An employee survey is a structured set of questions a company asks its employees, usually anonymously, to measure how they feel about their work, their manager, their team, and the company, and to find out what should change. It is the only reliable way to hear from everyone at once rather than from the people who speak up, and it is only worth running if someone intends to act on the answers.
In 2026 the practical questions are not whether to survey but how: which types to run, how often, how many questions, over which channel, and how to close the loop fast enough that people keep answering. This guide is Matter’s complete answer: the eight survey types and when each fits, question design with a dozen examples, anonymity and trust, delivery in Slack, Microsoft Teams, and email, response rates, closing the loop, and the honest case for when a free form tool is enough. We build Matter, survey software that runs inside Slack and Teams, so read the software section with that in mind. The advice is vendor-neutral, every third-party figure links to its primary publisher, and the Matter data comes from our published research, including a survey of our own that we use here as a design example.
Last verified September 2026.
The short version
Employee surveys are how a company listens to everyone at once, and they work when they are short, anonymous, regular, delivered where people already are, and followed by visible action. Most companies need three kinds: an annual engagement survey to map the drivers, a monthly or quarterly pulse to track them, and lifecycle surveys at onboarding and exit to catch what the others miss; eNPS gives a single trend number, and custom surveys handle everything else. The one practice that matters most is closing the loop: share the results and two commitments within two weeks, or response rates fall and the next survey measures cynicism. It costs nothing to start: a free form tool and a spreadsheet run a credible program for a small team, and chat-native survey software earns its price when you need anonymity people trust, trend tracking, reminders, and same-day responses. Start with five questions, one open-ended, sent this month.
What is an employee survey?
An employee survey is a questionnaire, sent to some or all employees, that measures attitudes, experiences, and intentions on a consistent scale so the results can be compared across teams and over time. The defining features are anonymity or confidentiality (so people answer honestly), a fixed set of questions (so the results are comparable), and an owner who will act on the results (so it is worth answering). A survey without any one of those is a poll: harmless, and mostly useless.
Surveys sit inside a wider practice called employee listening, which also includes one-on-ones, skip-levels, focus groups, suggestion channels, and the informal reading a good manager takes every day. Surveys are the part of listening that scales, and the part that gives you numbers. They are not a substitute for the conversation; they tell you where the conversation is needed.
Pulse survey vs. engagement survey vs. eNPS
An engagement survey is the census: 20 to 40 questions across every driver, run once or twice a year, benchmarked against the last run. A pulse survey is the speedometer: one to five questions, run weekly to monthly, tracking a few drivers over time. eNPS (employee Net Promoter Score) is one question, how likely are you to recommend this company as a place to work on a 0 to 10 scale, with the score calculated as the percentage of promoters (9 to 10) minus the percentage of detractors (0 to 6). They answer different questions and most programs need all three. Our guides to the employee engagement survey and the pulse survey go deeper on each.
What an employee survey is not
It is not a performance review, a vote, or a legal record. It is not anonymous if fewer than five people answered from a team and the report shows the team. It is not a listening program if the results go to a folder. And it is not the source of engagement: a survey measures what managers and peers create every week; it does not create anything itself.
Why employee surveys matter: what the research shows
The case for surveying rests on two facts: the things that drive engagement and retention are mostly invisible to leadership until someone asks, and by the time they become visible in turnover it is too late to fix them. The research below is linked to its primary publisher in every case.
The engagement gap you are measuring. Global employee engagement fell to 20% in 2025, down from 21% in 2024 and a 23% peak in 2022 and 2023, according to Gallup’s State of the Global Workplace 2026. Actively disengaged employees rose to 20%, and only 34% of employees worldwide are thriving in their overall lives. A survey is how you find out which side of that line your own teams sit on.
Why results are reported by manager. The manager alone accounts for 70% of the variance in team engagement. Gallup first published the finding in its 2015 State of the American Manager report and restates it in its current engagement guidance. The practical consequence for survey design is that company-wide averages hide the story; the useful result is the spread between teams.
