Exit Interview Process, Questions and Template for India
A complete exit interview playbook for Indian SMBs: who should ask, a themed question bank, a copy-paste form template, and how to turn exit data into retention action.
The exit interview, and the three jobs it actually does
Every HR team in India runs some version of an exit interview. Very few get anything useful out of it. The resignation lands, the notice period starts, someone books thirty minutes in the final week, asks eight questions off a form that has not changed in years, writes "better opportunity" in the reason field, and files it. Three months later the same manager loses another engineer for the same unspoken reason, and nobody connects the two.
This guide is about fixing that — designing the conversation for candour, then building the unglamorous back-end that turns candour into retention action.
One warning first. If you are not prepared to change anything based on what you hear, do not run exit interviews. An unactioned exit interview is worse than none, because it teaches the leaver — and everyone they talk to — that feedback here is theatre.
An exit interview does three different jobs at once, and most teams never separate them.
Diagnosis. Why did this person leave? Only useful if it is specific enough to act on. "Lack of growth" is not a diagnosis. "I asked for a defined path to senior twice in eighteen months, my manager said he would check with the founder, and I never heard back" is — it names a broken process you can fix on Monday.
Closure. The employee deserves a dignified place to put down a pile of unexpressed frustration. This matters more for the people staying than for the person leaving: freeze someone out of meetings the day they resign, and the remaining team learns that loyalty here is conditional.
The alumni relationship. The person walking out is a potential rehire, a referral source, and — critically in India — the subject of a background verification call in two years. Nobody boomerangs to a company that made their exit miserable.
Why most exit interviews produce polite, useless answers
You ask why they are leaving. "Better opportunity, growth reasons." You ask about their manager. "No, no, manager was very supportive." Everyone smiles. Nothing is learned.
This is not dishonesty. In that moment, candour has no upside and several plausible downsides.
- They are protecting the reference. Verification calls in India often go to the last manager or HR contact, sometimes informally.
- The relieving letter is pending. Asking for candour while holding documents they need is a structural conflict, and they read it correctly.
- HR is not neutral. Employees know that "confidential" often means "unattributed, but your manager will guess in four seconds."
- They have checked out. By the last week the decision is old news and frustration has cooled into indifference.
- The questions invite generic answers. "Any feedback for us?" invites "no, everything was good."
- There is no closed loop. They have given honest exit feedback before and watched nothing change.
Designing for candour
You cannot remove these pressures. You can reduce them.
- Decouple the paperwork. Confirm in writing, early, that the relieving letter, experience certificate and settlement are process-driven and not contingent on this conversation. Then behave that way.
- Move it earlier — mid-notice, not the last two days — and make sure whoever signs off the settlement is not the one asking why they left.
- Ask about the system, not the people. "What made your work harder than it needed to be?" beats "did you have problems with your manager?"
- Anchor on timelines. "Take me back to the first time you seriously thought about leaving — roughly when, and what was happening?" This is the highest-yield question in the bank.
- Promise something narrow and keep it. Say exactly what gets shared, with whom, in what form.
- Show the loop. Name a change that came from a previous exit conversation.
- Ask again later. An anonymous alumni pulse survey 60 to 90 days after the last working day often beats anything said inside the building.
Where the exit interview sits in the offboarding timeline
A conversation held at the wrong point cannot succeed no matter how good the questions are. Here is the flow for an Indian SMB with a 30 to 90 day notice period.
