Why Exit Interviews Don't Tell You Anything Useful

Your HR head forwards the exit interview summary two days after the last working day. You read it between meetings. Reason for leaving: better opportunity. Would you recommend the company to others: yes. Anything we could have done differently: nothing as such, sir — it was a personal decision. Overall experience: positive.

That is the fourth one this year, about four different people, and it has said the same thing every time. Four capable people walked out and the file records that nobody did anything wrong.

At some point you should stop reading that as good news. Four independent people did not all leave for a reason that has nothing to do with you. What you are looking at is not a finding. It is a genre.

The usual response is to improve the instrument — better questions, a neutral third party, an anonymous form. That is the wrong repair. The exit interview is not a flawed tool that sharper questions would fix. It is a perfectly reasonable tool pointed at the wrong stage of the sequence.

You are asking at stage five about a decision taken at stage two

A resignation is the last event in a chain that has been running for the better part of a year. We set that chain out in full in why your best people leave — five stages, roughly nine to fourteen months from beginning to end.

The decision does not happen in stage five. It happens in stage two, the recalculation, when a capable person quietly stops treating the job as something they are building and starts treating it as something they are doing. Everything after that is logistics.

The exit interview arrives at the very end of stage five, when the laptop is being handed back. It asks a person to explain a decision they finished making nine to fourteen months earlier, in a room where they have four things still pending with you, on a form that will sit in their personnel file.

Every part of that sentence is a problem.

The Wrong Instrument

Four failures, and they compound. Fixing one does not help, because the other three still hold.

Wrong time

You are asking about a decision made a year ago.

Assume total honesty — a person who wants to help you and has nothing to gain or lose. They still cannot give you an accurate answer, because memory does not store decisions, it stores outcomes and builds a reason backwards from them. By the time someone has resigned, negotiated a package, told their parents and explained the move to eleven people at a farewell lunch, the account has been rehearsed into a clean shape. Better role. More responsibility. Right time.

That account is not a lie. It is a summary written after the fact, smoothed by every retelling, and it has written over the thing you actually need — the specific week eleven months ago when something shifted.

You are not collecting a cause. You are collecting a conclusion.

Wrong incentive

In India this is sharper than almost anywhere, and it has nothing to do with character.

At the moment you ask the question, the person needs several things from you. A relieving letter. A full-and-final settlement that lands correctly and on time. A clean experience letter with the right designation on it. A reference call handled well when the new employer's background check firm rings in three weeks. And beyond all of that, a peaceful exit from an industry where the senior circle is small, everyone has worked with everyone, and one WhatsApp group can carry a reputation across four companies in an afternoon.

Now weigh that against the benefit of telling you the truth. The truth costs them something real and immediate and gains them nothing — whatever improvement it produces will be enjoyed by people who are not them, at a company they are leaving.

So they are polite. They say better opportunity, they thank you sincerely, and they go.

Do not read that as cowardice, and do not let anyone in your business moralise about it. It is correct behaviour. A rational adult with a family, a home loan and a notice period does not hand a departing employer a candid critique in exchange for nothing. If your process depends on people acting against their own interest to work, your process does not work.

Wrong person asking

HR conducts the exit interview. HR cannot change the thing that caused the exit — the reporting line, the scope, the role that had no next step, the promise made in a review and never revisited. Both people in the room know this, which quietly converts the conversation into paperwork. The employee gives answers that will close the file. HR records them so the file closes.

The obvious fix is for you to conduct it yourself. That makes it worse. The moment the promoter is in the room, the cost of honesty goes up — the person is now being candid to the one individual whose good opinion determines the reference, the settlement and whether the door stays open. Seniority does not create candour. It suppresses it.

Wrong question

"Why are you leaving" asks for a reason. A reason is a single, tidy, portable thing, which is exactly what a person under mild pressure will happily supply.

But the useful information is not a reason. It is a sequence. What changed, and when. Which ask went unanswered, in which quarter, and what the person concluded from the silence. How long they carried on after that before the first recruiter call was returned.

Nobody asks for the sequence, so nobody learns it — which is why exit interviews across an entire industry produce the same four answers and no company ever improves because of one.

The data is not useless. It is actively misleading.

This is the part that costs money.

If the exit interview merely produced nothing, you would ignore it. What it produces is a systematically skewed record. It over-reports the socially safe causes — better opportunity, growth, family reasons, salary — because those cost the departing person nothing to give. It under-reports the real ones: there was no path forward here, a specific commitment made to me was not honoured, a decision I was given was taken back six weeks later in a meeting I was not in. Saying "there was nowhere for me to go here" implies a judgement of your business and exposes the person's own ambition to someone who could later call it arrogance.

