How does the recommendation engine work?

How does the engine calculate recommendations?

For each employee, the engine checks which reviews are in Poplee Performance (including imported review dates). To identify a review type, the engine uses the category associated with the campaign. This enables it to determine, for example, whether an employee has already had a professional review or not.

The engine calculates the recommendations each time you view the "recommendations" tab, to keep them up to date.

If an employee has a review in progress in the rule category, they will not be recommended. In fact, even if the review has not been completed, the employee will not be recommended, as the engine considers the employee to be in the process of completing the review. Therefore, it makes no difference whether the interview is in preparation, execution, approval or validation. 

Rejected reviews are taken into account by the engine, we consider a rejected review to be a valid review.

Archived reviews are NOT taken into account by the engine. In fact, if the review has been archived, the employee no longer has the opportunity to complete the review, so the engine does not take this review into account. 

The engine does not take into account employees who are excluded from the rule at the date you are planning. 

The engine recommends only one review for each rule per employee. For example, if an employee has missed several professional reviews, the engine will only recommend one professional review for the employee. 

For a recurring rule, the engine uses:

- the reference date (e.g. contract start date) if no review has been conducted

- the date of the last review, if a review has already been conducted. The engine will then look at the preferred completion date you defined as HR administrator when you added the reviews to the campaign, and not at the date where the review was actually conducted. 

For example, if you have a recurring annual review rule that is conducted every year, and as HR admin you have scheduled a review for Valentin with a completion date of May 1, 2021. In fact, Valentin was delayed and completed his review on July 1, 2021. Valentin will again be recommended on this rule from May 1, 2022 and not from July 1, 2022. We would like the engine to be based on the date you defined, to avoid future reviews being postponed if the employee completes the review late.

Why is an employee not recommended to me? 

There are several reasons why you may not see an employee as a recommendation: 

  • The employee already has a review in progress in the same category as the rule. 
  • The employee is excluded (permanently or temporarily) from the rule on the date you projected. 
  • The employee is not part of your scope "create a campaign and administer campaigns where the manager is within the scope...".

As you can only create reviews for employees who are within your scope, if an employee is not within your scope, you will not see them in the recommendations.

  • The employee is not part of the target population defined by the rule filters (department, establishment, occupation category, contract type).
  • The employee does not have a reference date used by the rule.

For example, if the rule is based on the contract end date and the employee has no contract end date, then they cannot be recommended.

  • The employee does not have an active contract date today or an active contract in the future. 
  • If it is a one-off rule:
    • the rule was activated after the date when the employee should have completed the review.

For example, take a specific rule for the end of trial period reviews, which must be conducted 3 months after the contract start date. Valentin arrived on January 1, and you activate the rule on June 1. Valentin should have completed his review on March 1, as the rule was not active at the time. Therefore, Valentin will not be recommended.

    • the employee already has a completed review in the rule category 
  • If it is a recurring rule: 
    • the employee already has an review in the same category as the rule, within the rule's observation period.

For example, if you have a recurring professional review rule that applies every two years from the start of the contract, and Valentin already had a professional review last year, then he will not be recommended for today since it will have been less than two years since he had his review. 

    • the employee has a reference date within the observation period

For example, if you have a recurring professional review rule that applies every two years from the contract start date, and Valentin arrived less than two years ago then he will not be recommended. 


Page content

Was this article helpful?
0 out of 0 found this helpful