| Stage | Timing | Owner | What must happen |
|---|---|---|---|
| Resignation and manager conversation | Day 0–2 | Employee → manager | Written, dated, on record. Understand the reason; decide if a counter-offer is even on the table. |
| Retention attempt (optional) | Day 1–5 | Manager + skip-level | Decide fast. A dragged-out counter-offer poisons the notice period. |
| Formal acceptance and announcement | Day 3–10 | HR + manager | Acceptance letter, last working day, notice or buyout in writing. Team told before the grapevine. |
| Knowledge transfer | Day 7 → LWD−7 | Manager + employee | KT checklist, named receivers, deadlines, credential inventory. |
| Exit interview | 60–70% through notice | Skip-level or HRBP | Structured conversation; survey sent 2–3 days earlier. |
| Asset and access recovery | LWD−5 → LWD | IT + admin | Laptop, ID card, SIM, licences; access revoked on LWD. |
| Last working day | LWD | Manager + HR | Farewell; relieving letter and experience certificate handed over. |
| Full and final settlement | LWD + 30–45 days | Payroll | Final salary, leave encashment, gratuity if eligible, recoveries, Form 16 timing. |
| PF and statutory records | LWD + 30–60 days | Payroll | EPFO exit date updated, UAN shared, ESIC updated. |
| Alumni follow-up | LWD + 60–90 days | HR | Anonymous pulse survey, alumni group, boomerang flag. |
Too early in that sequence and the person is still in negotiation mode, reluctant to burn a bridge they may still be standing on. Too late and they have mentally left. Mid-notice finds someone who has settled into the decision, still remembers the detail, and has enough runway that a fix could still be attempted.
One sequencing rule: agree the KT plan before the exit interview, or the employee will suspect the conversation is a covert attempt to extract handover information.
Who should conduct the exit interview
This matters more than the question list. The same question gets a different answer depending on who asks.
| Interviewer | Pros | Cons | Best used when |
|---|---|---|---|
| Direct manager | Knows context, can probe specifics | Often the reason for the exit; nobody criticises them to their face | Almost never as the official interview. Fine as a separate goodbye. |
| Skip-level manager | Senior enough to act, distant enough for candour about the manager | Can intimidate; may defend their own org | Strong default for most SMBs. |
| HRBP or HR manager | Trained, consistent, owns the data and follow-up | Seen as management's agent; often holds the paperwork | The workhorse — if HR has visibly acted before and does not control the settlement. |
| Founder or CXO | Big signal of respect; can authorise a fix in the room | Not scalable; power gap suppresses honesty | Senior exits and regretted attrition. |
| Neutral third party | Highest candour; no reference risk | Costs money; lacks internal context | Sensitive exits; a deep dive when attrition spikes inexplicably. |
| Anonymous survey only | Cheap, scalable, good for ratings and sensitive topics | No probing; anonymity is fragile in small teams | High-volume, seasonal or field roles; or half of a hybrid. |
A practical default for a 20 to 200 person company:
- Standard exits: anonymous survey, then a 40-minute conversation with the HRBP.
- Regretted exits: survey, then the skip-level, plus a short founder conversation if senior.
- Manager suspected as a factor: skip-level conducts it, and the manager is not in the scheduling loop.
- Second or third exit under the same manager in a quarter: escalate the pattern to the quarterly review regardless of what the individual says.
One rule almost nobody follows: whoever conducts the exit interview should not be whoever signs the settlement or the relieving letter.
Format and timing: survey, conversation or both
The written form gives consistent, aggregatable data and works for distributed teams, but cannot probe. "Rated manager support 3/5" tells you nothing about why.
The live conversation lets you follow the thread — the most valuable sentence is almost always the answer to a follow-up, not the first answer. But it is inconsistent across interviewers and hard to aggregate.
The hybrid is the recommendation. Send the survey two to three days ahead, read it before you walk in, and use the conversation to probe the low ratings. The employee has already committed something to writing, you get structured data and depth, and the pre-work signals that this is serious.
Logistics: 40 to 60 minutes, mid-morning, private and neutral — not the manager's cabin, not a glass room the team can watch. One interviewer; two feels like a tribunal. Ask before recording. Take notes in their words, not your interpretation.
How to open the conversation
"Thanks for making time. This is not a retention attempt — your decision is accepted. It has no bearing on your relieving letter, experience certificate or settlement; all of that is in motion and I will confirm dates before you leave today. What I want is your honest read on what it was like to work here, especially the parts that were harder than they needed to be. I would rather hear something uncomfortable than something polite. I write up themes, not quotes, and share them with leadership. Nothing is attributed to you by name without asking first. If something needs a formal process — harassment, safety, misconduct — I will stop and tell you, because that has a separate route. One more thing: someone in this chair told us last quarter that our onboarding buddy system was not working, and we rebuilt it. So this does go somewhere."