Turnover is preventable, and surveys are the warning system. 75% of employee departures were preventable. That is the finding of Work Institute’s 2025 Retention Report, based on more than 120,000 exit interviews (Work Institute, 2025). Gallup’s data on leavers explains why so few are prevented: 45% of voluntary leavers say neither a manager nor another leader proactively discussed their job satisfaction, performance, or future with them in the three months before they left. 44% of those who discussed their intention to leave never raised it with their direct manager, and 77% either left within three months of starting to search or never actively searched at all (Gallup, 2024). The window for a stay conversation is short, and a pulse survey is often the only thing that opens it in time.
What engagement is worth. Engagement is associated with 21% lower turnover in high-turnover organizations and 51% lower turnover in low-turnover organizations. Top-quartile units also show 70% higher employee wellbeing, measured as the share of thriving employees (Gallup Q12 meta-analysis, 2024). The Q12 is itself a twelve-question employee survey, which is the strongest argument in the literature that a short, stable, well-designed questionnaire can predict business outcomes.
Onboarding is where the first survey belongs. 70% of new hires decide whether a job is the right fit within the first month, including 29% who know within the first week. 44% of employees say they had regrets or second thoughts about the job within the first week, which is why BambooHR concludes that companies have about 44 days to influence a new hire’s long-term retention (BambooHR, 2023, survey of 1,565 adults, April 2023). An onboarding survey at day 30 is the cheapest retention tool a company has.
Nobody asks. Only 10% of employees have been asked by someone at work how they like to be recognized. Employees who strongly agree they get the right amount of recognition are 4x as likely to say someone at work encourages their development (Gallup, 2022). The figure is about recognition, but the lesson is general: the questions that would change how a manager behaves are rarely asked, and a survey is the one place they get asked of everyone.
What Matter’s data adds. In The Recognition Habit, Matter’s 2026 Benchmark Report, drawn from aggregate, de-identified platform data across thousands of teams and 76,000+ members, over half of pulse responses arrive within the first hour and 87.5% within a day; coin rewards roughly double completion (39.6% vs 19.6%, survey-weighted across 84,000+ pulse surveys); 98.2% of pulse surveys run anonymous; and 98% of teams attach coin rewards on custom surveys. Habit teams, the ones that recognize on a weekly ritual, are twice as likely to run pulse surveys. The findings are correlational: they describe what surveys look like on teams with different habits, not a controlled experiment. All of our studies, with methodology, live on Matter’s research hub.
Types of employee surveys
Eight survey types cover almost everything a company needs to ask. They differ in cadence, length, anonymity, and what the result is for. The table summarizes; the sections after it say when each earns its place.
| Survey type | Cadence | Length | Anonymous | Best for |
|---|---|---|---|---|
| Engagement survey | Annual or twice a year | 20 to 40 questions | Yes | Mapping the drivers and the gaps between teams |
| Pulse survey | Weekly to monthly | 1 to 5 questions | Yes | Tracking a few drivers over time; early warning |
| eNPS | Quarterly | 1 question plus a follow-up | Yes | A single trend number for leadership |
| Onboarding survey | Day 30, 60, and 90 | 5 to 10 questions | Confidential | Catching a bad start before it becomes an exit |
| Exit survey | At departure | 5 to 10 questions | Confidential | Learning the real reason, and the preventable share |
| Lifecycle survey | At each milestone | 5 to 10 questions | Confidential | Promotions, manager changes, returns from leave |
| Custom survey | As needed | Varies | Depends | A specific decision: a policy, a benefit, an event |
| 360 feedback | Annual or on request | 10 to 20 questions | Confidential, attributed to the subject | Development feedback on one person from all sides |
1. Engagement surveys
The engagement survey is the full census of every driver: expectations, tools, recognition, manager relationship, development, voice, mission, colleagues, and growth. Run it once or twice a year, keep the core questions identical between runs, and report the spread between teams rather than the company average. Its weakness is speed: a result in March describes February, and the fix ships in June. Treat it as the map. Our employee engagement guide covers what the drivers are and how to act on them.
