Mentoring Matching: Why the Algorithm Behind Your Pairs Matters
Every mentoring program makes matches. The question is whether those matches are made by instinct, by spreadsheet, or by science.
At MentorPRO, we built our matching algorithm on thirty years of peer-reviewed research into what actually makes a mentoring relationship work. The difference that makes is not theoretical. It shows up in the data.
What Is Mentoring Matching and Why Does It Matter?
Mentoring matching is the process of pairing a mentee with a mentor based on shared characteristics, goals, and compatibility factors. It sounds straightforward — but the research on what actually predicts a strong, lasting mentoring relationship is more nuanced than most programs realize.
Match on the wrong variables and you get pairs that close early, disengage, or never develop the trust needed to produce real growth. Match on the right ones and you get relationships that go the distance.
Most mentoring platforms offer matching as an administrative convenience. MentorPRO treats it as the most consequential decision in the entire mentoring process — because the evidence says it is.
Built by the World’s Leading Expert on Mentoring Matching
MentorPRO’s matching algorithm was developed by Dr. Jean Rhodes, Professor of Psychology at the University of Massachusetts Boston and Director of the Center for Evidence-Based Mentoring. Dr. Rhodes is the author of Older and Wiser: New Ideas for Youth Mentoring in the 21st Century (Harvard University Press, 2020), winner of the 2023 Eleanor Maccoby Book Award from the American Psychological Association — the field’s most prestigious recognition for contributions to developmental psychology.
Over the course of her career, Dr. Rhodes has published more than a dozen peer-reviewed studies specifically on mentor-mentee matching, compatibility, and match longevity, examining questions of race, gender, shared interests, screening, premature closure, and relationship quality across thousands of mentoring pairs.
That body of work is not background reading for our algorithm. It is the algorithm. No other mentoring platform was built by someone who has spent thirty years studying exactly this question.
What the Real-World Data Shows
Among pairs matched through MentorPRO, only 10.87% ended early. Among pairs assigned manually by experienced staff, 29.41% closed before completion — nearly three times the rate.
The strongest evidence for MentorPRO’s matching system comes not just from the research literature but from the programs using it in the field. Zing Programme, a leading mentoring organization based in Spain, completed an internal analysis comparing mentoring pairs matched through MentorPRO’s algorithm versus those assigned manually by experienced staff. Around 40 percent of their pairs were matched through MentorPRO, giving them a meaningful comparison group.
The findings were striking:
- MentorPRO-matched pairs: 10.87% early closure rate
- Manually assigned pairs: 29.41% early closure rate — nearly three times higher
- MentorPRO-matched pairs also scored higher on relationship quality
Research consistently shows that premature match closures can be harmful to mentees, particularly young people from marginalized backgrounds (Rhodes, 2020). Matches that stay together produce better outcomes.
Algorithmic Matching vs. Manual Assignment
| MentorPRO Algorithm | Manual Matching | |
| Variables considered | 15 research-driven variables | Varies by coordinator |
| Consistency | Uniform across all pairs | Subject to human variability |
| Scalability | Scales to any program size | Labor-intensive at scale |
| Early closure rate (Zing data) | 10.87% | 29.41% |
| Relationship quality | Higher | Lower |
| Evidence base | 30 years of peer-reviewed research | Practitioner judgment |
| Customizable | Yes | Yes |
Programs that rely on manual matching are not doing something wrong. Experienced coordinators bring real judgment and contextual knowledge to the process, and they always have the final say. But manual matching is limited by time, cognitive bandwidth, and the number of variables any one person can weigh at once.
An algorithm built on decades of research does not get tired, does not carry unconscious preferences, and does not run out of capacity when a program scales.
How MentorPRO Matching Works
MentorPRO’s matching process is customizable, transparent, and grounded in more than 15 research-driven variables. Programs work with our team to identify the variables that matter most for their specific population and context.
The process follows four steps:
- Identify your variables — Program managers select from our research-backed library, with the option to add program-specific variables
- Generate your intake form — MentorPRO creates a customized form linked directly to the platform
- Collect responses — Mentors and mentees complete the form; a designated point person can review submissions
- Confirm your pairs — Program managers review the algorithm-generated match list, confirm pairings, and grant platform access
Pairs can be re-matched in future cycles as needed. The platform also integrates with our Check-In, Goal Setting, and Flash Mentoring features for a fully connected mentoring experience.