That removes the retention agenda, decouples the paperwork, licenses discomfort, states confidentiality honestly, pre-flags escalation, and proves the loop closes — in about ninety seconds.
The exit interview question bank
Pick 12 to 18 of these, not all. Good exit interviews cover fewer topics in more depth. Arrows mark the probing follow-ups, which is where the information actually lives.
1. Reason for leaving
- What made you start looking? Not the final trigger — the first moment you thought "maybe I should see what else is out there." → Roughly when, and what was happening at work?
- If one thing had been different, you would still be here. What is it? → Did you raise it? What happened?
- Push or pull — were you leaving something, or going toward something?
Skip the long version for higher studies, relocation or family medical reasons. Interrogating a genuinely external cause wastes goodwill and yields nothing.
2. The role and the work itself
- How closely did the job match what was described in hiring? → What was different, and when did you notice?
- What part of the work did you find most engaging? What did you dread?
- Did you have what you needed: information, access, decisions, budget, people?
3. Manager and feedback
Frame these around behaviour and process, not personality. You want observable facts, not character verdicts.
- How often did you have a one-on-one, and how useful were they? → Status updates, or your development? → How often were they cancelled?
- When you did something well, how did you find out? And when something needed to improve? → Timely, or months later in a review cycle?
- If you raised a problem with your manager, what usually happened next?
- What one change in how you were managed would have made the biggest difference?
Soften or skip if the employee is visibly uncomfortable. Do not push someone into naming their manager as the villain.
4. Growth and career
- When you joined, what did you expect the next two years to look like? → How did that compare?
- Did you have a clear sense of the next level and what it would take? → Written down, or just talked about?
- Was there a role here you wanted and did not get? → Did you ask? What happened?
5. Compensation and benefits
Probe past this one — compensation is the socially safest reason to give, because it requires criticising nobody.
- How did you feel about your pay relative to the work? And did you understand how increments were decided?
- Was total compensation explained clearly, including variable pay, PF and gratuity? → Did take-home match what you expected?
- If compensation had matched your new offer, would you have stayed? → The key probe. A firm yes points at a pay-band problem. A pause means money was the excuse. Follow the pause.
6. Team and culture
- How would you describe the culture to a friend considering joining? → And to a close friend, off the record?
- How were disagreements handled on your team?
- Did you see people treated differently for reasons that were not about performance?
- Who made your job easier? We would like to recognise them.
7. Workload and wellbeing
- What did a typical week look like in hours? → And a bad week? How often?
- Could you disconnect evenings and weekends? → Expectations about responding after hours, written or unwritten?
- Did you take your leave? If not, why not?
- If you had said you were overloaded, what do you think would have happened?
8. Tools and process friction
Chronically underused, and often the fastest source of fixable wins.
- What took far longer than it should have because of a process or a tool?
- How was getting approvals — leave, expenses, purchases, hiring, deployment?
- What one process fix would make your old team meaningfully faster?
9. The new offer
Be curious, not competitive. This is free market intelligence, and people share it happily.
- What is the new company offering that we were not? What sold you, beyond the offer letter?
- How did they run their hiring process? How long from first conversation to offer?
Do not ask for the exact CTC unless they volunteer it. Ask about the band.
10. Boomerang and referral
- Under what circumstances would you consider coming back? → What would need to be different?
- Would you recommend this company to someone in your network? → For whom would it be a bad fit?
- Is there anyone in your network we should be talking to?
Ask that last one every time. Leavers on good terms are often generous with referrals.
11. Closing questions
- What should we keep doing? What should we stop?
- Is there anything I did not ask that I should have?
- Is there anything you want to say that you have not said yet?
That middle question is often the most productive in the bank — people use it to raise the thing they have been circling for forty minutes.