Which brings you to the mismatch that matters. Culture Amp's analysis of 175 teams across hundreds of fast-growing companies found that development opportunities drove 52% of departures, leadership 28%, the direct manager 12% and pay 11%. The largest single cause of people leaving is having nowhere to go — and it is close to the least-reported answer on any exit form, for exactly the reasons above.

So consider what happens to a promoter who runs his retention strategy off exit interview themes. His forms say money and better opportunity. He responds with a salary correction, a revised increment policy, an engagement initiative, a culture offsite. He spends real rupees and changes nothing structural — because the structural thing, that four of his senior roles have no next step attached to them and everybody in those roles has worked it out, never appeared in the data he was reading.

This is also how the two most confident beliefs in this category survive. Exit forms say pay, so pay looks like the cause, which we test in do employees really leave for money. Exit forms say the manager when they say anything critical at all, which keeps a badly compressed maxim alive — we trace it in do people leave managers or companies.

How to test it

Ask HR for the last ten exit interview forms. Read only the reason field. Count how many of the ten named a cause inside your control — a role with no next step, a promise you did not honour, a specific manager, a decision that was reversed, a scope that never widened.

Not "better opportunity". Not "personal reasons". Not "relocating". Those are exits from the conversation, not causes.

If your count is zero or one out of ten, hold that against Gallup's finding that 52% of voluntarily exiting employees say their manager or organisation could have done something to prevent them from leaving. Just over half of the people who left you would have stayed, and something specific would have held them.

Both of those cannot be true. Either Gallup is wrong about your business specifically, or your form is not detecting the thing it exists to detect. You already know which.

What to do instead

Not a better exit interview. Different conversations, at different times, with different incentives attached.

Move the question earlier

The conversation that produces real information happens with people who are staying, and it is not an engagement survey. Surveys ask people to rate statements. You need people to name things.

Two questions, asked by you and not by HR, twice a year, in a fifteen-minute conversation with each person you would be genuinely damaged to lose.

• "What have you asked me for that hasn't happened?" Not what do you want — what did you already ask for. This directly interrogates stage one of the Exit Clock. If the answer is "nothing, sir, everything is fine", that is not comfort. People who have stopped asking have usually stopped for a reason.
• "What is the next thing for you here — and does it actually exist?" The second half is the honest half. Sometimes the correct answer is that it does not exist yet, or will not for two years. Saying that out loud is uncomfortable, but a person who knows the real timeline can plan around it. A person left guessing assumes the worst and starts looking.

Both questions are unpleasant to ask. That is precisely why they work — a question that is comfortable to ask is usually one whose answer you already have.

Ask the sequence, not the reason

If you do run an exit conversation, replace the reason question with a timing question: when did you first think about leaving, and what was happening around then?

That single question does more than the whole standard form. It moves the person from justifying a decision to remembering a period, which is easier to answer honestly and much harder to answer from a rehearsed script. You will often get a month, a project, a meeting or a specific reversal — the stage-two event you have been trying to find.

The ninety-day call

Three months after someone has left, ring them yourself.

The timing is the whole point. By then they have been paid. The relieving letter is issued, the full-and-final is settled, the reference check is done and the new job is real. Nothing is outstanding between you in either direction, so the incentive structure that made honesty expensive on their last day has dissolved.

Three rules make it work. You make the call, not HR. You say at the start that it is not a re-hiring approach, and you mean it — the moment it becomes one, the honesty ends. And you ask about the sequence, not the reason: when did it start, and what was going on then.

People who gave you eight polite words in the exit interview will give you twenty candid minutes on that call. It is the cheapest source of retention information available to an Indian promoter, and almost nobody uses it.

Read the pattern, not the instance

Each individual account may be unreliable. The distribution is not.

Departures cluster — around particular managers, particular roles, and particular points in tenure, such as the eighteen-month mark or the year after a promotion that did not come. Plot your last two years of exits on one sheet: name, manager, role, months of tenure, month of resignation. You are not looking for reasons. You are looking for a shape.

If four of seven exits report to the same person, you have learned something no individual form told you, and it is worth reading alongside why your managers leave. If they cluster at a tenure point, you have a structural ceiling rather than a people problem. Either way the cluster tells you which team to watch before the next resignation rather than after it, which is where the flight risk indicators become readable.

Where to start

Do the ten-form count this week. It takes twenty minutes and it will settle the argument about whether your exit process is worth running.

Then pick the three people whose resignation would genuinely hurt and ask them the two questions — with the door shut, and without defending yourself in the room. What holds people who have options is a separate discipline, set out in how to retain your key people.

If you want the structural version — which roles in your business have no next step attached to them, and where progression quietly stops — the Business Pulse diagnostic maps it. That is the answer your exit interviews were never going to give you, because it is the one nobody says out loud on their last day.