Free template
Mentorship Program Survey Questions
Baseline, midpoint, and close-out question sets that produce numbers leadership will accept, plus the questions to cut because they only measure politeness.
- Time to complete
- 30 minutes to adapt all three surveys
- Use it when
- Before launch. The baseline has to go out before anyone is matched.
Three surveys, three jobs
| Survey | When | Length | What it is for |
|---|---|---|---|
| Baseline | Before matching | 3 minutes | The comparison point that makes every later number mean something |
| Midpoint | Halfway | 2 minutes | Catching lapsed pairs while there is still time to fix them |
| Close-out | Final two weeks | 5 minutes | The numbers you report, and the raw material for the next cycle |
The baseline is the one programs skip. Without it you can report how people feel at the end, which is not the same as reporting that anything changed. Send it with the match announcement email, before the first session happens.
Survey 1: Baseline
Sent to matched participants before their first session. Three minutes.
1. How clear are you on what you need to do to progress to the next level here? [1 to 5: Not at all clear / Completely clear]
2. How many people more senior than you, outside your direct reporting line, would take a call from you about a work problem? [0 / 1 to 2 / 3 to 5 / More than 5]
3. In the last six months, how often have you had a conversation about your career that was genuinely useful? [Never / Once / Two or three times / Monthly or more]
4. How likely are you to still be at [COMPANY] in 12 months? [1 to 10]
5. In one sentence, what do you want to be different in 12 months? [FREE TEXT]
Questions 1, 2, and 4 are the three you will re-ask at close-out. Keep the wording identical, word for word, or the comparison is worthless.
Question 4 is sensitive. Say plainly in the survey intro that responses go to [NAMED PERSON] only and are reported in aggregate, never individually, and never to managers. If you cannot honestly promise that, cut the question rather than break the promise.
Survey 2: Midpoint
Two minutes. The job of this survey is operational, not analytical: find the pairs that have stopped meeting while there is still half a cycle to save them.
1. How many sessions have you had so far? [0 / 1 / 2 / 3 / 4 or more]
2. When is your next session booked? [It is in the diary / We have agreed a rough time / Nothing booked]
3. Is this still worth your time? [Yes / Not sure / No]
4. Anything you want the program manager to know? [FREE TEXT, OPTIONAL]
Act on this within 48 hours. Anyone answering 0 or 1 to question 1, or "nothing booked" to question 2, gets a direct message that week. Anyone answering "no" to question 3 gets a conversation, not a nudge.
A midpoint survey you collect and do not act on is worse than not running one, because you have now asked people to tell you something is broken and then visibly done nothing.
Survey 3: Close-out
Five minutes, sent in the final two weeks.
Repeat questions
Identical wording to the baseline. These four give you your change numbers.
1. How clear are you on what you need to do to progress to the next level here? [1 to 5]
2. How many people more senior than you, outside your direct reporting line, would take a call from you about a work problem? [0 / 1 to 2 / 3 to 5 / More than 5]
3. In the last six months, how often have you had a conversation about your career that was genuinely useful? [Never / Once / Two or three times / Monthly or more]
4. How likely are you to still be at [COMPANY] in 12 months? [1 to 10]
Program questions
5. How many sessions did you have in total? [NUMBER]
6. Did you achieve the goal you set at the start? [Yes, fully / Yes, partly / No, but the goal changed / No]
7. Was your match a good fit? [1 to 5]
8. What is one thing you do differently now as a result of this program? [FREE TEXT]
9. Would you recommend this program to a colleague? [1 to 10]
10. What would you change about how the program is run? [FREE TEXT]
Question 8 is the most useful question on this survey. A participant who cannot name one behavioural change had a pleasant six months and nothing else, and the count of blank answers to question 8 is a far more honest measure of program impact than question 9.
Questions to cut
These get asked constantly and measure almost nothing:
- "How satisfied were you with the program?" People report satisfaction with things that changed nothing, particularly when a colleague gave up time for them. Satisfaction here measures politeness.
- "Did you enjoy your sessions?" Same problem, more so.
- "How would you rate your mentor?" Puts participants in the position of scoring a colleague who volunteered, so the distribution collapses to the top two options. Ask about fit (question 7) instead, which is about the pairing rather than the person.
- Anything with more than five points on the scale, except the two deliberate 10-point questions above. Extra granularity you will not act on just costs completion rate.
What to report
Report three layers. Most programs report only the first, and then struggle to get funded again.
Layer 1: Did it run
- Match rate: mentees matched against mentees who applied
- Meeting completion: sessions held against sessions the agreed cadence implies
- Pairs that reached close-out without ending early
Layer 2: Did it work for participants
- Goal completion: question 6, "fully" plus "partly"
- Change in clarity: question 1 at close against question 1 at baseline
- Change in senior network: question 2 at close against baseline
- Named behavioural change: proportion of question 8 answers that are not blank
Layer 3: Did it move a business number
- 12-month retention for participants against a matched non-participant group
- Promotion rate over the same period, same comparison
State the selection effect openly. People who volunteer for a mentoring program are not a random sample, and a CFO will spot that before you finish the slide. Match your comparison group on level, tenure, and function, say plainly that this is a matched comparison rather than a randomised trial, and you will be believed. Claim causation and you will not be.
The full reporting frame, including what you committed to at approval, is in the mentorship program proposal.
Questions about this template
- How do you measure the success of a mentorship program?
- Measure three layers: whether it ran (meeting completion), whether it worked for participants (goal completion and self-reported change against a baseline), and whether it moved a business number (retention or promotion rate against a matched comparison group). Most programs report only the first layer, which is why most programs struggle to get funded a second time.
- What questions should a mentorship program survey ask?
- Ask about meeting frequency, goal progress, relationship quality, and one behavioural change question tied to the participant's own goal. Avoid satisfaction questions as your headline metric: people report satisfaction with things that changed nothing, because they are being polite about a colleague who gave up time for them.
- When should you survey mentorship program participants?
- Three times: a baseline before matching, a short midpoint check, and a close-out. The baseline is the one programs skip and the one that makes every later number meaningful, because without it you can only report end-state feelings rather than change.
- How do you prove the ROI of a mentorship program?
- Compare participants against a matched non-participant group on retention and promotion at 12 months, matching on level, tenure, and function. This is not a randomised trial and you should not present it as one, but a matched comparison with the selection effect stated openly is far more credible to a CFO than a satisfaction score.