Questions to avoid, and why
| Avoid | Why | Ask instead |
|---|---|---|
| "Was it because of your manager?" | Leading; forces a verdict on a person | "What change in how you were managed would have made the biggest difference?" |
| "Who on the team was difficult to work with?" | Invites naming individuals unnecessarily; creates a record you may have to act on | "Were there working relationships that made the job harder? What was happening structurally?" |
| "Don't you think you're making a mistake?" | A retention attempt in disguise; destroys the frame | Nothing. That belongs earlier, with the manager. |
| "How much are they paying you?" | Intrusive; reads as counter-offer prep | "Was the compensation band meaningfully different?" |
| "Are you leaving for personal or family reasons?" | Strays into health, marital status, pregnancy | "Is there anything about the timing we should understand?" |
| Anything on caste, religion, community or family planning | Legally risky and inappropriate | Nothing. No version of this belongs here. |
Two more rules: do not relitigate a performance rating or a PIP here, and do not use an exit interview to gather evidence against a current employee. If you need to investigate someone, run an investigation.
A copy-pasteable exit interview form template
Send this two to three days before the conversation. Adapt it, but resist making it longer.
---
EXIT INTERVIEW FORM
Responses will be summarised as themes for leadership. Nothing will be attributed to you by name without asking you first. Completing this form has no bearing on your relieving letter, experience certificate or full and final settlement.
Section A — Basic details (filled by HR)
| Field | Entry |
|---|---|
| Name / Employee ID | |
| Department, designation, manager, skip-level | |
| Joining date / resignation date / last working day / tenure | |
| Exit type | Voluntary / Involuntary / End of contract |
| Regretted attrition | Yes / No / Partial |
| Interviewer, date, format |
Section B — Primary reason for leaving (select one; add a secondary if relevant)
☐ Compensation ☐ Career growth ☐ Learning and development ☐ Relationship with manager ☐ Team or culture ☐ Workload or burnout ☐ Role mismatch ☐ Company direction ☐ Commute or location ☐ Work arrangement ☐ Relocation ☐ Higher studies ☐ Health or family ☐ Own venture ☐ Career change ☐ Other: ______
- In your own words, why are you leaving?
- When did you first start considering leaving, and what was happening at the time?
Section C — Ratings (1 = strongly disagree, 5 = strongly agree, NA if not applicable)
| Statement | 1 | 2 | 3 | 4 | 5 | NA |
|---|---|---|---|---|---|---|
| The role matched what was described during hiring | ||||||
| I had the tools, access and information to do my job well | ||||||
| My manager gave me useful, timely feedback and backed my development | ||||||
| I could raise concerns without worrying about consequences | ||||||
| I had a clear picture of my growth path here | ||||||
| I was recognised fairly for good work | ||||||
| My compensation was fair, and I understood how increments were decided | ||||||
| My workload was sustainable and I could take leave when needed | ||||||
| I could disconnect outside working hours | ||||||
| I felt I belonged, and people here are treated with respect | ||||||
| Internal processes helped rather than slowed my work | ||||||
| Onboarding set me up well in my first 30 days | ||||||
| I would recommend this company as a place to work |
Section D — Open questions
- What one thing, had it been different, might have kept you here?
- What should we keep doing? What should we stop?
- What made your work harder than it needed to be — any process, tool or approval that was consistently painful?
- Was there anything you raised that went unaddressed?
- Who here made your work better? We would like to recognise them.
- Anything else you would like us to know?
Section E — Alumni
Would you consider returning? Yes / Maybe / No — if maybe, what would need to be different? ____ Would you refer someone? Yes / No · Join the alumni group? Yes / No · Personal email: ____
Section F — Sensitive matters
If you experienced or witnessed harassment, discrimination, retaliation, an unsafe condition, or financial or ethical irregularity, tick here: ☐ — these are handled through a separate formal process, not this form. The appropriate committee will contact you confidentially. You may also write to: [escalation email].
Section G — Interviewer notes (internal only)
Quotes worth keeping · Themes · Primary and secondary reason codes · Regretted: Y/N/Partial · Manager-related flag · Preventable: Y/N/Partial, and why · Recommended actions · Owner assigned · Escalation route if any · Rehire recommendation
---
India-specific practicalities
- Notice period and buyout tension. Notice of 30, 60 or 90 days rarely aligns with the new joining date, so a waiver or buyout is often still being negotiated when the exit interview is scheduled. An employee hoping for a 15-day waiver will not tell you the founder's decision-making is chaotic. Settle it in writing first — a decided no beats an open question.
