An employee survey program is the calendar, the rules, and the follow-through a company puts around asking its people how work is going: which surveys run (annual engagement, pulse, eNPS, onboarding, exit), how often, where they are delivered, how anonymity is protected, and what happens after the results come in. A single survey is an event. A program is the loop that turns answers into visible changes, so that people keep answering.
This guide is for the HR lead, founder, or people ops manager designing a survey program for a company of 20 to a few thousand people, or repairing one whose response rates have collapsed. It draws on Matter’s published research (aggregate, de-identified data across 84,000+ pulse surveys and 76,000+ members, plus Matter’s own survey studies of 1,021 U.S. professionals) and on studies from Gallup, Work Institute, and BambooHR, each linked to its publisher. We build Matter, which delivers surveys inside Slack and Microsoft Teams, so read the software section with that in mind. The steps are vendor-neutral, and we say plainly when a free form builder and a spreadsheet are enough.
Last verified September 2026.
The short version
A good employee survey program asks a few questions often, protects anonymity in a way people can verify, and shows within weeks what changed because of the answers. The design decisions that matter most are cadence (a short pulse every month or quarter beats a long survey once a year), anonymity (a minimum group size for reporting, stated up front), and visible follow-through (a “you said, we did” note after every survey). Delivery channel decides whether anyone answers: in The Recognition Habit, Matter’s 2026 Benchmark Report, over half of pulse responses arrive within the first hour and 87.5% within a day. A realistic budget is $40 to $80 per employee per year for software and completion rewards, and close to nothing for a small team on a free form builder. Plan four to six weeks from decision to first pulse, and a full year before the program has a trend line worth acting on.
What is an employee survey program?
An employee survey program is a planned, repeating set of surveys that together tell a company how its people are doing and where to act, with agreed rules for anonymity, reporting, and response. An annual engagement survey, a monthly pulse, a quarterly eNPS question, and onboarding and exit surveys are five instruments; the program is the calendar that keeps them from colliding, the question bank that keeps them comparable, and the loop that closes each one.
It is not a suggestion box, a performance review, or a one-off poll, although a program can absorb all three. The components are the same on an enterprise listening platform or a free form and a spreadsheet:
- People: who is surveyed (everyone, including frontline and part-time staff), who sees results at which level, and who owns the follow-through for each team.
- Cadence: a calendar that spaces the annual survey, pulses, eNPS, and lifecycle surveys so no one is asked twice in the same week.
- Questions: a stable core set that repeats so trends are real, plus a rotating slot for what is current.
- Anonymity and trust: how responses are separated from identities, the minimum group size below which results are not shown, and who can see comments.
- Follow-through: the commitment that results are shared within a set number of days and that each survey produces at least one visible action.
- Tooling and measurement: where surveys are delivered and answered, and the response rate, time-to-response, and trend lines that say whether the program is working.
Types of employee survey programs
Most programs combine several of these. Add a pulse first, because it builds the answering habit and the follow-through muscle; then eNPS and onboarding; then the annual survey, once the company can act on what it hears.
Annual engagement survey
A long survey, 30 to 60 questions, run once a year across the whole company, measuring engagement drivers such as clarity, recognition, manager support, and growth. It suits companies of a few hundred people and up that want benchmarks and driver analysis, and it is the wrong first survey for a company that has never closed a feedback loop.
Pulse surveys
Short surveys, one to five questions, sent weekly, biweekly, or monthly, usually anonymous, tracking a few core items over time. Pulses suit every size of company and are the backbone of a modern program because they give managers something to act on this month rather than next year. In the 2026 Benchmark Report, 98.2% of pulse surveys run anonymous, which is the default to keep.
eNPS
One question (“How likely are you to recommend this company as a place to work?”) on a 0 to 10 scale, with an optional follow-up asking why, run monthly or quarterly. It suits companies that want a single number leadership will track; on its own it says that something changed, not what, so it works best inside a pulse.
Onboarding surveys
Surveys triggered at day 7, day 30, and day 90 for every new hire, asking whether they have what they need, whether the job matches the offer, and whether they have been welcomed. They suit any company hiring more than a handful of people a year, because the first weeks are when the decision to stay is made: 70% of new hires decide whether a job is the right fit within the first month, including 29% who know within the first week (BambooHR, 2023, survey of 1,565 adults, April 2023).
Exit surveys
A short survey sent to every leaver, ideally after the last day so answers are candid, asking why they left and what would have kept them. The case for them is blunt: 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).
Custom and ad hoc surveys
One-off surveys on a specific question: the office move, the benefits renewal, the new on-call rotation. They are the type most likely to cause fatigue if they are not on the calendar, so give them a budget of slots per quarter and attach a small reward to completion: in the 2026 Benchmark Report, 98% of teams attach coin rewards on custom surveys.
