The predicted count is calculated by applying the predicted rate of the event for an employee population (e.g., actual exits, internal moves, promotions) to the number of employees in that population. The output is rounded to the nearest whole person.
For example, if the predicted rate of the event for a population of 100 employees is 10%, then the predicted count would be 10. Statistically, we expect that 10 of the 100 employees will exit, move, or be promoted in the next year.
When you view details for this population, you will see all 100 employees listed even though the predicted count is 10. This is because our methodology uses probabilities and does not specify exactly which 10 people the prediction will apply to. What you will see in view details is a list of employees sorted by their prediction scores. The data is presented this way to allow you to make a fully informed assessment based on a broad understanding of all the employee’s attributes.