- Relieving letter and experience certificate. Many employers will not confirm onboarding without them, so the perception that they can be withheld is a permanent drag on candour. Publish an offboarding SLA stating they are issued on the last working day subject only to asset return and completed KT, and hit it every time. This single piece of hygiene does more for exit interview quality than any question redesign.
- Full and final settlement timing. Settlement typically lands 30 to 45 days after the last working day — final salary, leave encashment, gratuity where tenure qualifies, notice adjustments, recoveries. Give a written breakdown with a committed date before the last working day, showing the computation, not just the number.
- Background verification calls, later. Have a written reference policy and share it: "We confirm designation, tenure and last drawn CTC. We do not offer subjective commentary."
- Counter-offer dynamics. Decide within the first week. If a counter-offer is warranted, make it once, make it real, make it fast. Never make one during the exit interview — it invalidates the frame and teaches everyone that resigning is the only route to a raise.
- Higher studies, relocation and family reasons. Code these as non-preventable and keep the conversation short and warm. But ask one question: "Was there anything about this job that made the decision easier?" Sometimes the plan was accelerated by a bad quarter. Set a reminder for when a two-year programme ends.
- Contractors, field and plant teams. Rarely tracked, and often where attrition is highest. Run a 15-minute version on scope clarity, process friction and payment timeliness; use a mobile-friendly form in the local language for distributed staff. Keep the data in a separate bucket.
- Statutory close-out. EPFO exit date updated, UAN shared, gratuity computed, ESIC updated, Form 16 timing communicated, and any bond or buyout recovery documented with the calculation shown. None of this is the exit interview's job, but all of it decides whether it goes well.
Handling sensitive disclosures
Sometimes the exit interview stops being an exit interview. Triggers: sexual harassment or anything under your POSH obligations, discrimination, bullying or retaliation, safety hazards, fraud, data or IP theft, any criminal matter.
- Stop, and say so. "I want to pause. What you have described needs a formal process, and I want to make sure it gets one."
- Do not investigate. No probing, no corroborating detail, no asking them to name witnesses. Improvised questioning can compromise a later inquiry.
- Explain the route — the Internal Committee for POSH matters, your whistleblower channel or the board for financial irregularity, the responsible function immediately for safety.
- Ask what they want, but be honest that for some categories, particularly POSH, the organisation has obligations independent of the complainant's preference. Do not promise to bury it.
- Document facts only — date, time, their words. Store it separately with restricted access and route it the same day, not after the last working day.
Do not tell the manager, especially if the manager is the subject. Do not let it affect the settlement or relieving letter in either direction. And do not treat a disclosure as less credible because it arrived at exit — people speak up when they no longer have anything to lose. If such disclosures consistently surface at exit and never before, your internal channels are not trusted, and that is a finding in its own right.
Confidentiality: what actually gets shared
The most common failure is a vague promise of confidentiality broken within a week. Someone says something specific, HR relays it, the manager works out who said it, and word spreads that exit interviews are not safe. Be precise instead.
| What is shared | With whom | Detail level | Timing |
|---|---|---|---|
| Reason code and regretted flag | HR, leadership | Coded data only | At exit |
| Rating scores | Leadership dashboard | Aggregated where team size allows | Monthly |
| Process and tooling feedback | Function owner | Specific; attribution asked for | Within 2 weeks |
| Manager feedback | Skip-level first | Themes only, never verbatim, never attributed | Quarterly, or immediately if serious |
| Praise and recognition | The named individual and their manager | Attributed — and worth attributing | Within a week |
| Sensitive disclosures | Designated committee only | Per the formal process | Same day |
In a team of four, "themes only, unattributed" is a fiction. Two honest options: tell the employee the truth — "your manager will likely know this came from you, do you still want me to pass it on?" — and let them decide, or hold the feedback and batch it with input from other exits and stay interviews.
When you do share manager feedback, deliver it one-on-one, lead with what the leaver said the manager did well, present themes across several exits rather than single-source criticism, and agree one specific change with a date. A manager who feels ambushed spends the quarter hunting for the source instead of fixing the problem.
