--- title: "Employee survey questions: 48 examples by theme" description: "These 48 employee survey questions cover six themes. Each statement to rate has an open question that explains a low rating. Seven others are best left out." url: "https://raconte.ai/en/blog/employee-survey-questions" --- # Employee survey questions: 24 statements to rate, and the 24 open questions that explain the ratings These 48 employee survey questions cover six themes. Each statement to rate has an open question that explains a low rating. Seven others are best left out. October 10, 2026 Godefroy Key takeaways * A statement rated from 1 to 5 shows how many people share a problem. An open question shows what the problem is, so write the two as a pair. * Ask the open question about a recent, real moment ("the last time you asked for help") and people tell you what happened. Ask for an opinion and you get one word. * Leave out any question whose answer you cannot act on, any statement that covers two subjects, and any profile question you will not use to compare groups. * The people who rate a statement low have the most to explain. Invite them to a five-minute voice interview, where a follow-up question gets the example nobody types in a comment box. * Two or three open questions are enough in a written survey. With more, people skip them or answer in two words. An employee survey needs two kinds of questions. A closed question, usually a statement rated from 1 to 5, measures how many people share a problem. An open question collects what happened. Most surveys have forty closed questions and one comment box, so they come back with scores and no reasons. This article gives 48 employee survey questions in six themes (workload, management, recognition, growth, meaning and process). Each of the 24 statements to rate comes with the open question that explains a low rating. You will also find the questions to leave out, and a way to ask the open questions out loud to the people who rated low. ## A 1-to-5 rating shows a problem without its cause A rating measures how many people agree with a statement. If the support team gives “My workload is reasonable” an average of 2.4 out of 5, you know the team mostly disagrees. You do not know which of these four situations they are in: * two people left in the spring and were never replaced * the new ticketing tool adds a step to every request * one customer takes up half of everyone’s week * priorities change every Monday morning Each situation calls for a different decision: a hire, a tool setting, a conversation with a customer, or a planning rule. All four produce the same 2.4. Survey platforms try to fix this with a comment field under each rating. Culture Amp, an employee survey platform, reports from its own data that [about 6% of people add a comment to a rating question](https://www.cultureamp.com/blog/employee-survey-comments), and that people who disagree with the statement are eight times more likely to comment. A separate open question does better, with around 60% participation in the same data. Typed answers stay short, though, and nobody can ask the author of “too many meetings” which meetings they mean. So write each open question at the same time as its closed question, with the same care. The closed one shows you where to look, and the open one is the question you would ask if you could sit down with the person. This article is the question bank. To choose the cadence, get a good response rate and use the results, read how to [run an employee engagement survey](/en/blog/employee-engagement-survey). ## 48 employee survey questions in six themes Each table pairs a statement to rate with the open question that explains a low rating. Have people rate the statements on a five-point scale, from “Strongly disagree” to “Strongly agree”. The open questions all ask about a recent, real moment. “Tell me about the last time you asked your manager for help” gets a story. “What do you think of your manager?” gets “fine”. None of them starts with “why”, because [an open-ended question that asks for an episode](/en/blog/open-ended-questions-qualitative-research) gets what happened, while “why” gets a justification. You do not need all 48. Pick two or three statements per theme, and keep two or three open questions for the written survey. Use the other open questions in a follow-up interview with the people who rated low. ### Workload Statement to rate Open question that explains a low rating My workload is reasonable. Describe your last heavy week. What filled it? I can finish my work within my normal hours. What did you last finish in the evening or at the weekend, and what pushed it there? I have the tools and information I need to do my job. What did you last have to wait for, or work around, to get a task done? I can take time off without work piling up. What happened to your work the last time you took a few days off? A workload statement can score low for very different reasons, so its open question is the one you need most. With “Describe your last heavy week”, you can tell a team that is short of people from a team that loses its time to a tool or to one customer. ### Management Statement to rate Open question that explains a low rating My manager helps me when I need it. Tell me about the last time you asked your manager for help. What happened next? I know what my manager expects of me. When did a priority last change for you? How did you find out? My manager gives me feedback that helps me improve. What is the last piece of feedback that changed the way you work? I can disagree with my manager without it being held against me. Think of the last decision you disagreed with. What did you do? Every statement here is about the direct