Exit interviews: how to run them so they actually change something

Most companies run an exit interview exactly once per employee, on the way out the door, squeezed between a goodbye lunch and a laptop handoff. It produces a paragraph or two of generic answers and a polite thank you. Then it goes into a folder nobody reopens until someone building a report decides to count how many people cited "a better opportunity" as their reason for leaving.

That is a remarkable amount of wasted information. A departing employee has nothing left to protect. They are not angling for a raise, a promotion, or a good reference from the manager they are about to stop reporting to, which makes them one of the only people in the entire organization with no real incentive to soften the truth about what is actually broken.

Exit interviews done well are not a farewell ritual. They are a direct line into the parts of your hiring, management, and retention process that never show up anywhere else, and most companies are sitting on that information without ever using it.

Why most exit interviews produce nothing useful

The standard exit interview fails for a predictable, structural reason: it happens too late, with the wrong person asking, and no real expectation that anything will come of it. By the time someone sits down for the conversation, they have already accepted another offer, mentally moved on, and have every incentive to keep the exchange pleasant and vague. "Wanted a change" and "better opportunity" are safe, low-effort answers that end the conversation quickly, and most exit interviews are not structured well enough to get past them.

Who asks the questions matters just as much as when. An employee leaving because of a difficult manager is not going to volunteer that to someone who reports to the same manager, or to HR if they suspect HR will share it internally before they have cleared out their desk. SHRM's guidance on employee relations is consistent on this point: the credibility of the person conducting a sensitive conversation determines how honest the answers will be, and that holds just as true on the way out as it does in any other difficult workplace conversation.

The deeper issue is that most companies never close the loop. An employee who gives real, substantive feedback and then watches nothing change has taught every future departing employee the same lesson: it is not worth the effort. Onboarding is the last mile of hiring. The exit interview is the last mile of retention, and treating it as a formality throws away the one data point most companies collect specifically to find out what retention is actually missing.

What a good exit interview can tell you that nothing else can

Engagement surveys, pulse checks, and skip-level meetings all measure how current employees feel while they still have a stake in the answer being read generously. An exit interview measures something different: how someone who no longer has anything to lose actually experienced working there, described with enough distance to be honest but not so much that the details have faded.

That makes exit data uniquely useful for catching problems that never surface through normal channels: a manager who looks fine in a skip-level but is quietly driving out strong performers one at a time, a workload that was sustainable on paper and unsustainable in practice, a promise made during the offer stage that quietly stopped being true. The true cost of a bad hire gets calculated constantly. The cost of losing a good one to a problem nobody flagged in time rarely gets calculated at all, and exit interviews are one of the few places that cost becomes visible before it repeats.

Exit data is also one of the only sources of information that is directly comparable across an entire department, because roughly the same questions get asked to every departing employee. A single exit interview is an anecdote. A year of them, reviewed together, is a dataset that can point to a specific team, a specific manager, or a specific policy with a precision that an engagement survey, averaged across hundreds of people who are not leaving, simply cannot match.

Timing it right: before, at, and after the departure

Running the exit interview on the employee's literal last day is the single most common mistake, because it is also the moment they are least likely to say anything substantive. A better window is after the resignation is final but before the last week turns into a blur of handoffs and goodbyes. Scheduling the conversation for the first week of a two-week notice period, rather than the last day, consistently produces more detailed, more useful answers.

A second, underused window sits weeks or months after departure, once the new job has had time to become a real point of comparison rather than a hopeful guess. An employee three months into a new role can tell you, specifically, what their new employer does differently and better, in a way they genuinely could not have articulated on the way out. A short follow-up survey at that mark catches insight the original exit conversation structurally cannot.

This staggered approach also sets up a cleaner path back to the person later. Rehiring former employees works best when the relationship stayed warm after departure, and a thoughtful, well-timed exit process, followed by a genuine check-in later, is one of the most natural ways to keep that door open without it feeling transactional.

The questions that produce signal, and the ones that just produce small talk

"Why are you leaving" is the question every exit interview leads with, and it is almost worthless on its own, because it invites exactly the vague, diplomatic answer most departing employees have already rehearsed before they sit down. The questions that actually produce signal ask about specific moments and specific people rather than overall verdicts: what made you start looking, what would have changed your mind two months ago, what did your manager do well and what did they miss.

Comparative questions work especially well because they force a concrete answer instead of a vague one. Asking what the new role offers that this one did not gets a far more specific response than asking whether the person was happy here, because it requires naming an actual difference rather than rendering an overall judgment that is easy to soften on the way out the door.

Indeed's guidance on running effective exit interviews makes a point worth building the whole process around: open-ended questions paired with a few closed, ratable ones give you both the texture of a real story and a number you can actually track over time. That matters because the most common failure of exit data is collecting rich answers that nobody ever aggregates into something a team can act on.

Who should conduct it, and who should stay out of the room

The departing employee's direct manager should almost never conduct their own exit interview. It is the fastest way to guarantee a sanitized answer, because even an employee with a genuinely good relationship with their manager will soften criticism delivered directly to the person it is about, and an employee with a bad relationship will say almost nothing real at all.

A neutral party, HR, a skip-level manager, or in smaller companies a founder with no direct stake in the departing employee's day-to-day, consistently gets further. Employer branding depends heavily on what departing employees say about the company afterward, publicly and privately, and an exit conversation that felt genuinely safe and well-run is one of the more reliable ways to keep that word of mouth from turning sour on the way out.

For sensitive departures, performance-related exits, harassment complaints, or anything with legal exposure attached, the conversation should be handled carefully, often with HR and legal counsel involved together rather than left to whichever manager happens to be available. Getting this wrong does not just cost you data. It can turn a routine departure into a liability the company did not need to create.

Closing the loop: turning answers into action

An exit interview that produces a write-up nobody reads is functionally the same as not running one at all, except it costs the departing employee their honesty for nothing in return. The fix is procedural: someone needs to own reviewing exit data on a fixed schedule, quarterly at minimum, and that review needs to produce at least one concrete action a quarter, not just a slide summarizing themes.

The most useful version of this loop routes specific, recurring feedback back to the people who can actually act on it. A pattern of departures citing the same manager, the same unmet promise from the offer stage, or the same broken process should reach that manager or process owner directly, not get absorbed into a company-wide summary where the specific signal disappears into an aggregate nobody traces back to its source.

Recruiting metrics that actually matter usually stop at the point of hire, but the loop should not. Exit data is retention's version of the same discipline: a number that means nothing until someone is accountable for moving it. A company that reviews exit interviews without ever changing a policy, a manager's coaching plan, or a compensation band because of what it found is running a listening exercise, not a retention program.

Measuring whether your exit interview program is actually working

The honest test of an exit interview program is not how many employees complete one. It is whether the answers, tracked over time, actually predict and reduce future turnover. A program with a high completion rate and no measurable change in why people leave or how long they stay is producing paperwork, not insight.

The useful audit compares exit reasons against actual retention outcomes a year or two later. If the same reason, a specific manager, a specific team, a specific unmet expectation, keeps showing up in exit interviews and the turnover rate for that team has not moved, the data is being collected and ignored rather than acted on. Glassdoor's research on what drives employee turnover is a useful outside check here too, because what employees say anonymously in public reviews after leaving often tracks closely with what they were willing to say in a well-run internal exit interview, and a gap between the two is worth investigating on its own.

The programs that actually move retention treat exit data the way a strong sales team treats a lost-deal review: not as a formality to close out a file, but as the most honest feedback the organization is ever going to get, collected at the one moment people have no reason left to hold back. Most companies have that moment available every single week. Very few have built a process that actually listens to it.