Turning exit data into retention action
This is where most programmes fall apart: decent conversations, no mechanism to convert them into decisions.
Step 1: Build a reason-code taxonomy
Free text cannot be aggregated. You need a small, fixed, mutually exclusive set of codes — a taxonomy with forty codes gets applied inconsistently and tells you nothing.
| Code | Category | Preventable? | Typical owner |
|---|---|---|---|
| C1 | Compensation — base salary | Yes | HR + Finance |
| C2 | Compensation — structure (variable, ESOP, benefits, CTC clarity) | Yes | HR |
| G1 | Growth — no path to next level, or promotion denied | Yes | Manager + HR |
| G2 | Growth — work stopped being developmental | Yes | Manager |
| G3 | Growth — internal mobility blocked | Yes | HR |
| M1 | Manager — lack of support or availability | Yes | Skip-level |
| M2 | Manager — behaviour, respect, favouritism | Yes | Skip-level + HR |
| M3 | Manager — no feedback or recognition | Yes | Skip-level |
| R1 | Role — expectation mismatch versus hiring | Yes | Hiring manager + TA |
| R2 | Role — unclear scope, shifting priorities | Yes | Manager |
| W1 | Workload — hours, on-call load, burnout | Yes | Manager + HR |
| T1 | Team and culture — dynamics, exclusion, collaboration | Yes | Manager + HR |
| L1 | Leadership, strategy or stability concerns | Partially | Leadership |
| P1 | Process and tools — friction, bureaucracy, missing systems | Yes | Function owner |
| WA1 | Work arrangement or commute | Partially | Leadership |
| N1–N5 | Non-preventable — relocation, studies, health, career change, own venture | No | — |
| I1–I3 | Involuntary — performance, restructuring, contract end | N/A | Manager / HR |
Step 2: Code every exit consistently
- One primary code and at most one secondary. Not three, not five.
- Code from the whole conversation, not the dropdown they ticked. If they selected "compensation" but the story was eighteen months of a stalled promotion, the primary code is G1 with C1 secondary.
- Have the same person code all exits in a quarter. Consistency beats accuracy — an imperfect taxonomy applied consistently reveals trends; a perfect one applied inconsistently reveals nothing.
- Record one line of justification per code, in the employee's own words, and never default to "other."
Step 3: Separate regretted from non-regretted attrition
| Classification | Definition | Response |
|---|---|---|
| Regretted | You wanted to keep them — strong performer, hard to replace, critical knowledge | Root-cause every single one |
| Partially regretted | Good in the role but not a long-term fit | Track for patterns |
| Non-regretted | Performance or fit issues; would not rehire | Feed back to hiring, not retention |
| Non-preventable | Reasons outside your influence | Exclude from preventable-attrition metrics |
Decide the classification before the exit interview, using the last performance review and the manager's honest input. Deciding afterwards invites rationalisation — it is very easy to relabel a painful loss as "not really regretted" once you have heard the reasons. Then track two headline numbers: voluntary attrition, and regretted voluntary attrition. The second belongs on the leadership dashboard.
Step 4: Cut the data in ways that reveal something
- Manager — the most revealing cut. Attrition concentrates; a company at fifteen percent overall may have one manager at forty and the rest near zero.
- Tenure band — exits under six months point at hiring and onboarding; 12 to 18 months at growth and manager quality; three years plus at ceiling and pay drift.
- Performance rating — losing top performers while keeping weak ones is the most dangerous pattern in the dataset, and an average hides it completely.
- Team, hiring source, joining cohort, pay-band position, location and work arrangement.
