Data-driven method for identifying top candidates
What is Candidate Match?
Developed by two Wharton professors, Candidate Match uses a diagnostic tool combined with a machine learning algorithm to identify top candidates for job openings based on your preferences. First piloted in the fall of 2016, Candidate Match is open to all employers recruiting at Penn this year. Candidates are undergraduate seniors from the College of Arts & Sciences, School of Engineering & Applied Science, and the Wharton School.
Since the tool uses data-driven methods to identify candidates, Candidate Match is a useful complement to your existing recruiting strategies.
Candidate Match is part of a broader initiative focusing on Analytics at Wharton, bringing the tools of cutting edge research to firms and policymakers.
Diagnostic tool to identify firm preferences
Machine learning + real resumes of Penn Seniors
Custom recommendations of candidates
How does Candidate Match work?
- Employers complete the Candidate Match diagnostic tool.
The diagnostic tool takes about 30 minutes to complete and involves evaluating 40 hypothetical resumes. The content of the resumes is tailored to efficiently identify preferences for different candidate characteristics.
- The tool uses a machine-learning algorithm to identify candidates.
Results from the diagnostic tool will identify multiple dimensions of employer preferences. Using a machine-learning algorithm, employer’s preferences will be matched with resumes from real Penn seniors to provide custom recommendations of candidates.
- Employers receive resumes that match their reported preferences.
The contact information and resumes of top candidate matches will be sent to employers for consideration. Employers can choose to invite candidates to visit them at a career fair and/or apply for a job at their organization. Candidates are not notified of any matches.
Who should complete the diagnostic tool?
- Respondents should be those involved in screening resumes for a position.
- Multiple people from one firm can participate.
- Different people should complete the diagnostic tool for different positions.
- More responses for one position can yield higher matching accuracy and precision!