manager, the level where someone can act on the answer. “Senior leadership listens to employees” gives a score that nobody on the team can move. ### Recognition Statement to rate Open question that explains a low rating My work gets the recognition it deserves. When did you last feel your work was noticed, and by whom? When I do good work, someone tells me. Which piece of work from the past three months went unnoticed? I am paid fairly for the work I do. What do you compare your pay with when you judge whether it is fair? Promotions here go to the people who earned them. Think of the last promotion on your team. What did it tell you about what counts here? A low recognition score can come from pay, praise or promotion, and each of the three depends on a different person. So each one has its own statement. The question on pay comparisons shows what people measure their pay against: a job ad, a colleague hired last month, a former employer. ### Growth Statement to rate Open question that explains a low rating I have learned something new in my job this year. What is the last thing you learned at work, and how did you learn it? I can see what my next step here could be. What would you like to be doing in two years, and what stands in the way? Someone at work talks to me about my progress. When did you last talk about your career with someone here? What came out of it? Most days, I get to do what I do best. Which part of your job would you gladly do more of, and which would you hand over? Several of these statements echo [Gallup’s Q12](https://www.gallup.com/q12/), the best-known engagement questionnaire, which includes “In the last six months, someone at work has talked to me about my progress.” A time frame in the statement (“this year”, “in the last six months”) makes people think of facts rather than of a general feeling. ### Meaning Statement to rate Open question that explains a low rating I understand how my work contributes to the company’s goals. Which of your tasks would nobody miss if you stopped doing it? I can do quality work. Describe the last time you delivered something you were not proud of. What forced your hand? I trust the direction the company is taking. Which recent company decision did you understand the least? I would recommend this company as a place to work. What would you tell a friend who was about to apply here? “I can do quality work” picks up the cases where people are asked to ship something they find poor. “I would recommend this company as a place to work” is the eNPS question (employee Net Promoter Score), which many companies have people rate from 0 to 10 and track from year to year. ### Process Statement to rate Open question that explains a low rating Decisions that affect my work are explained to me. What is the last decision that changed your work? How did you hear about it? Our meetings are a good use of my time. Which recurring meeting would you cancel, and what would be lost? I can get a decision quickly when I need one. What are you waiting on right now, and since when? Our processes help me do my work. Which step in your work exists only because of a rule or a tool? How do you get around it? Process questions are the easiest ones to act on, since an answer usually names a meeting, an approval step or a tool. Start with this theme if your last survey changed nothing: a cancelled meeting is a result people see within a week. ![Six pairs of employee survey questions, one per theme: a statement to rate on the left and the open question that explains it on the right](/.netlify/images?url=_astro%2Fpairs-en.70EPJeK1.jpg&w=1376&h=768&dpl=6ad1cc428d394a5c89e0155b) Statements to rate have a known flaw: people tend to agree with whatever is put to them. Jon Krosnick, professor of communication and political science at Stanford, and Stanley Presser, sociologist at the University of Maryland, reviewed the studies on [agreement scales](/en/blog/quantitative-research-questions-examples) and put this bias at about ten percentage points. Keep the exact same wording from one round to the next. The bias then stays the same in every round, so the change between two rounds is more reliable than the level of a single one. You still get this bias if you swap the scale for a yes or no, because [people lean towards yes on a two-option question](/en/blog/dichotomous-questions-examples) too. If your survey has to feed a psychosocial risk assessment, a grid built for that purpose serves you better than these six themes. Our [social barometer questionnaire](/en/blog/social-barometer) follows the six risk areas used in French occupational health. ## Questions to leave out of an employee survey Remove a question when you cannot use its answer. Question Problem Ask this ”My manager is available and sets clear goals.” Two subjects share one rating, so nobody knows which one a 2 is about One statement on availability, another on goals ”How much do you value our generous remote-work policy?” The wording contains the expected answer ”Our remote-work policy suits the way I work." "I am happy with where the office is.” (the lease was just renewed) Nothing can change, and people see their answer lead nowhere A question on what you can change: days on site, commute support ”Would you use a mentoring programme?” People say yes to an idea and never sign up ”Who did you last ask for career advice?" "Which colleague slows the team down?” The survey becomes a way to report on colleagues, and people stop trusting it ”What are you waiting on right now, and since when?” Age, gender and exact job title In a team of eight, three criteria are enough to recognise someone Department and a tenure band, nothing more ”Any other comments?” The question is too broad, so