Step 5: Roll it up into one table
The numbers below are illustrative placeholders showing the structure — fill them with your own.
| Reason code | Exits | Regretted | Concentration | Owner | Action committed | Review |
|---|---|---|---|---|---|---|
| G1 Growth — promotion | 6 | 5 | Engineering, 1–2 yr tenure | VP Engineering | Publish levelling and promotion criteria | Next quarter |
| M1 Manager — support | 4 | 3 | One sales team | Head of Sales | Coaching plan; 1:1 audit | Next quarter |
| C1 Compensation — base | 4 | 2 | Below band midpoint | HR Head | Benchmark two roles; correction budget | Next cycle |
| R1 Role — expectation mismatch | 3 | 1 | Exits under 6 months | TA Lead | Rewrite two JDs; add role-reality call | 6 weeks |
| W1 Workload — hours | 3 | 2 | Support on-call rota | Support Manager | Rebalance rota; one backfill | 8 weeks |
| N1–N5 Non-preventable | 4 | — | Spread | — | Informational only | — |
The useful columns are not the counts. They are Owner, Action committed and Review. Without those three this is a report; with them it is a plan.
Step 6: Run a quarterly attrition review
Book a recurring 60-minute meeting — founder, HR head, every function head, same slot every quarter.
- Headline numbers (5 min): voluntary and regretted attrition, versus last quarter and the same quarter last year.
- Reason-code rollup (10 min): top three codes and what moved.
- The cuts (10 min): manager, tenure, team, performance rating. Name the concentrations out loud.
- Last quarter's actions (15 min): every commitment, its owner, its status — done, in progress or dropped, and why. The most important item on the agenda.
- New commitments (15 min): maximum three, each with a named owner and a date.
- Watchlist (5 min): which teams or managers worry you next quarter, and what are you doing before someone resigns?
Step 7: Close the loop visibly
The loop closes when the organisation knows an action happened because of exit feedback. Publish a short quarterly note — "here is what people told us on the way out, and what we changed." When a fix lands, say what prompted it. Mention it in the next exit interview. Candour is a response to evidence.
Learn before people quit: stay interviews and onboarding feedback
The best exit interview is one you did not need, because you found out six months earlier.
Stay interviews are 30-minute conversations with current employees about what keeps them and what might push them away. Run them with your strongest performers twice a year, conducted by the skip-level or HRBP. Most questions are exit interview questions moved earlier:
- What makes you look forward to a workday here? What makes you dread one?
- If a recruiter called tomorrow, what would make you take the call?
- What one thing would make your job meaningfully better?
- Where do you want to be in two years, and does that look possible here?
- Is there anything you have raised that has not been addressed?
Do not run them unless you will act — an unactioned stay interview is more corrosive when the person is still watching. And do not let them become performance reviews.
Onboarding check-ins catch the causes of early exits while they are fixable. Day 15: do you have what you need, and do you know who to ask? Day 45: is the job what you expected, and what is still unclear? Day 90: do you know what success looks like?
Reuse the same phrasing and rating scales across all three, and you can track one statement — "I have a clear picture of my growth path here" — from day 45 through the stay interview to the exit interview. When that score decays around month twelve across several people, you have found something no exit interview alone would show.
Common mistakes
- Scheduling it on the last working day. They are returning a laptop and thinking about Monday.
- Letting the direct manager run it. If the manager is a factor, this guarantees a useless conversation.
- Treating it as a retention attempt. The moment they sense it, they start managing you.
- Promising confidentiality you cannot deliver, especially in small teams.
- Accepting the first answer. "Better opportunity" is a placeholder, and depth only comes from follow-ups.
- Recording nothing structured. If every exit lives in its own document, twenty conversations never become one insight.
- Coding what they ticked instead of what they said. The dropdown is a social artefact; the story is the data.
- Mixing preventable and non-preventable exits in one number, and treating all attrition as bad when some exits correct a hiring mistake.
- Never looking at manager-level data. Attrition concentrates, and the company average hides it.
- Sharing verbatim feedback with the manager. You burn the source, and the next three people say nothing.
- Ignoring exits under six months, and skipping contractors, interns and field staff — the most diagnostic groups you have.
- Handling a serious disclosure informally, which can compromise a formal process later.
- Letting paperwork friction poison everything. Someone chasing a PF update for three weeks will not refer anyone.
- Doing nothing with the findings. The most common failure of all.
Frequently asked questions
Are exit interviews mandatory in India?
