Employee Movement Model
Learn how external and internal employee movement is calculated.
Overview
Employee movement can be broken down into two categories: external and internal. External movement is a result of employees who are joining or leaving the organization (starts and exits). Internal movement is a result of changes to employee attributes (for example, changes to an employee's location or organization).
Understanding internal movement
Internal movement is based on changes to employee attributes and the selected analysis context (filters that you've applied and time period that you've selected). To calculate internal movement, the solution looks at changes in primary attributes (attributes that are defined by your organization) and changes to secondary attributes (the other attributes that have been applied to your analysis context). By default, the primary attributes that are examined for changes are Location and Organization.
Internal movement can be broken down into the following components:
- Moves in: Employees who have moved into the analysis context as a result of changes to their primary attributes.
- Moves out: Employees who have moved out of the analysis context as a result of changes to their primary attributes.
- Moves within: Employees who remain within the analysis context even with changes to their primary attributes.
- Others in: Employees who have moved into the analysis context as a result of changes to their secondary attributes.
- Others out: Employees who have moved out of the analysis context as a result of changes to their secondary attributes.
Note: To change the labels displayed in movement visualizations, see Create a Movement Concept.
Example: Calculating internal movement
Let's say you're looking at employee movement for March 2018 with the following filters applied:
- Location: New York
- High Performers (key group based on the Performance Rating attribute that includes employees who have a performance rating of Level 4 or higher).
To calculate internal movement, the solution will look for changes in the Location, Organization, and Performance Rating attributes for the employees in the analysis population. If a change is detected, the solution will then determine if the change caused the employee to move in, out, or within the analysis context. The analysis population is based on your analysis context, so the solution will look at employees in New York who are High Performers for March 2018.
The following table shows the employees in the analysis population and the attribute value changes that were detected for March 2018:
| Employee | Location | Organization | |
|---|---|---|---|
| A | New York to Los Angeles | Sales | Level 5 |
| B | New York | Sales to Marketing | Level 4 |
| C | New York | Product | Level 3 to Level 5 |
Based on the analysis context, you would see a total of 3 internal moves:
- Employee A will be counted as a Move out because they are no longer a part of the analysis context as a result of the change in location from New York to Los Angeles.
- Employee B will be counted as a Move within because the solution detected changes to the organization during the period. However, the change did not result in the employee moving out of the analysis context.
- Employee C will be counted as Others in. Although there were no changes to the primary attributes (Organization and Location), the change to the secondary attribute (Performance Rating) resulted in the employee moving into the analysis context.
Updates to employee movement counts
You may see changes to your movement counts as a result of the updates that were made to the Employee Movement model for the Summer 2019 release.
Support for non-categorial attributes
We've added support for non-categorial attributes, so movement can be detected for any attribute that you add to your analysis context.
This includes:
- Movement for time sensitive attributes (for example, Tenure and Age).
- Movement as a result of employees moving from one bucket to another (for example, Performance Group and Compa Ratio).
- Movement for referenced attributes where changes in another subject's attribute values cause employees to move in or out of the analysis context (for example, Direct Manager > Location). In this example, movement is based on changes to the location of an employee's manager instead of the location of the employee.
To detect changes in these attributes, the solution will sample data at set intervals based on the time granularity that you've selected. For example, if you set your time granularity to monthly, the solution will compare the value at the beginning of the month to the value at the end of the month. This means if the attribute value changes multiple times in the month, it will only be counted once. You will see the net result of the changes.
Note: The sampling level is capped at monthly intervals. If you set your time granularity to year, multiple moves may be counted if the attribute changes multiple times within the year.
Example: Internal movement based on tenure
Let's say you want to look at internal movement for employees who have been with the company for 5 to 10 years.
To conduct your analysis, you set the following analysis context:
- Apply a filter for Tenure: 5 to 10 yrs
- Select a time period from January 2018 to December 2018 at a monthly granularity
Based on the monthly granularity, the solution will sample the data at monthly intervals to detect changes to the Tenure attribute value. An internal move will be counted if it causes the employee to move in or out of the analysis context.
The following illustration shows the analysis population and the changes to their tenure during the time period.
Based on the analysis context, you would see 2 internal moves:
- Employee A moves into the analysis context in February 2018 as their tenure increases from 4 to 5 years. This is counted as an Others in.
- Employee B moves out of the analysis context in July 2018 as their tenure increases from 10 to 11 years. This is counted as an Others out.
Changes to movement counts due to reorganization
Now that we sample data at set intervals to detect changes for referenced attributes, you may see different counts for movement associated with reorgs. Previously, each change to the organization attribute that was a result of restructuring was counted as a move. With the changes in sampling, not all attribute changes will be detected. For example, if you set your time granularity to monthly, the solution will compare the value at the beginning of the month to the value at the end of the month. This means if the attribute value changes multiple times in the month, only one move will be counted.
Example: Multiple reorgs in a month
The following illustration shows changes to the organization attribute value as a result of multiple reorgs. Let's assume that you're looking at a monthly granularity.
- Before the Summer 2019 release, 2 internal moves would be counted (IT to Product and then Product to Finance).
- After the Summer 2019 release, only 1 move is counted (IT to Finance).
Example: Multiple reorgs in a year
The sampling level is capped at monthly intervals. This means if you set your time granularity to year, multiple moves may be counted if the attribute value changes multiple times within the year.
- Before the Summer 2019 release, 3 moves would be counted (IT to Product, Product to Finance, and Finance to Product).
- After the Summer 2019 release, 2 moves are counted (IT to Finance and Finance to Product). Since the solution samples the data at a monthly interval, the change from IT to Product that occurred on January 5 will not be detected.