most people skip it or write “no" "What costs you the most time or energy in your work?” Culture Amp warns against questions like the one about the office in its [tips for writing survey questions](https://www.cultureamp.com/blog/good-employee-survey-questions): “You’re put in a difficult position if you ask a question about something you can’t change (or aren’t ready to discuss).” With each question, you promise people that their answer will be used. People who see that promise broken once answer the next survey with less care. Questions about harassment or discrimination need a decision before the survey goes out. If an anonymous “yes” comes back, someone has to know what happens next, and people have to be pointed to the formal reporting channel on the same screen. Without that, leave the question out of the survey and promote the reporting channel separately. ![Seven kinds of question to leave out of an employee survey, from the statement covering two subjects to the final "Any other comments?"](/.netlify/images?url=_astro%2Fleave-out-en.BhKrxiNm.jpg&w=1376&h=768&dpl=6ad1cc428d394a5c89e0155b) ## Ask the open question out loud to people who rated low The people who rated a statement 1 or 2 are the ones who can explain the score. A written survey cannot go back to them, since they are anonymous. You can invite them to come forward instead, in one of two ways. * Inside the survey, add one line on the last page, with a link to a five-minute voice interview: “If you rated your workload low, tell us what happened. It takes five minutes, out loud, and we ask for no name.” If your survey tool has conditional logic, show that line only to people who rated the statement 1 or 2. * After the survey, send the same interview to the team where the theme scored lowest. This step belongs [between two rounds of a workplace climate survey](/en/blog/social-barometer), before the results message goes out. The interview is spoken because people say more out loud than in writing. A 2023 study published in _Social Science Computer Review_ put the same open questions to about a thousand smartphone users, half typing and half speaking. [Typed answers averaged around 21 words and spoken answers around 55](https://journals.sagepub.com/doi/full/10.1177/08944393231160961). A spoken interview also allows a follow-up: when someone says “too many meetings”, the interviewer asks which ones. ```mermaid graph TD A[Statement rated in the survey] --> B{Rating of 1 or 2?} B -->|No| C[The survey ends as usual] B -->|Yes| D[Link to a five-minute voice interview] D --> E[Open question on a recent, real moment] E --> F{Answer still general?} F -->|Yes| G[Follow-up: ask for an example] G --> F F -->|No| H[Transcript read by HR] H --> I[Message to the team: what we heard, what changes] ``` The interview with people who rated their workload low can start from this prompt. It reuses two open questions from the workload table, and you can swap in the questions of any other theme. > You are running a short, confidential interview with an employee who rated their workload low in our latest survey. Your goal is to understand what makes the workload heavy. Start by asking them to describe their last heavy week and what filled it. Then ask what they last had to finish in the evening or at the weekend, and what pushed it there. Whenever an answer stays general, ask for a recent example. Never ask them to name a colleague, and do not defend the company. After about five minutes, ask which single change would help the most, then stop. [Raconte](/en), an AI voice interview tool, runs this kind of interview. You paste the prompt, turn on a [reusable share link](/en/docs/share-link) set to ask for neither name nor email, and put the link on the last page of the survey. Each person opens it in a browser and talks, and the AI follows up whenever an answer stays vague. For each interview you get the transcript, the audio, a summary and the sentiment of every answer. Raconte delivers the interviews one by one. It computes no score and builds no report across interviews, so the ratings stay in your survey tool. To read a whole round of transcripts together, connect an AI assistant through the [MCP server](/en/docs/mcp). Interviews run in English or French. For a sense of cost, say 50 people open the link and talk for six minutes each. That makes 300 minutes. The first 60 minutes of the month are free, and the other 240 are billed at [$0.12 a minute](/en/pricing), so the round costs $28.80. Keep the voice interview for explaining low ratings. If all you need is a rating from 300 people, the survey gets it faster. An open conflict on a team, or a report of harassment, needs a person who can act: a manager, HR, or the formal reporting channel. ## Anonymity and response rate depend on the questions too Most advice on anonymity is about the tool. The questions matter as much. * Every profile question makes the groups smaller. Ask for the department and a tenure band, and hide any group under five people. * An open answer can give away its author through a detail, such as a project name or a date. Ask people in the survey introduction to name no colleague, and have HR pass on themes and paraphrased quotes to managers. * When a link is shown only to low raters, whoever listens knows that each speaker rated low. Say so in the invitation, along with who listens to the recordings. * A colleague can recognise a voice. Present the voice interview as confidential, since no tool can make a voice anonymous. Many employees doubt that a survey is anonymous. In a [survey of 1,000 US employees](https://www.hrdive.com/news/engagement-surveys-paint-a-distorted-picture/739260/) published in February 2025 by Visier, a publisher of HR