No. What is governed by law and contract is the machinery around it — notice period obligations, full and final settlement, PF and ESIC record updates, gratuity where tenure qualifies, and the handling of complaints under POSH or similar frameworks. The exit interview is a voluntary practice, and participation is voluntary for the employee too. You cannot make it a condition for a relieving letter or a settlement. If someone declines, offer the written form instead.
Who should conduct it — HR or the manager?
For the official, recorded interview, use HR or a skip-level manager. The direct manager is too often part of the reason, and even when they are not, the employee softens their answers. The manager should still have a separate goodbye — just do not use it to collect the official data.
When exactly should it be scheduled?
Roughly 60 to 70 percent of the way through the notice period — around day 40 of a 60-day notice, day 20 of a 30-day one. Late enough that the decision has settled, early enough that details are fresh and action is still possible. Never on the last working day.
How do I get honest answers instead of "better opportunity"?
Decouple the conversation from the paperwork and say so out loud. Ask about timelines rather than causes. Ask about systems rather than people. Always ask a follow-up — the first answer is a summary, the second is the truth. Prove the loop closes by naming a change that came from a previous exit conversation. And send an anonymous survey 60 to 90 days later, once the relieving letter is in hand.
What is the difference between regretted and non-regretted attrition?
Regretted attrition is losing someone you wanted to keep; non-regretted is a departure that is neutral or positive. The distinction matters because a blended percentage is not actionable — eighteen percent made up largely of non-regretted exits is a hiring problem, while the same eighteen percent made up of top performers is an emergency.
How much of what an employee says gets shared with their manager?
Themes, not verbatim quotes. Patterns across multiple exits, not single-source criticism. Nothing attributed by name unless they explicitly agree. In small teams, tell them honestly that the manager may infer the source and let them decide. Praise is the exception — pass that on generously.
What if someone reports harassment or misconduct during the exit interview?
Stop immediately and say that this needs a formal process. Do not investigate, probe for corroboration, or ask them to name witnesses. Route it the same day — the Internal Committee for POSH matters, your whistleblower channel for financial irregularity, the responsible function for safety. Document only what was said, store it separately, do not tell the manager, and do not let it touch the settlement.
Should exit data change how we run hiring?
Yes — one of its highest-return uses. Exits inside the first six months coded as role-expectation mismatch point straight at your job descriptions, your interview process and what hiring managers promise in the room. Send those codes to your talent acquisition lead every quarter. The same applies to hiring source.
Bringing it together
An exit interview is not a form. It is a system with five parts, and all five have to work:
- A timeline with the conversation mid-notice, after KT planning and after the paperwork is decoupled in writing.
- The right interviewer — usually skip-level or HRBP, never the direct manager, never whoever signs the settlement.
- A hybrid format — structured survey first, then a real conversation for depth.
- Questions designed for candour — timelines instead of causes, systems instead of people, always the follow-up.
- A closed loop — reason codes, regretted classification, manager and tenure cuts, a quarterly review with named owners.
Miss any one and you are back to the polite, useless conversation described at the top of this article. And keep the honest perspective: exit interviews are a lagging indicator, so the teams that get the most from them use exit data to build stay interviews and onboarding check-ins, where the same questions still have time to matter.
Starting from zero? Do three things this quarter: move exit interviews to mid-notice, adopt a reason-code taxonomy of about twenty codes, and book a quarterly review where every committed action from last quarter gets read out.
Make offboarding the easy part
Most of the friction around exit interviews is not the questions. It is the relieving letter that is late, the settlement nobody can explain, the PF exit date that never got updated, and feedback sitting in twelve documents nobody can aggregate.
CozyHR handles the offboarding machinery so your team can focus on the conversation — resignation workflows and notice period tracking, knowledge transfer checklists, asset and access recovery, automated full and final settlement with a line-by-line breakdown, relieving letter and experience certificate generation, statutory record updates, configurable exit interview forms with reason-code tagging, and attrition dashboards that cut by manager, team, tenure and regretted status.
If you are running employee offboarding across spreadsheets and email threads, it is worth a look. Start a free trial of CozyHR, or book a walkthrough and we will show you the exit interview process and attrition analysis end to end.
Your departing employees already know what is wrong with your company. The only question is whether you have built a way to hear it.
