MentorPairing, the mentor matching tool that hands you a draft instead of a verdict · Updated September 8, 2026

What MentorPairing actually does

The workflow has four moves. First, two linked application forms (one for mentors, one for mentees) collect the same fields on both sides: industry, function, years in role, city, meeting cadence and three topics the person wants to work on. Conditional questions and resume uploads are supported, and applications close on a date you set, which spares you the annual ritual of chasing the last few stragglers.

Second, the matching engine scores the eligible pairs and proposes a full cohort, which the site says takes under a minute. Third, you review the suggestions with the reasoning shown next to each one, lock the pairs you like and re-run the rest as often as you want. Nothing is final until a human says it is.

Fourth comes launch: introduction emails go out with bios and a scheduling link, pairs log their meetings (the site describes it as two taps), inactive pairs get an automatic nudge after 30 days, and a dashboard keeps a running count so you are not reconstructing the year in June from memory and email threads.

Weighted matching, explained without the mystique

Every question you collect gets sorted into one of four buckets: hard rule, strong preference, tie breaker, or ignore. Hard rules eliminate impossible pairs outright (same employer, wrong campus, whatever your program cannot allow). Preferences and tie breakers move the score around. That is the entire mechanism, and MentorPairing is refreshingly unbothered about saying so: the site describes the engine as doing the part you would otherwise do with a highlighter.

There is no AI claim anywhere on the page, which in 2026 counts as a personality trait. Asked directly whether an algorithm matches better than a coordinator does, the site answers no, and adds that it does not try to. The engine filters out what cannot work and ranks what can, and MentorPairing states that coordinators typically override 5 to 10 percent of the suggested pairs. If you were hoping to stop thinking about your cohort entirely, this is not that product.

Life after the pairing: nudges, reporting and rematches

Pairing is the part every program plans for, and the months afterward are where participants quietly go missing. MentorPairing covers that stretch too. Reporting counts participation, meetings held, hours contributed, retention by cohort and check-in scores, filterable by chapter, campus or class year, and it leaves as a CSV or as a read-only web link for the board members who are never going to log in.

People also drop out, so there is a rematch flow: mark a pair closed with a reason, and the engine proposes a replacement from the waitlist while the historical record stays attached to both participants. Exports are logged as well, with the account that ran them and the time, which is a small feature that saves a large argument later.

MentorPairing publishes its own program statistics on the page (pairs matched to date, meetings logged, median hours saved when building a cohort). Those are the vendor's figures rather than an independent audit, and this page is not one either. Read them the way you read any other number printed next to a buy button.

Pricing, fit and the fine print

Pricing is public and tied to active pairs rather than staff accounts: $79 a month for up to 60 pairs, $179 a month for up to 300 pairs, and $359 a month with the ceiling removed. All three tiers include unlimited coordinator seats, the matching engine, meeting logs and CSV export, billed monthly with no setup fee and no annual commitment. The site does not advertise a free trial.

Fit is a question of scale. The programs described on the site are professional associations, universities and colleges running intakes on an academic calendar, employee resource groups, and nonprofit or civic programs with rolling intake, generally in the tens to hundreds of pairs. If your mentoring program is currently a handful of pairs and a group chat, you would be buying machinery you do not need yet.

On the housekeeping: participant data sits in US data centers, encrypted in transit and at rest, and the site states it is never sold, rented or used to train anything. Access is role-scoped, university partners can sign data agreements (the page calls itself FERPA aware), and the Institution plan adds single sign-on plus a read API. Two edges worth knowing before you commit: cancelling stops the plan at the end of the month you have already paid for, with no prorating, and your data stays exportable for 90 days after closure. Pull the CSV before that clock runs out.