Note: The candidate grading functionality is only used for external candidates. Internal candidates aren't assigned a grade when they apply. The system doesn't display candidate grades for candidates who reside in New York City, to comply with New York City legal requirements.
When the candidate grading functionality is enabled, Dayforce uses artificial intelligence (AI) to assign a grade to candidates when they apply for jobs at your organization. These grades are displayed in the Recruiting feature along with a report card with additional details.
As part of the initial configuration for this functionality, you must initiate a one-time learning process for the candidate grading AI using the Learn from Historical Data button displayed for the Candidate Grade Enabled client property in Recruiting Setup > Client Properties. This learning process involves analyzing historical recruiting data, including job application resumes, job requisition details, job posting details, and job description details.
This process also uses the settings in the Progress Indicator column of the Candidate Status feature and the Decline Reason Type column of the Decline and Offer Rejection Reason feature to learn how your organization uses candidate statuses to screen and move candidates through the recruiting process. See Configure Candidate Grading.
To avoid introducing bias during the grading process, the following steps are taken:
- Hiring Status: Historical recruiting data is considered for all candidates, not just for those who were hired.
- Name: Candidate names are removed from job applications before they are screened to help avoid implicit bias.
- Years of Experience: Years of experience is only considered up to ten years for each candidate to help avoid age bias.
- Location: Specific zip or postal codes aren't used. However, the grading AI does consider a radius around the job location (if location is determined to be relevant). Candidates who are closer to the job's location are weighted higher for that element of the overall grade, starting from approximately 100 kilometers or 62 miles away from a job's defined location.
With these bias-reducing measures in place, the grading AI learns from your organization's past hiring practices and constantly evolves to provide the most accurate grading possible. This removes the need for recruiting and hiring managers to manually screen candidates because the AI takes care of it using the same learned logic.
If your organization doesn't have any historical recruiting data to learn from, the AI grades candidates based only on how closely their job application matches the job description. The accuracy of these grades will vary depending on the role and the quality of the job description. As time goes on and you hire and decline more job applications, the grading AI will be able to learn from the status and decline reason mappings configured using the Progress Indicator column and the Decline Reason Type column.
Learning occurs automatically when enough recruiting data, 20 positive training events and 20 negative training events, have occurred for a job requisition. Positive training events are when a candidate is moved forward in the recruiting process and negative training events are when a candidate is declined (using a mapped status) without having moved into the interview process.
Grade Elements
The grade elements that are displayed in a candidate's report card depends on how much training data was available for the job requisition at the time of grading. If the 20 positive and 20 negative training events have yet to occur, some of the grading elements might not be displayed. The weight attributed to each element is also dependent on the recruiting data available at the time of grading and is based on the likelihood that a certain element will lead to an interview. For example, the Location element might be assigned a higher weight because a candidate's proximity to a job creates a predictable pattern for grading when training data for the other elements isn't available.