2. Pulse surveys
A pulse survey asks one to five questions on a fixed cadence, weekly to monthly, and its value is entirely in the trend. Keep two or three anchor questions stable for a year, rotate one topical question, and always include one open-ended prompt. Delivered inside chat, a pulse gets answered the same day, which is what makes the trend honest: in the 2026 Benchmark Report, over half of responses arrive within the first hour. The full question bank is in our guide to the employee pulse survey, and Matter’s pulse surveys page shows how they look in Slack and Teams.
3. eNPS surveys
eNPS is one question with a 0 to 10 scale and a follow-up asking why. It is popular with leadership because it is one number that moves, and it is misused for the same reason: a small team’s eNPS swings 40 points when two people change their minds. Run it quarterly, report it at ten or more respondents, and read it with the follow-up text, which is where the information is. Our eNPS guide covers the calculation and the benchmarks.
4. Onboarding surveys
A new hire decides early, so ask early: a day-30 survey on clarity, tools, welcome, and manager contact, a day-60 survey on workload and belonging, and a day-90 survey on growth and whether they would choose the job again. Keep these confidential rather than anonymous (the manager needs to know who is struggling), and route the results to the manager and HR within a day. See our onboarding survey questions for a full set.
5. Exit surveys
An exit survey asks the person who is leaving what actually happened, and it works better as a short written survey sent before the last day than as an interview on the last afternoon, because people are more candid in writing and less inclined to be polite. Ask for the main reason, the moment they started looking, what would have kept them, and whether they would return. Read them in aggregate every quarter; a single exit is a story, ten are a pattern.
6. Lifecycle surveys
Lifecycle surveys are short check-ins at the moments most likely to change how someone feels about the company: a promotion, a manager change, a return from leave, a reorganization, a first anniversary. They are cheap to run because the trigger is a date in the HR system, and they catch what the annual survey misses because it asks everyone the same thing at the same time regardless of what just happened to them.
7. Custom surveys
Custom surveys exist for a specific decision: should we change the holiday schedule, which benefits do people actually use, how did the offsite land. Keep them short, tell people what decision the survey informs, and report the decision when it is made. On Matter, 98% of teams attach coin rewards on custom surveys, which is a hint about how the best teams treat them: as a small ask worth a small thanks.
8. 360 feedback surveys
A 360 survey collects feedback on one person from their manager, peers, direct reports, and sometimes themselves. It is a development tool, not an engagement measure, and it needs the most care of any survey here: confidential to the reviewers, focused on behaviors rather than traits, and delivered by someone trained to discuss it. Our 360 feedback questions guide covers the question set.
How to run employee surveys well: 7 principles
These are the practices that separate programs whose response rates rise year over year from the ones that quietly die. Each runs on a free form tool as well as on dedicated software.
1. Short
Every question you add costs responses, and the responses you lose are not random: the busiest and the most disengaged drop out first, which biases the result toward the people who already like it here. Cap pulse surveys at five questions and engagement surveys at 40, and cut any question whose answer would not change what you do. One open-ended question is worth more than five scale questions you will not act on.
2. Anonymous, and visibly so
People answer honestly when they believe the answer cannot be traced, and belief is the operative word. State the anonymity rule in the survey itself, set a minimum group size for any reported breakdown (five is the usual floor), and never ask demographic questions that would identify someone on a small team. On Matter, 98.2% of pulse surveys run anonymous, which reflects how teams actually choose when the option is one toggle. Onboarding, exit, and 360 surveys are the exception: confidential to HR and the manager, and say so.
3. Regular, on a cadence people can predict
A survey that arrives on the same day every month becomes a habit; one that arrives when leadership is worried becomes a signal that something is wrong. Publish the cadence, keep it, and keep the anchor questions stable so the trend means something. If a question must change, change it at the start of a year and note the break in the series.
