Hiring guideWritten for frontline employers · 16 September 2026
Stay Interview Questions
A stay interview asks somebody who is still working for you what would make them leave, while you can still do something about it. It is a fifteen-minute conversation, run by the person's own supervisor, once or twice a year, off-cycle from performance reviews — and it is the cheapest retention instrument that exists, because it costs nothing but the fifteen minutes. The twenty questions below are grouped by what they are for, with what to listen for and the follow-up that matters. One rule governs all of them: a stay interview creates an obligation. Ask what would make someone leave, change nothing, and you have confirmed to your best people that raising something is pointless.
What a stay interview is, and the two things it is not
A stay interview is a short, structured, one-to-one conversation with a current employee about why they stay, what frustrates them, and what would make them leave. It is not a performance review and it is not an exit interview, and confusing it with either is how the practice fails.
It is not a performance review. The moment a conversation is attached to a rating, a raise or a record, you get answers calibrated for the rating. Run stay interviews off-cycle — different month, different tone, no form that goes in a file.
It is not an exit interview. An exit interview asks somebody who has already gone why they left, which is the least useful moment to ask: the decision is made, the relationship is over, and the answer is calibrated for the reference. The stay interview asks the same question of somebody who has not decided yet.
The one-line difference
An exit interview tells you why the last person left. A stay interview tells you why the next one will.
The rule that decides whether it works
A stay interview creates an obligation. If you ask what would make someone leave and then change nothing, you have not gathered data — you have confirmed to your best people that raising something is pointless, and you have accelerated the exact departure you were trying to prevent. Do not run stay interviews you are not prepared to act on.
The practical discipline is to leave every conversation with three things: one thing you will change, one thing you will go and find out about, and one thing you have said plainly that you cannot change and why. The third is not a failure. Being told a straight no, with a reason, is far better received than being asked and then ignored — and it is the part managers most often skip, because it feels like the conversation has gone badly. It has not.
If you cannot name one thing you will change before you walk into the room, you are not running a stay interview. You are running a survey with a witness.
Who runs it, and when
Two mechanics decide most of the outcome.
Run it with the direct supervisor, not HR. The supervisor is the variable most likely to be the answer, and the conversation is far less useful when that person is not in the room. It is also the person who can actually act on seven of the twenty answers before the end of the week. The obvious objection — what if the supervisor is the problem — is real, and the answer is a second route rather than a different owner: make sure everyone has somebody above or beside their supervisor they can raise something with, and notice when a whole team's answers cluster.
Run it off-cycle and on a rhythm. Once or twice a year per person, scheduled like anything else, fifteen to twenty minutes, and never in the same month as reviews. For frontline teams on shift patterns the practical version is a rolling schedule — two or three conversations a week until the team is covered, then start again.
- New hires: at 30 and 90 days, when the gap between what was advertised and what the job is will still be fresh.
- Everyone else: once or twice a year, spaced away from reviews.
- After a change: a new supervisor, a schedule change, a site move, a busy season — the month after, not the month of.
- Never: in the week somebody has resigned. That is an exit interview and it is too late for this.
The twenty questions
The order matters more than the wording. The first group is easy to answer and establishes that this is a conversation rather than an appraisal. The middle groups are where the useful answers live. The last question is the one people remember being asked.
1. Opening: establish it is a conversation
| Ask | What you are listening for | The follow-up that matters |
|---|---|---|
| 1. What do you look forward to on a shift here? | Whether anything positive is specific rather than general | “When was the last time that happened?” |
| 2. What is the most frustrating part of a normal day? | Friction they have stopped reporting because nothing changed | “Who else does that affect?” |
| 3. What is one thing we do that you would stop tomorrow? | The process everyone works around | “What would you do instead?” |
| 4. What made you say yes to this job in the first place? | Whether the job is still the one they accepted | “Is that still true?” |
2. The core: why they would leave
These four are the reason the conversation exists. Ask them slowly and leave silence after each one.