Getting Started
MentorPRO’s matching system is available as part of the full platform, which also includes flash mentoring, goal-setting, check-ins, course delivery through MentorPRO Academy, and program management tools. Our team works with every partner to customize the matching variables, intake form, and workflow to fit their specific program model and population.
Frequently Asked Questions About Mentoring Matching
What is mentoring matching?
Mentoring matching is the process of pairing mentors and mentees based on shared characteristics, goals, and compatibility factors. When done well, it produces more stable relationships and better outcomes for both parties.
Is algorithmic matching better than manual matching?
Data from Zing Programme found that MentorPRO-matched pairs had an early closure rate of 10.87% compared to 29.41% for manually assigned pairs, with higher relationship quality scores as well. Manual matching draws on valuable practitioner judgment but is limited by time and cognitive capacity in ways a research-driven algorithm is not.
What variables does MentorPRO use for matching?
MentorPRO draws on more than 15 research-driven variables, including demographics, language fluency, goals, interests, availability, and program-specific criteria. All variables are customizable in consultation with our team.
How does MentorPRO matching support equity?
Young people from historically marginalized backgrounds benefit most from well-matched mentoring relationships and are most vulnerable to harm from poor matches or early closures. MentorPRO’s algorithm was developed specifically with diverse youth populations in mind, drawing on three decades of research.
Can programs of any size use MentorPRO matching?
Yes. MentorPRO’s algorithm scales from small cohorts to programs with hundreds of pairs, maintaining consistency across every match without increasing staff workload.
Is MentorPRO matching available for international programs?
Yes. Zing Programme in Spain is one example of an international partner using MentorPRO’s matching system with strong results across a culturally diverse population.
Selected Publications on Mentor-Mentee Matching by Dr. Jean Rhodes
Ramadurai, R., Deng, Y., Werntz, A., Yowell, C., and Rhodes, J. E. (2025). Mentor-mentee similarity in college peer mentoring: Untangling conflicting results. Studies in Higher Education. https://doi.org/10.1080/03075079.2025.2572504
Raposa, E. B., Ben-Eliyahu, A., Olsho, L., and Rhodes, J. (2019). Birds of a feather: Is matching based on shared interests and characteristics associated with longer youth mentoring relationships? Journal of Community Psychology, 47(2), 385–397. https://doi.org/10.1002/jcop.22127
Spencer, R., Gowdy, G., Drew, A. L., and Rhodes, J. E. (2018). Who knows me the best and can encourage me the most? Matching and early relationship development in youth-initiated mentoring relationships. Journal of Adolescent Research.
Kupersmidt, J., Stump, K., Stelter, R., and Rhodes, J. (2017). Predictors of premature match closure in youth mentoring relationships. American Journal of Community Psychology.
Kupersmidt, J., Stump, K., Stelter, R., and Rhodes, J. (2017). Mentoring program practices as predictors of match longevity. Journal of Community Psychology. https://doi.org/10.1002/jcop.21883
Rhodes, J., Schwartz, S., Willis, M. M., and Wu, M. V. (2017). Validating a mentoring relationship quality scale: Does match strength predict match length? Youth and Society, 49(4), 415–437.
Grossman, J. B., Chan, C. S., Schwartz, S. E. O., and Rhodes, J. E. (2012). The test of time in school-based mentoring. American Journal of Community Psychology. https://doi.org/10.1007/s10464-011-9435-0
Kanchewa, S., Rhodes, J. E., Schwartz, S. E. O., and Olsho, L. (2014). The influence of same- versus cross-gender matching in formal youth mentoring programs. Applied Developmental Science.
Rhodes, J. E., Reddy, R., and Rappaport, N. (2005). Promoting successful youth mentoring relationships: A preliminary screening questionnaire. Journal of Primary Prevention, 26(2), 147–167.
Rhodes, J. E., Reddy, R., Grossman, J. B., and Lee, J. M. (2002). Volunteer mentoring relationships with minority youth: An analysis of same- versus cross-race matches. Journal of Applied Social Psychology, 32(10), 2114–2133. https://www.rhodeslab.org/wp-content/uploads/2016/04/rhodes-et-al-2002-same-versus-cross-race-matches.pdf
Rhodes, J. E. (2020). Older and wiser: New ideas for youth mentoring in the 21st century. Harvard University Press. https://www.hup.harvard.edu/books/9780674292277
