.
Example: If 100 employees actually resign, and the model only succeeds in predicting 70 of them, then the recall is 70%.
These two metrics are important because they provide a balanced picture of model performance. High precision with low recall could mean the model is too careful and only predicts resignation when it is very confident, but misses many other cases. On the other hand, high recall with low precision could mean that the model predicts resignation too often, resulting in many wrong predictions.

The following is an explanation for each of the information presented on the Employee at Risk page. 

A. Summary Predictive Attrition Risk

In this first section, you can see a brief summary per category of Attrition Risk predictions shown in percentage and number of employees.
This summary information is grouped by:

B. Overview of Attrition Risk

In this second part, overall information will be presented regarding predictions regarding the possibility that employees will resign or leave the company in the next month. The information presented is grouped by level: 

• Low risk, the likelihood of employees resigning is below 25%.
• Medium risk, the likelihood of employees resigning is between 25-50%.
• High risk, the likelihood of employees resigning is 50% -75%.
• Critical risk, the likelihood of employees resigning is 75% -100%

You can download the diagram in spreadsheet form by clicking  “Download”  which will direct you to the page  Export history.

C. Attrition Risk Trends

In this third section, you can see a brief summary per category of Attrition Risk predictions shown in percentage and number of employees.

- Show percentage number:  Check if you want the diagram to show the Attrition risk trends percentage number. You can click again to remove the percentage number.
-   Download: You can download the diagram in spreadsheet form by clicking this button.
- Notification Alert bar:
Yellow  means that the system predicts that there will be an increase in the percentage of employees leaving the company compared to last month.

Green means the system predicts there will be a decrease in the percentage of employees leaving the company compared to last month. 

White means that the system predicts that there will be a certain percentage of employees leaving the company compared to last month.

D. Attrition Risk Breakdown

In this fourth section, predictions will be presented based on employee grouping filters, such as Organization, Branch, Gender, and so on. You can see more detailed information on the right side of the screen.

- Search bar: You can type certain keywords to search for specific information in this part of the diagram.
- Download: You can download the diagram in spreadsheet form by clicking this button.
- Breakdown by: Customize the diagram display based on your needs. The example in the image above is a breakdown based on Organization. If you select Breakdown by Organization (and certain breakdown categories), then, you can add details View by.

E. Risk vs Protective Factor

In this fifth section, you can see the 10 (ten) factors that most influence employees' conditions for resigning (Risk factor) or settling (Protective factor). That way, companies can find out what factors need to be improved within the company to retain their employees.

-  You can download the diagram in spreadsheet form by clicking  “Download”.
- You can hover over any of the bars to see the details. The red factor with a risk value (Risk score) positive (moving to the right) means that the larger the factor, the higher the probability of an employee resigning. The percentages shown are a comparison with last month. Following are the details of the Risk factor (red).

Meanwhile, here are the details of Protective factors (green). The green factor with risk value (Risk score) negative (moving to the left) means that the larger the factor, the lower the probability of an employee resigning. The percentages shown are a comparison with last month.

F. Employee Attrition Details

In this sixth section, you can see a list of employees, along with complete personal data information and the level of possibility of resigning.

This is an overview of Employee at Risk. Furthermore, you can learn about Employee Performance here.