Concept: Optimizing Suggested Job Openings in the Career Hub
Machine Learning Input Data
When you enable suggested job opportunities for Opportunity Marketplace in
Career Hub
, Workday uses several explicit and inferred data points as machine learning (ML) inputs to determine relevant job openings in the section. We assign the most weight to the match between the skills on the worker’s profile and the explicit skills on the job requisition, but we also consider:
- Similarity between the job requisition title and the worker’s current job title. We only suggest jobs that have a certain level of similarity.
- Skills that we infer from the job requisition.
- The worker’s experience in a skill, which we infer from their job history.
- Whether the job matches the worker’s current job family, job family group, or job interests.
For a full list of input data for ML suggestions in the
Career Hub
, see Workday AI Fact Sheet: Career Hub.Seniority Modeling
When you use the recommended
Default (ML - Generated)
for the Hierarchy
option on the Edit ML Preferences
tab of the Maintain Career Hub and Opportunity Marketplace
task, we also infer the seniority levels of each job profile using:- Its job level name.
- Its management level name.
- The names of the job profiles in the same job level and management level.
We surface job suggestions from the worker’s current level of seniority, determined by their job profiles, and up to 2 levels higher in our seniority modeling.
When a job profile doesn’t have either a job level or a management level, we don’t consider seniority when suggesting its open jobs to workers, or suggesting job openings to workers with this job profile.
This setting doesn’t apply to job profile suggestions in the
Explore
page and Career Path Builder. Use the Edit Explore Preferences
tab on the Maintain Career Hub and Opportunity Marketplace
task to change your hierarchy settings for ML suggestions in the Career Path Builder. See Configure Career Path Builder for details.Optimizing ML Input Data
Follow these best practices to increase the number and quality of suggested job openings for your workers:
- Add 5-10 skills to job profiles. We prioritize these explicit skills over the skills that we infer from the job requisition.
- If you use job levels, management levels, or both, Workday recommends the default ML setting for determining job seniority. To match workers to job openings at appropriate seniority levels with this setting, ensure that:
- Your active job level hierarchy and management level hierarchy have meaningful level names that reflect seniority in your organization.
- Your job profiles have meaningful names.
- Your job profiles have accurate job or management levels. You can use theAll Active Job Profilesreport in theJob Architecture Hubto check the hierarchies for your active job profiles.
Filtering Suggested Jobs by Hierarchy Level
Instead of the recommended default setting, you can select
Job Level
, Management Level
, or Compensation Grade
for the Hierarchy
option on Maintain Career Hub and Opportunity Marketplace
> Edit ML Preferences
. In this case:- We limit jobs directly by theOrderfield in the selected hierarchy, without using machine learning to infer job seniority. Workday suggests jobs at the worker's current grade or level, or 1 order higher.
- We do not surface suggestions without a grade or level in a hierarchy.
- You also need to specify theHierarchy Direction Order:
- Ascending: The most junior grade or level has the lowest order number.
- Descending: The most senior grade or level has the highest order number.
To configure hierarchies for directly filtering job suggestions, see the Improving Job Suggestions Using Hierarchies section in Workday AI Fact Sheet: Talent Marketplace.
Example: You don’t use job or management levels, so you select
Compensation Grade
. We only suggest jobs with a compensation grade on the job requisition that is the same as or 1 grade above the compensation grade for the worker’s position.