What the research says about employee survey programs
Engagement is low and not recovering. Global employee engagement fell to 20% in 2025. That is 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 (Gallup, State of the Global Workplace 2026). The same report finds manager engagement dropped to 22% in 2025, from 27% in 2024 and 31% in 2022, while non-manager engagement sits at 19%, so the gap between managers and the people they lead has nearly closed.
The manager is the unit of analysis. 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 (Gallup). That is the argument for team-level reporting above a minimum group size, with managers owning the follow-through.
Engagement moves turnover. 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).
Leavers were rarely asked. 42% of employees who voluntarily left their organization say their manager or organization could have done something to prevent it. Gallup’s July 2024 study surveyed a random sample of 717 U.S. employees who had voluntarily left an employer within the past year (Gallup, 2024). In the same study, 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. A pulse is the cheapest way to start that conversation in time.
Delivery channel and a small reward decide whether people answer. In the 2026 Benchmark Report, drawn from aggregate, de-identified Matter platform data across thousands of active workspaces over the trailing twelve months, over half of pulse responses arrive within the first hour and 87.5% within a day, and coin rewards roughly double completion (39.6% vs 19.6%, survey-weighted across 84,000+ pulse surveys). Habit teams, those with a weekly recognition 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. Methodology for all of Matter’s studies is on our research page.
People will be honest with AI analysis if a human reads it. In Matter’s Did AI Kill the Thank You survey, 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%); only 13% would not trust AI-analyzed feedback under any condition. If your platform summarizes comments with AI, say so, and say who reads it.
How to build an employee survey program in 7 steps
These steps work on a listening platform, a chat-native survey tool, or a free form builder. Where Matter’s data or product bears on a decision we say so; the advice does not depend on any vendor.
Step 1: Decide what the program is for and what you will report
Write the purpose in one sentence: “Find the teams where people are struggling early enough to help, and prove to leadership that we act on what we hear.” Then pick two numbers: one about the program (response rate, or share of surveys with a published action) and one about the business (voluntary turnover, an engagement index, or eNPS). Baseline both before the first survey.
Step 2: Set a budget, including the follow-through
Three lines: software, completion rewards, and time. Software runs from nothing (a free form builder) to a few dollars per user/month for a chat-native tool, and to quote-based annual contracts for enterprise listening platforms. Completion rewards are optional and small. The line companies forget is time: someone reads the results, writes the summary, and chases the actions, and every manager needs an hour per survey to discuss results with their team. If that time is not in the budget, run fewer surveys.
Step 3: Build the question bank and the anonymity rules
Choose five to eight core questions that repeat in every pulse (recommend, workload, manager support, recognition, clarity) and keep their wording fixed for a year so the trend line is real; add one rotating question for what is current. Then write the anonymity rules down and publish them: responses are separated from identities; results are shown for groups of five or more; managers see their team’s scores and see comments only above the threshold; nobody can filter to a single person. A rule people cannot verify is not anonymity, so explain the mechanism. For a model of how to state a survey’s basis, Matter’s own The State of Employee Recognition and Rewards 2026 states its sample (1,021 employed U.S. professionals), its recruitment (an independent third-party research panel, Cint), and its field window (June 30 to July 13, 2026); an internal survey deserves the same discipline.
Step 4: Set the calendar and protect it
A workable default for most companies: a three-to-five-question pulse monthly, an eNPS question quarterly inside the pulse, onboarding surveys at day 7, 30, and 90, an exit survey for every leaver, and an annual engagement survey only once the monthly loop has run for six months. Put every survey on one calendar, including the ad hoc ones, and enforce a rule that nobody is asked more than once in a week. A pulse rides best on an existing rhythm. Matter’s recognition prompt, Feedback Friday™ (US Patent 12,199,935), prompts the team weekly; give-coins reset weekly, earned coins never expire; a pulse scheduled on the same day inherits the habit. A free-form program can do the same with a recurring Monday post and a fixed close date.
Step 5: Decide where surveys are delivered and answered
Response rate is mostly a delivery decision. A survey answered inside Slack or Microsoft Teams, with the answer buttons in the message, gets answered the same day; a survey linked from an email gets answered when someone remembers. Frontline and shift staff without email need a QR code, SMS, or kiosk path or the results describe only the office. The platform-specific details (who can install apps, whether a tenant admin has to approve, what Microsoft Forms and Viva Pulse do and do not do) are covered in our guides to Slack survey apps and Microsoft Teams apps for employee surveys. Chat-native tools also start fast: in the 2026 Benchmark Report, 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.
Step 6: Launch with a pilot, and train managers on the follow-through
Run the first two pulses with two or three teams whose managers volunteered, and use them to test the threshold, the close date, and the summary format. Before company-wide launch, train every manager on a 30-minute routine: read the team result, pick one thing to change, tell the team what it is and when, and report back at the next pulse. Give them a template for the “you said, we did” note. Launch by having leadership share their own team’s first result, including the lowest score, because trust in a survey comes from seeing that low scores are safe to give.