| Ask | What you are listening for | The follow-up that matters |
|---|---|---|
| 5. If you were going to leave, what would be the reason? | The real answer, which is almost never pay alone | Silence. Let them fill it. |
| 6. What would another employer have to offer you? | Your actual competitive gap, in their words | “Is that the money or the hours?” |
| 7. Have you thought about leaving in the last six months? | A yes is information, not a betrayal — react accordingly | “What stopped you?” |
| 8. What would make you stay for another two years? | Whether the answer is something you can actually do | “Which of those matters most?” |
3. The job itself
| Ask | What you are listening for | The follow-up that matters |
|---|---|---|
| 9. Do you have what you need to do the job properly? | Equipment, cover, information — usually the cheapest thing to fix | “What is missing most often?” |
| 10. What part of your job uses what you are best at? | Whether the person is in the right seat | “How much of the week is that?” |
| 11. What part of the job did you not expect when you started? | The gap between the posting and the reality | “Would you have taken the job knowing it?” |
| 12. Is the workload what you expected, more, or less? | Chronic short-staffing that has become normal | “Which shift is worst?” |
| 13. What takes longer than it should? | Process friction the supervisor cannot see from where they sit | “Show me tomorrow.” |
4. The schedule, the pay and the people
Three subjects managers avoid in case they cannot fix them. Avoiding them does not make them stop being the reason people leave.
| Ask | What you are listening for | The follow-up that matters |
|---|---|---|
| 14. Does the schedule work for your life outside work? | The real driver of frontline turnover, more often than pay | “What one change would help most?” |
| 15. When you needed time off, how did that go? | Whether the written policy and the practice match | “Who did you ask, and what happened?” |
| 16. Do you feel fairly paid for what you do here? | Not the number — the comparison they are making | “Compared with where?” |
| 17. Who here makes your job easier, and who makes it harder? | Team dynamics you will not hear about any other way | “What would fix that?” |
5. Forward, and the question that opens everything
| Ask | What you are listening for | The follow-up that matters |
|---|---|---|
| 18. What would you like to be doing here in two years? | Whether they can see a next step, and whether it exists | “Has anyone talked to you about that?” |
| 19. What could I do differently as your manager? | Costs nothing, and the answer is usually specific and small | “Starting when?” |
| 20. What should I have asked you that I did not? | Everything the first nineteen missed | Write it down in front of them. |
Do not ask all twenty
Pick eight to twelve for a fifteen-minute conversation, always including questions 5, 9, 14 and 20. A forty-minute interrogation gets shorter answers than a fifteen-minute conversation does.
What to do with the answers
Within 48 hours, write three lines per person: the thing you will change, the thing you will find out about, and the thing you said no to. Then tell them what happened, by name and by date. The telling is the part that produces the effect — an employee who sees one small thing change because they mentioned it will raise the next one early, which is the entire point of the exercise.
Across a team, read the answers for clusters rather than incidents. One person mentioning the schedule is a preference. Five people mentioning the schedule is a staffing model. One person mentioning a colleague is friction; five people mentioning the same colleague is a supervision problem you have been paying for in turnover.
Two cautions. Do not build a scoring system out of this — the moment answers become data points with a number attached, the conversation becomes an assessment and the honesty goes. And do not promise what you will have to walk back; a straight no in the room is cheaper than a yes that quietly expires.
What we can and cannot claim
Most pages on this subject attach a percentage to stay interviews — a retention lift, a turnover reduction, a return on investment. Those figures generally come from vendor surveys of self-selected respondents, and we could not verify them, so they do not appear here.
What we can say is narrower and more useful. We have no measured evidence for the effect size of stay interviews, recognition programs, mentorship schemes or manager training. Stay interviews are cheap enough to be worth trying regardless — fifteen minutes, twice a year, no software, no budget line — and the mechanism is plausible: the reasons frontline employees give for leaving are mostly known to them in advance, and nobody is asking. That is an argument from logic, not from data, and we would rather label it as such than dress it up.
The same applies to the cost of replacing someone. There is a widely repeated figure expressed as a share of salary; its provenance is usually unstated and it varies by an order of magnitude across the roles it gets applied to. Our employee retention strategies page sets out what we can and cannot support in more detail, and its calculator asks for your own figure rather than supplying one.
When a stay interview is not the answer
Three situations where the conversation will not help and may do harm.
- When you are below market on pay and know it. Question 16 will produce an answer you cannot act on, and the conversation becomes a reminder. Fix the number or do not raise the subject.
- When the turnover is in the first ninety days. Early departures are usually a hiring and onboarding problem — the job was not what the posting said, or the first week was chaotic. That is fixed in the job description and the first-week plan, not in a conversation six months in.
- When you have run them before and changed nothing. Starting again without acknowledging that is worse than not starting. Say plainly what did not happen last time and what is different now, or wait until something is.