analytics software, 37% said they do not believe engagement surveys are ever really anonymous. To get candid answers, say precisely who reads what before anyone answers. No setting in the tool does as much. The questions also set the length of the survey, which affects the response rate. Culture Amp finds that [ten minutes or less is ideal](https://www.cultureamp.com/blog/good-employee-survey-questions). Its standard engagement template has 57 questions and takes most people [just under ten minutes](https://support.cultureamp.com/en/articles/7048331-engagement-survey-template-overview). Twelve to eighteen statements from this article, plus two or three open questions, stay well under that length. A short survey does not protect the response rate on its own, though. Participation falls mostly when [people saw nothing change after the previous round](/en/blog/employee-engagement-survey). ## Where to start Build your first version in this order: 1. Pick two or three statements per theme from the six tables, twelve to eighteen in all. 2. Add two or three open questions, taken from the themes you expect to score lowest. 3. Check every question against the table of questions to leave out. 4. Limit the profile questions to department and tenure band. 5. Write the follow-up interview for one theme, and put its link on the last page. 6. Set the date of the results message before the survey goes out. Keep the wording of the statements unchanged in the next round. Change the open questions and the interview theme as often as you like, since they are not compared from one round to the next. With [Raconte](/en), the first 60 interview minutes each month are free, with no credit card. That covers a first follow-up interview on a small team. After that, [you pay per minute of interview](/en/pricing). ## Frequently asked questions ### What should an employee survey include? An employee survey should include statements to rate on the themes that drive how people feel about their work (workload, management, recognition, growth, meaning and process), one or two profile questions such as department and tenure, and two or three open questions. Each statement covers one subject and uses the same wording in every round. Each open question asks about a recent, real moment. ### How many questions should an employee survey have? Aim for a survey people finish in ten minutes or less. A full engagement survey often has 40 to 60 statements, and Culture Amp’s standard template has 57. A shorter survey of 12 to 18 statements and two or three open questions is enough to follow six themes. Take the survey yourself to check how long it takes. ### What are good open-ended questions for an employee survey? Good open-ended questions ask about a recent, real moment: “Describe your last heavy week”, “What are you waiting on right now, and since when?”, “Which recurring meeting would you cancel, and what would be lost?”. They avoid “why”, which makes people justify themselves, and they replace the final “Any other comments?”, which most people skip. ### Is an employee survey really anonymous? An employee survey is anonymous only if the results of small groups are hidden, commonly under five respondents, and if profile questions stay few. In a team of eight, age, gender and job title together identify a person. A detail in an open answer can also give its author away, so HR should pass on themes to managers and keep the raw answers to itself. ### Is an employee survey mandatory? No law requires a company to run an employee survey, and employees are generally free to skip it. Employers do have a duty to protect their employees’ health, and in France they must keep a written assessment of psychosocial risks, which a survey can feed. Culture Amp treats a 100% response rate as a warning sign, because it often means people were pushed to answer. ### What questions should you avoid in an employee survey? Avoid statements that cover two subjects, wordings that contain the expected answer, hypothetical questions (“Would you use…”), questions about things you cannot change, questions that ask people to name or judge a colleague, and profile questions you will not use to compare groups. Replace the final “Any other comments?” with one concrete open question. Ready to launch your first interview? Describe your interview, share the link: an AI agent runs it by voice with each participant. You get back the transcript and analysis. [Try for free](/login) ## Continue reading [![Employee engagement survey: what it measures and misses](/.netlify/images?url=_astro%2Fthumbnail.BmAwo4dk.jpg&w=1376&h=768&dpl=6ad1cc428d394a5c89e0155b) October 7, 2026 ## Employee engagement survey: what it measures and misses What an employee engagement survey measures, why response rates fall, why few people write a comment, and the quarterly voice interview that explains a score. ](/en/blog/employee-engagement-survey)[![Social barometer: indicators, questionnaire, cadence](/.netlify/images?url=_astro%2Fthumbnail.CSL8vL2n.jpg&w=1376&h=768&dpl=6ad1cc428d394a5c89e0155b) October 9, 2026 ## Social barometer: indicators, questionnaire, cadence A social barometer tracks how employees see their work. Read it with your HR figures, adapt our sample questionnaire and collect the reasons behind the scores. ](/en/blog/social-barometer)[![45 open-ended questions for qualitative research](/.netlify/images?url=_astro%2Fthumbnail.x5EdDQ3f.jpg&w=1376&h=768&dpl=6ad1cc428d394a5c89e0155b) August 22, 2026 ## 45 open-ended questions for qualitative research A bank of 45 open-ended questions grouped by research goal, the reason each one is phrased the way it is, and the four ways a question closes back up. ](/en/blog/open-ended-questions-qualitative-research)