4. Delivered where people already are
Response rates track friction. A survey that opens inside Slack or Microsoft Teams, where the person already is, gets answered between tasks; a link in an email gets opened later or never. The Benchmark Report figures above (over half of pulse responses within the first hour, 87.5% within a day) come from surveys delivered in chat. Email still has a place for people who do not live in a chat app, for exit surveys, and for the annual census, so the practical rule is chat first, email as the fallback, and never a separate portal login.
5. Acted on within two weeks
The single largest determinant of the next survey’s response rate is what happened after the last one. Within two weeks of closing, share the headline result, name one or two things that will change, and say who owns them. Teams that skip this step see the rate fall by a few points every run until the survey measures only the people who never gave up. Our guide to increasing survey participation covers the tactics in detail.
6. Closed-loop, not just reported
Reporting is telling people what they said. Closing the loop is telling them what you did about it, and then asking again. The loop has four parts: the result, the commitment, the progress note in the next survey, and the retest of the same question. When people see a low score on a question lead to a change and then a higher score, they learn that answering works. That lesson is the whole program.
7. Small thanks for the ask
A survey is a favor, and a small acknowledgment changes the completion math. On Matter, a coin reward attached to a pulse survey roughly doubles completion (39.6% vs 19.6%). The reward does not need to be money: a note of thanks from the leader who reads the results, a share of the results, or a visible decision all count. What matters is that the person sees the ask was noticed.
Employee survey examples: questions by survey type
Twelve questions across the survey types, written to be used as they are. Scale questions use a 1 to 5 agreement scale unless noted; keep wording identical between runs.
Engagement survey questions
- I know what is expected of me at work.
- I have the tools and information I need to do my job well.
- I can see a path to grow my career at this company.
These three cover expectations, resources, and growth, the drivers most often at the bottom of a first survey. The full set of employee engagement survey questions has 40 more, grouped by driver.
Pulse survey questions
- My workload this week was manageable.
- In the past week, I received recognition for doing good work.
- What is one thing that would make next week better? (open-ended)
The first two are anchors to keep stable all year; the third is the open-ended prompt that catches what the scale misses. Our pulse survey questions post has the full rotation.
eNPS question and follow-up
- How likely are you to recommend this company as a place to work? (0 to 10)
- What is the main reason for your score? (open-ended)
Onboarding survey questions
- In my first week, I understood what was expected of me.
- I know who to ask when I am stuck.
Exit survey questions
- What was the main reason you decided to leave? (choose one, then explain)
- What could we have done differently that would have kept you here? (open-ended)
360 feedback question
- What is one thing this person should keep doing, and one thing they should start doing? (open-ended)
A design example: how Matter built its own survey
The clearest way to show question design is a survey we ran ourselves. The State of Employee Recognition and Rewards 2026 surveyed 1,021 employed U.S. professionals, recruited through an independent third-party research panel (Cint), between June 30 and July 13, 2026. Three design choices are worth copying. The sample was recruited by a panel rather than from our own customers, so the answers describe the workforce and not our users. Frequency was asked on a fixed scale (weekly, monthly or less, rarely or never), which is why the headline can be stated cleanly: 45% of employees are recognized weekly, 33% monthly or less, and 21% rarely or never, so 1 in 3 is thanked once a month or less. And the report states its correlations as correlations: it found that weekly-or-better recognition is more common at chat-first workplaces than at email-only ones, and says plainly that the finding describes an association, not a cause. Fixed scales, an honest sample description, and careful claims are the same three rules that make an internal survey worth trusting.
How to build an employee survey program
A program is the schedule and the ownership that turn individual surveys into a listening system. The compact version has five steps; the full step-by-step guide covers question banks, communication templates, reporting formats, and how to scale from one team to the company.
- Set the calendar. One engagement census a year, a pulse monthly or quarterly, eNPS quarterly, onboarding at 30, 60, and 90 days, and exit for every departure. Publish the calendar so people expect the ask.
