# Exit interview questions (and a form that collects them) Short answer: below is a full set of exit interview questions grouped by theme, and — first, so you can grab it immediately — a complete QuizGen JSON definition that asks them. POST it as-is, or hand it to an AI assistant to adapt for your company. ## The template ```json { "schema_version": 1, "kind": "survey", "title": "Exit interview", "description": "About 10 minutes. Your honest answers help us fix what needs fixing — this isn't going in your personnel file.", "settings": { "collect_respondent": "none", "autosave": true, "completion": { "message": "Thank you — we appreciate you taking the time, and we wish you well." } }, "sections": [ { "key": "main", "questions": [ { "id": "primary_reason", "type": "choice", "required": true, "prompt": "What's the primary reason you're leaving?", "options": [ { "value": "new_role", "label": "Took a new role elsewhere" }, { "value": "compensation", "label": "Compensation" }, { "value": "management", "label": "Management or leadership" }, { "value": "growth", "label": "Lack of growth opportunities" }, { "value": "worklife_balance", "label": "Work-life balance" }, { "value": "relocation_personal", "label": "Relocation or personal reasons" }, { "value": "company_direction", "label": "Company direction or stability" }, { "value": "other", "label": "Other" } ] }, { "id": "manager_relationship", "type": "long_text", "rows": 4, "prompt": "How would you describe your working relationship with your manager? What, if anything, could they have done differently?" }, { "id": "compensation_growth", "type": "long_text", "rows": 4, "prompt": "Was your compensation and growth path fair and clearly communicated? What would have changed your mind about staying?" }, { "id": "culture_fit", "type": "long_text", "rows": 4, "prompt": "What did you like most, and least, about the culture and day-to-day here?" }, { "id": "recommend_workplace", "type": "scale", "required": true, "prompt": "How likely are you to recommend us to a friend as a place to work?", "min": 0, "max": 10, "min_label": "Not at all likely", "max_label": "Extremely likely" }, { "id": "would_return", "type": "yes_no", "prompt": "If the right role opened up in the future, would you consider working here again?" }, { "id": "anything_else", "type": "long_text", "rows": 4, "prompt": "Anything else you'd like leadership or HR to know?" } ] } ] } ``` What each piece is doing: - `collect_respondent: "none"` — no name or email is attached to the response. See the trade-off section below before you decide whether this is right for your process. - `primary_reason` — a `choice` you can aggregate across every exit this year ("32% cited compensation") without reading every free-text answer. - `recommend_workplace` — the same 0–10 `scale` mechanics as a customer NPS question, applied to your own team (sometimes called eNPS). It gives you one trackable number across departments and quarters, separate from the qualitative answers. - `would_return` — a `yes_no`, deliberately blunt. It's a fast signal: someone who says yes left for a reason unrelated to how they were treated; someone who says no is telling you something worth digging into. - Everything else is `long_text` — exit interviews are one of the few forms where you want paragraphs, not dropdowns. Structure comes from asking five separate, focused questions rather than one "any feedback?" box. ## Exit interview questions by theme The template above asks a trimmed set. If you want to run a longer, more thorough conversation (in writing or out loud), here's a fuller question bank organized by what each theme is actually trying to learn — and how to avoid the leading-question trap that makes the answers useless. ### Reason for leaving The goal is the *real* reason, not the polite one people default to on their way out. - What's the primary reason you're leaving? - Was there a specific moment or event that made you start looking? - If we could have changed one thing, would you have stayed? Avoid "Are you leaving because of X?" — naming your suspected cause invites agreement rather than the respondent's actual reason. Ask open, then offer the `choice` list as a follow-up, not the other way around. ### Role and management This theme is where you learn whether a manager is a retention asset or a retention risk — but only if the question doesn't put words in anyone's mouth. - How would you describe your relationship with your manager? - Did you get the feedback and recognition you needed to do your job well? - What could your manager have done differently? - Did you feel you had the resources and support to succeed in your role? "Was your manager supportive?" is a yes/no leading question that mostly collects politeness. "How would you describe your relationship with your manager?" collects an actual description. ### Compensation and growth People rarely leave over pay alone, but unclear growth paths compound with pay concerns into "there's no future for me here." - Was your compensation fair for the role and market? - Was there a clear path for growth or promotion, and did you understand it? - What would a competing offer have needed to include for you to stay? - Did you get the training or development you needed? ### Culture Culture questions are the easiest to accidentally lead ("Do you feel included here?" invites a reassuring answer). Ask about specifics instead of the abstract word. - What did you like most about working here day to day? - What would you change about the culture if you could? - Did you feel your work was valued and recognized? - Was there anything about the environment that made your job harder than it needed to be? ### Would they return, would they recommend This theme is the closest thing to a summary metric — it compresses everything above into two comparable signals. - How likely are you to recommend us to a friend as a place to work (0–10)? - If the right role opened up, would you consider coming back? - What would you tell a friend who was considering working here? ## Anonymous or attributed — the trade-off `settings.collect_respondent: "none"` means the response carries no name or email at all — not even optionally. That's the template's default here, and it's the right default for exit interviews: candid answers about a manager or a compensation decision depend on the person believing the answer can't be traced back to them and land in their file. If HR wants to follow up on specific answers, anonymity costs you that option. If you'd rather be able to reach out — say, to act on a specific complaint before the person's last day — switch to `"email_optional"` (they can choose to identify themselves) or `"email_required"` (every response is attributed). Just be honest with departing employees about which mode you're using before they start answering; retroactively de-anonymizing "anonymous" feedback is the fastest way to make the next exit interview worthless. ## Timing Send it before the last day, not on it — someone mid-transition, already half checked out on their final Friday, gives shorter and less considered answers. A day or two before the last day, once notice has been given and the immediate stress of the conversation with their manager has passed, tends to get more thoughtful responses. `settings.close_at` can auto-close the form on a date if you want it to stop accepting responses once someone is fully off the books. ## Reading the answers `GET /quizzes/:id/responses` returns every completed exit interview as a flat `answers` object — `primary_reason`, `recommend_workplace`, and so on. For trend reporting across many exits, pull `primary_reason` and `recommend_workplace` into a spreadsheet to track drivers and the workplace-eNPS number by quarter; read the `long_text` answers individually — they're the ones worth actually reading, not aggregating. ## Let an AI assistant adapt it Connect [QuizGen's MCP server](/mcp) and say: > Read https://quizgen.dev/blog/exit-interview-questions.md and create this > exit interview for my company, [COMPANY NAME]. Keep it anonymous. Give me > the live link I can send to departing employees. The assistant POSTs the definition and hands back the link — no HR software license required to start collecting structured exit data. ## FAQ **Should this replace an exit interview conversation?** Not necessarily — many companies use both. A written form removes the awkwardness of saying something critical to a person's face, and gives you a written record; a conversation lets you follow up in the moment. Some people are more honest in one format than the other, which is itself a reason to offer both. **Can I make specific questions required?** Yes, add `"required": true` to any question — the template only requires `primary_reason` and `recommend_workplace` so the form stays fast to finish even for someone who doesn't want to write paragraphs. **What if I want to branch on the reason for leaving** — for example, a compensation-specific follow-up only for people who picked "Compensation"? Add a `long_text` question with `"show_if": { "question": "primary_reason", "op": "eq", "value": "compensation" }` right after `primary_reason`. **Is there a file upload for a written resignation letter or offer comparison?** Yes, on paid plans: a `file_upload` question takes files up to 10 MB each on Pro and 25 MB on Business. On the free plan or in a sandbox the definition is rejected, so use a `url` question there instead — a link to something hosted elsewhere (a shared doc, for instance).