And one situation where it is exactly the answer: when you have a small number of people whose departure would genuinely hurt, and you do not currently know which of them is already looking. That is the case this conversation was built for.
Where the recruiting side comes in
Said against our own interest: retention work is cheaper than advertising. Across our 2026 benchmark of 891 managed campaigns the median cost per applicant was $13.88, and applicants are not hires — the median campaign needs several applicants to produce one. A fifteen-minute conversation that keeps one experienced person is worth more than the campaign that would have replaced them, and it is worth more again in the weeks of a supervisor's time that a vacancy consumes.
Where advertising earns its place is the part retention cannot fix: growth, retirements, a new site, and the departures that were always going to happen. If that is the position you are in, the guide to finding employees when no one is applying is the right starting point, and interview questions to ask candidates covers the conversation at the other end of the relationship.
Frequently asked questions
What is a stay interview?
A stay interview is a short structured conversation with a current employee about why they stay, what frustrates them and what would make them leave. It is normally run by the employee's direct supervisor, lasts fifteen to twenty minutes, happens once or twice a year, and is deliberately held off-cycle from performance reviews so the answers are not calibrated for a rating.
What questions should you ask in a stay interview?
Eight to twelve of the twenty on this page, always including four: what would make you leave, do you have what you need to do the job properly, does the schedule work for your life outside work, and what should I have asked you that I did not. The first establishes the real reason, the second is usually the cheapest thing to fix, the third is the most common driver of frontline turnover, and the fourth catches what the others missed.
How often should you run stay interviews?
Once or twice a year for established employees, plus at 30 and 90 days for new hires while the gap between the job advertised and the job delivered is still fresh. Also run one the month after a significant change — a new supervisor, a schedule change, a site move. Do not run one in the week somebody has resigned; at that point it is an exit interview.
What is the difference between a stay interview and an exit interview?
Timing, and therefore usefulness. An exit interview asks somebody who has already resigned why they left, when the decision is made and the answer is calibrated for the reference. A stay interview asks somebody who has not decided yet what would make them leave, while you can still act on the answer.
Who should conduct a stay interview?
The employee's direct supervisor, not HR. The supervisor is the variable most likely to be the answer, and the conversation is far less useful when that person is not in the room. The safeguard against the supervisor being the problem is a second route — somebody else people can raise things with — plus watching for answers that cluster across a whole team.
Do stay interviews actually reduce turnover?
We have no measured evidence of an effect size, and we are not going to borrow a percentage from a vendor survey. What can be said is that they are close to free — fifteen minutes, twice a year, no software — and that the reasons frontline employees give for leaving are usually known to them well in advance and nobody asks. That is an argument from mechanism rather than from data.
How long should a stay interview be?
Fifteen to twenty minutes. A longer session turns into an interrogation and produces shorter answers, not longer ones. Pick eight to twelve questions, leave silence after the ones that matter, and take notes in front of the person rather than afterwards.
What should you do after a stay interview?
Within 48 hours, record three things: one thing you will change, one thing you will go and find out about, and one thing you have said plainly you cannot change and why. Then go back and tell the person what happened. The telling is what produces the effect; a stay interview that changes nothing is worse than none at all, because it confirms that raising something is pointless.
Should stay interview answers be recorded or scored?
Write notes for your own three actions, but do not build a scoring system. As soon as the answers become data points attached to a number, the conversation becomes an assessment and the honesty disappears. Read across the team for clusters instead — five people naming the schedule is a staffing model, not five preferences.
Are stay interviews worth it for hourly and frontline staff?
Particularly so, because the two things that most often drive frontline turnover — the schedule and the supervisor — are both things an employee will tell you about if asked and rarely volunteer otherwise. The practical version for shift-based teams is a rolling schedule of two or three conversations a week until the team is covered, then start again.
Retention is cheaper. Advertising is for the gap it cannot close.
We put frontline jobs in front of people who are not job hunting, and report cost per applicant with the management fee inside it. Median across 891 managed campaigns: $13.88.
Book a DemoCost-per-applicant figures: Boostpoint 2026 Social Job Advertising Benchmark, 891 managed campaigns and 1,334 campaign-months. No effect-size claim is made for stay interviews; where we have no measured evidence, this page says so. A shorter version of these questions previously appeared on our employee retention strategies page.