- Write the anchor questions. Five to eight questions that will not change for a year, covering expectations, recognition, manager, growth, and the recommend question, plus one open-ended prompt. Everything else rotates.
- Decide where it lives. Surveys that open inside the tool people already use get answered the same day. For Slack teams that means a survey delivered in Slack with a reminder in the channel; for Microsoft 365 teams it means the same inside Teams, with Microsoft Forms as the free baseline. Our roundups of Slack survey apps and Microsoft Teams apps for employee surveys compare the options on each platform, including running the program with no new software.
- Set the anonymity and reporting rules before the first send. Minimum group size for any breakdown, who sees raw text, how confidential lifecycle surveys are routed, and how long results are kept. Put the rules in the survey introduction.
- Schedule the loop. Results within one week, commitments within two, a progress note in the next pulse, and a retest of the lowest question next quarter. Assign an owner for every commitment. This is the step most programs skip and the one that determines whether the program lives.
How to measure employee survey success
A survey program has its own metrics, separate from what the surveys measure, and they tell you whether the program itself is healthy. Track these monthly, with the reference points from Matter’s platform data where we have them.
- Response rate. The share of invited people who completed the survey. The level varies by channel and team size, so the useful reading is the trend: a rate that falls run over run means people have stopped believing the survey changes anything. On Matter, a coin reward attached to a pulse roughly doubles completion (39.6% vs 19.6%).
- Time to respond. How quickly answers arrive after the send. Chat-delivered surveys on Matter see over half of responses within the first hour and 87.5% within a day; an email survey that trickles in for two weeks is measuring two different moods.
- Open-ended completion. The share of respondents who wrote something in the free-text box. Falling text is an early sign of fatigue even when the scale questions still get answered.
- eNPS trend. The quarter-over-quarter direction at ten or more respondents. Ignore single-quarter moves on small teams.
- Action rate. The share of survey commitments that were delivered by their date. Publish it; it is the number that tells employees whether the loop is real.
- Retention. Regrettable turnover by team, compared with the team’s survey scores two quarters earlier. Surveys should predict exits; if they do not, the questions are wrong.
Read the numbers by team, not by company, because the manager drives most of the variance. And report the survey program’s own metrics alongside the results: a rising engagement score on a falling response rate is not good news.
Employee survey software and tools
Survey software adds five things a form tool does not: anonymity people can verify, delivery inside Slack or Teams with reminders, a stable question bank with trend reporting, lifecycle triggers tied to start and end dates, and a way to say thanks for completing. It does not add listening. A team of 25 with a Google Form, a Microsoft Form, or a Typeform, a spreadsheet for the trend, and a leader who shares results within two weeks has a working program, and it should keep it until the manual parts slip: the reminder gets forgotten, the anonymity gets doubted, the trend lives in someone’s inbox, and the onboarding survey stops going out. Our comparison of employee survey software covers the category from free form tools to enterprise listening suites, and says where each fits; HR and People Ops teams who do not want to be the program’s engine are the usual buyers.
Matter is our own answer, so weigh this accordingly. Matter’s Surveys add-on runs pulse and eNPS, onboarding, and custom surveys inside Slack and Microsoft Teams, with email delivery as well, on a recognition platform with no separate login; templates and examples are on our employee surveys page. There is a Free plan for unlimited users for recognition and a 14-day free trial on paid plans: Basic is $1 per user/month, billed annually, Pro is $3 per user/month, billed annually, and the Surveys add-on is $2 per user/month, billed annually, with details on our pricing page. Feedback Friday™ (US Patent 12,199,935) prompts the team weekly; give-coins reset weekly, earned coins never expire. The start is fast: almost half of teams (46%) send their first recognition within 15 minutes of setup, 1 in 3 within 5 minutes, and the figures are identical on Slack and Teams. Two honest limits: surveys are a paid add-on rather than part of the Free plan, and dedicated employee-experience suites go deeper on analytics, benchmarks, and driver analysis, so a large company that needs those is better served by one of them. Chat-native employee listening is the fit.
