What are the key metrics for a mentoring program?

Track how the program runs and who takes part: invitations accepted, participants matched, first meetings held, meetings per pair per month, pairs with no meeting in a set window, goals set and completed, survey response rate, and the hours the owner spends. These are operating measures. They do not prove a business outcome.

How often should you review mentoring metrics?

Look at quiet pairs and upcoming meetings weekly. Review the funnel and meeting counts monthly. Review the full set at the end of each cohort. A weekly look costs a few minutes and catches a stalled pair while a nudge can still help.

Can a mentoring program prove retention or promotion results?

Not from program data alone. Retention and promotion depend on many factors outside mentoring. Report them as context with stated limits, and keep operating metrics as the core of your reporting. See the ROI post for how to present estimates with assumptions.

What metric shows a mentoring pair has gone quiet?

Days since the last completed meeting, compared with the cadence the pair agreed. MentorStack flags pairs that have gone quiet on the admin dashboard and lets you send both people a one-click nudge. It does not re-pair them, so the next step stays with you.

Mentoring Program Metrics and KPIs: What to Track in the First Year

A metric earns a place on your list if you would act on it. If a number goes red and you would shrug, drop it. A first-year program can run on about eight measures. All of them come from things people did, not things people felt.

This post covers what to track, what each number tells you, and what to do when it moves. It stays on operation and participation on purpose. For the harder question of what a program is worth, read how to measure mentoring ROI.

Why measure operation before outcome?

You can act on operation. If few invitees accepted, change the invitation. If pairs stopped meeting in week three, change the follow-up. You cannot act on a retention figure in week three, and you cannot isolate mentoring's share of it later.

Operating metrics also protect you in a leadership review. "Forty of forty-eight matched pairs held a first meeting within two weeks" is a statement you can check. It does not need a caveat. (Those numbers are an example, not a MentorStack result.)

What should you track from invitation to first meeting?

This is the intake funnel. It shows where people drop off before the mentoring begins.

Stage	What to count	What a drop tells you
Invited	People sent an invitation	Your starting size
Joined	Invitations accepted	The ask or the channel is weak
Matched	Participants with an active match	Matching is slow, or supply is short on one side
First meeting	Pairs that met once	The kickoff did not set a date
Goal set	Pairs with at least one goal	The first meeting had no structure

The MentorStack admin dashboard shows this as a member onboarding funnel with the same five stages. In a spreadsheet, you can build it with five columns and a date for each stage.

Watch the gap between "matched" and "first meeting." A long gap is an early stall to watch for. A kickoff message with a proposed date shortens it.

What should you track once pairs are meeting?

Meetings per pair per month. Compare it with the cadence you set. If you asked for every two weeks and the median is once a month, the cadence was too ambitious or the scheduling is hard. The meeting cadence post covers how to choose.

Pairs with no completed meeting in the window. Pick a window, such as three weeks, and list every pair past it. This is your weekly working list. In MentorStack the dashboard's at-risk panel builds it and a nudge sends both people a check-in. The product does not re-pair anyone. You decide what to do next.

Cancelled and missed meetings. One cancellation is noise. A run of them is a signal that one person's calendar cannot carry the commitment.

Goals set, and goals completed. Count both. A pair with no goal at day 30 needs a prompt. A pair with goals set and none moving at day 60 needs a conversation. The goal-setting post explains why direction should come before action items.

What should you ask in a survey?

Keep it short and ask at set points. A check-in at about 30 days and a close-out at the end of the cohort are enough.

Ask whether the conversations are relevant to the person's goals, whether they can keep to the agreed rhythm, and whether they want to continue. Track the response rate as its own metric. A low response rate means you are reading the opinions of the few who answered. The survey question template has wording you can copy.

What should you track about the owner's time?

Almost nobody does, and it matters most if you plan to grow.

Log the hours you spend each week on matching, reminders, follow-up and reporting. Do it for four weeks. Then divide by the number of active pairs. You now have a coordination cost per pair, in your own numbers.

That figure tells you what a second cohort will cost. If twelve pairs take you six hours a week, forty pairs will not take two. This is the signal behind when a pilot outgrows its spreadsheet, and the sum a leader will want before approving more budget.

What should you not track?

Skip numbers that reward activity for its own sake. A count of messages sent tells you little about whether a pair is useful to each other. A count of meetings can rise when people hold short, empty meetings to hit a target. Skip any figure you would not change your plan for. If you track it, someone will try to improve it, so pick numbers where improvement means the program is working better.

How do you set targets?

Set them from your own first cohort, not from another company's numbers. A target borrowed from a vendor report describes a different program, a different audience and a different definition of "active."

Run one cohort with no targets. Write down what you saw. Set the next cohort's targets a little ahead of that, and tell participants what you are measuring and why.

Where do outcome measures fit?

After operation is stable, add outcome measures as context. Retention, internal moves and engagement scores all have other causes. Show them next to your operating metrics, state that mentoring is one influence among several, and avoid saying mentoring caused the change. The ROI post shows how to present an estimate as a range with its assumptions.

How do you report this?

One page. The funnel, meetings held against plan, the count of quiet pairs and what you did about them, goal progress, survey results, and owner hours. A dated snapshot at the end of each cohort is enough for most leaders.

In MentorStack, report generation and export start on the Starter plan, and scheduled reports start on Growth. The Free Pilot has a basic dashboard. See the analytics features for what each view shows. If you are weighing a tool against your current spreadsheet, the spreadsheet comparison lists what each does with these metrics, and mentoring software for scaling a pilot describes the move in stages.

What should you do this week?

Pick three numbers you can pull today: joined, first meeting held, and quiet pairs. If you cannot pull them in under an hour, that tells you something about your tooling. A quick look at where your program stands overall is in the free mentorship maturity assessment.