Why most exit interviews produce nothing
The format is near-universal and the yield is close to zero in most organisations. Four reasons, and they compound.
- The interviewer. Conducted by HR who report into the same leadership the complaint would be about, or worse by the manager the person is leaving. Candour is not a reasonable expectation.
- The timing. Held while the final settlement, the relieving letter and a future reference are all still pending. The leaver has every incentive to say the new role was simply a good opportunity.
- The questions. Satisfaction scales and generic prompts produce generic answers. Nobody discloses a specific problem in response to a rating question.
- The output. Filed against the employee record and never aggregated, so even a candid interview changes nothing and the next person's experience is identical.
The result is a process that consumes time, generates a document, and tells the organisation what it already believed. That is worse than not running them, because it creates the impression that the question has been asked and answered.
What to ask instead
The useful questions are specific, about events rather than feelings, and oriented to the decision rather than to satisfaction.
- When did you first start thinking about leaving, and what prompted it? This is the single most valuable question, because the answer is usually months before the resignation and identifies the actual trigger.
- What would have changed your mind at that point? Note the emphasis on that point rather than now, which is usually too late.
- What did you tell your manager at the time, and what happened? Frequently the answer is that they raised it and nothing followed.
- What is the new role offering that this one did not? Concrete comparison rather than sentiment.
- What surprised you about the job after you joined, in either direction? This tests the recruitment and onboarding promise against the reality.
- What should we ask the next person that we have not asked you?
Two questions to avoid: would you recommend us, which yields a polite number, and is there anything else, which yields nothing at the end of a conversation that has not already built trust.
Design choices that change the yield
| Choice | Default | Better |
|---|---|---|
| Who conducts | HR business partner for that team | Someone outside the leaver's management chain, or a third party |
| When | Last week, settlement pending | After the settlement is agreed, or a follow-up some weeks after leaving |
| Format | Form, then a conversation | Conversation first, notes after, form only if it aggregates |
| Confidentiality | Unstated | Stated explicitly: what will be shared, with whom, in what form |
| Output | Filed | Themed and reported in aggregate, with the manager dimension visible |
The follow-up some weeks after leaving is the most underused of these. A former employee who has started the new job, been paid by someone else and has nothing pending is markedly more candid than one still waiting for a settlement.
Turning it into a signal
Individual exit interviews resolve nothing. The pattern across them is the management product, and producing it requires only that the notes be themed consistently.
- Code each exit against a small set of causes: manager, progression, pay, workload, role clarity, relocation, external opportunity. Resist a long taxonomy, which prevents patterns emerging.
- Report by team and by manager, not only organisation-wide. Attrition concentrates, and an organisation-wide average conceals exactly the concentration that matters.
- Track time from the stated trigger to the resignation. A consistently long gap means there was a window in which intervention was possible.
- Compare against what was said at recruitment. Where leavers repeatedly cite something as a surprise, the promise made at hiring is wrong.
- Report regretted and non-regretted separately. They are different problems and mixing them makes both invisible.
None of this requires a system beyond a spreadsheet, and most organisations still do not do it, which is why the same causes recur for years.
The limits of exit data
Exit interviews are a lagging indicator by construction. By the time someone is answering, the decision is made, the replacement cost is incurred, and the causes they describe have been operating for months.
Two consequences follow.
- Pair exit data with something that surfaces the same causes while people are still present: skip-level conversations, engagement surveys read by team, or simply asking managers whose people are at risk. Exit data validates those signals; it should not be the first place a problem appears.
- Do not over-fit to the articulate. Leavers who give a clear, structured account are more memorable than those who say little, and their reasons are not necessarily representative.
There is also a category the process misses entirely: the people who considered leaving, were retained, and whose reasons never surface because they never resigned. Those are the interviews most worth having and almost nobody conducts them.
Frequently asked questions
What is an exit interview?
A conversation with a departing employee about why they are leaving and what their experience was. Its purpose is to surface causes the organisation can act on, which requires it to be conducted and used quite differently from how most organisations run it.
Are exit interviews mandatory?
No statute requires one. They are entirely a management practice, which also means there is no obligation on the departing employee to participate or to be candid if they do.
Who should conduct an exit interview?
Someone outside the leaver's management chain, since the most common cause of leaving is the manager. An interview conducted by the person being complained about, or by someone reporting to them, produces polite answers.
When is the best time to do it?
After the final settlement is agreed rather than while it is pending, and ideally with a follow-up some weeks after the person has left. A former employee with nothing outstanding is markedly more candid than one still waiting to be paid.
What is the most useful question to ask?
When did you first start thinking about leaving, and what prompted it. The answer is usually months before the resignation, identifies the real trigger, and tells you there was a window in which something could have been done.
How Engage handles exits
Engage keeps exit responses against a consistent set of themes rather than as free text on a form, so the pattern by team and by manager is visible instead of being buried in individual records. Because exits, attrition and tenure sit in the same system, the exit reason can be read alongside how long people stayed and where they were, which is what turns individual conversations into something a leadership team can act on.
See people analytics in Engage