One 2026 note on AI and trust. In Matter’s Did AI Kill the Thank You survey of 1,021 U.S. professionals, 30% of employees would be more honest in an employee survey if they knew AI was analyzing the open-ended responses, versus 16% less honest; the number one trust condition, chosen by 51%, is that a human reviews the AI’s summary before any decision is made, 20 points ahead of anonymity (31%), and only 13% would not trust AI-analyzed feedback under any condition. If your survey tool summarizes free text with AI, say so in the survey, and say that a person reads the summary before anything is decided.
Frequently asked questions about employee surveys
What is an employee survey?
An employee survey is a structured questionnaire, usually anonymous, that a company sends to its employees to measure how they feel about their work, manager, team, and company, and to find out what should change. The main types are engagement surveys, pulse surveys, eNPS, onboarding and exit surveys, lifecycle surveys, custom surveys, and 360 feedback. A survey is only useful when someone acts on the results and tells employees what changed.
How often should you survey employees?
Run a full engagement survey once or twice a year, a pulse survey monthly or quarterly with one to five questions, eNPS quarterly, an onboarding survey at 30, 60, and 90 days, and an exit survey for every departure. Publish the cadence so people expect it, keep the anchor questions stable so the trend is comparable, and never survey more often than you can act.
Should employee surveys be anonymous?
Engagement, pulse, eNPS, and most custom surveys should be anonymous, with a minimum group size of about five before any breakdown is reported and no demographic questions that could identify someone on a small team. Onboarding, exit, and 360 surveys work better as confidential surveys, where HR and the manager can see who answered, because the response needs a personal follow-up. Whichever you choose, say so in the survey itself.
What is a good response rate for an employee survey?
The level depends on channel, team size, and how much people trust the program, so the trend matters more than any single number: a rate that falls run over run means employees have stopped believing the survey changes anything. Surveys delivered inside Slack or Microsoft Teams are answered faster than email surveys, and on Matter a small coin reward roughly doubles pulse completion (39.6% vs 19.6%). The best way to raise the rate is to close the loop on the last survey.
What questions should an employee survey ask?
Ask a short, stable set covering expectations (I know what is expected of me), resources (I have the tools I need), recognition (I received recognition this week), the manager (someone at work cares about me), growth (I can see a path to grow here), and the recommend question for eNPS, plus one open-ended question such as what is one thing we should change. Keep the wording identical between runs and drop any question whose answer would not change what you do.
What is the difference between a pulse survey and an engagement survey?
An engagement survey is the annual or twice-yearly census with 20 to 40 questions across every driver; it maps where the gaps are. A pulse survey is short (one to five questions) and frequent (weekly to monthly); it tracks a few drivers over time and gives early warning when a team changes. Most programs need both, plus eNPS as a single trend number for leadership.
Do you need employee survey software?
Not at first. A free form tool and a spreadsheet run a credible program for a small team, as long as a leader shares results and commitments within two weeks. Software earns its price when you need anonymity people can verify, delivery inside Slack or Teams with reminders, lifecycle triggers tied to start and end dates, trend reporting across many teams, and a way to thank people for completing. Chat-native tools such as Matter start at a few dollars per user; enterprise listening suites are quote-priced.
The bottom line
Employee surveys are how a company hears from everyone, and they work when they are short, anonymous, regular, delivered where people already are, and followed by visible action within two weeks. Run the census once a year, the pulse every month or quarter, and the lifecycle surveys at the moments that decide whether someone stays, and read every result by team rather than by company. The research says most turnover is preventable and most of it is never discussed; a survey program is the mechanism that gets it discussed in time. If your team lives in Slack or Microsoft Teams and you want pulse, eNPS, onboarding, and custom surveys answered where people work, next to recognition, Matter is our answer; if a form and a spreadsheet are working, keep them working.