Step 7: Close the loop every time, then review the program quarterly
Publish company-level results within ten working days of each close, and team-level results within a week of that. Every survey produces at least one visible action, even a small one, and the next survey opens with a line about what changed. Review the program each quarter: response rate by team, time-to-response, and whether any survey on the calendar has gone two cycles without an action. Retire that one.
Employee survey program examples
These are templates, not named companies: designs we see working, with a calendar, budget, and metric so you can copy the shape. The numbers are design targets, not results.
A 40-person agency: one monthly pulse in Slack
Calendar: a three-question anonymous pulse on the first Friday of each month (workload, recommend, one rotating question), plus an exit survey for every leaver. Budget: a free form builder or a chat-native tool at published per-user rates; the founder spends an hour a month on the summary. Follow-through: results and one action posted in the all-hands channel within a week. Metric: response rate, target 85%, and the month-over-month workload score.
A 120-person SaaS company: pulse, eNPS, and onboarding
Calendar: a five-question monthly pulse with eNPS in the quarterly edition; onboarding surveys at day 7, 30, and 90; an annual engagement survey in Q1 after the first year of pulses. Budget: a chat-native survey tool at about $3 to $5 per user/month, billed annually, plus a small coin reward per completed pulse. Follow-through: each manager posts a “you said, we did” note in their team channel within two weeks. Metric: eNPS trend and day-90 onboarding score.
A 600-person manufacturer: shift-friendly pulses
Calendar: a two-question pulse every two weeks, answered from a QR code posted at the shift board and from supervisors’ phones, with questions on safety and being heard; a quarterly five-question edition; exit surveys by SMS. Budget: software at published per-user rates plus rewards attached to completion. Follow-through: results printed and posted at the shift board with the one action taken. Metric: response rate by shift, so the night shift is visible, and the safety item trend.
A 250-person professional services firm: quarterly listening with comments
Calendar: a ten-question quarterly survey with an open comment field, eNPS included; onboarding at day 30 and 90; a manager-change survey fired 60 days after any reporting-line move. Budget: a listening platform or chat-native tool, plus a partner’s hour per quarter per practice group. Follow-through: comments summarized by theme and read by a human before anything is shared; practice leads present one action each at the quarterly meeting. Metric: comment participation rate and the manager-change survey score.
A 1,000-person company: the full calendar
Calendar: an annual engagement survey in September; monthly pulses that rotate through the annual survey’s lowest three drivers; quarterly eNPS; onboarding at day 7, 30, and 90; exit surveys; event-triggered surveys after reorgs. Budget: a listening platform or a chat-native tool, a completion reward budget, and a half-time owner. Follow-through: a published action register, one row per team per quarter, with owner and status. Metric: share of teams with a closed action each quarter, target 90%, and the annual engagement index.
A remote-first team of 80: a weekly one-question micro-pulse
Calendar: one question every Friday, rotating through a bank of eight, always anonymous, with a monthly recap; eNPS quarterly; onboarding at day 7 and 30. Budget: a chat-native tool at published rates; the people lead spends 20 minutes on the Monday recap. Follow-through: the recap names the question, the score, the change from last month, and what is being done. Metric: time-to-response (target: 80% within a day) and the recommend score.
What an employee survey program costs
Software. Matter’s Surveys add-on is $2 per user/month, billed annually, on top of Basic at $1 per user/month, billed annually, or Pro at $3 per user/month, billed annually, so Pro plus Surveys totals $5 per user/month; it covers pulse and eNPS, onboarding, and custom surveys delivered in Slack, Teams, and email, with a 14-day free trial and a Free plan for unlimited users on the recognition side. For comparison, verified September 2026 on each vendor’s pricing page: Lattice sells Engagement at $4 seat/month, billed annually, covering pulse, onboarding and exit, eNPS, and action planning; Polly has a free plan for Slack polls with unlimited responses and 45-day results history; Microsoft Forms is included in Microsoft 365 licenses at no separate price; Workleap Officevibe is sold in suite bundles starting at $4,999 USD per year; and Culture Amp is quote-based and billed annually. Plan details for Matter are on Matter’s pricing page.
Completion rewards. Optional, and small. In the 2026 Benchmark Report the median reward redemption is $30 and 40% of redemptions are under $25; a survey coin worth a dollar or two is in proportion, and coin rewards roughly double pulse completion. A working assumption of $1 to $2 per completed survey, across a dozen pulses and a couple of lifecycle surveys a year at 70% completion, is roughly $15 to $20 per employee per year.
Time. One owner at two to four hours per survey for a few hundred people (summary, actions, chasing), plus roughly an hour per manager per survey to discuss results with their team. This line is larger than the software line at every company size.
The table is a sample annual budget. The arithmetic is ours: software at Matter’s Basic plus Surveys rate of $3 per user/month, billed annually ($36 per user per year); completion rewards assumed at $18 per employee per year; owner time as hours because the rate is yours. Substitute your own tool’s published rate.
| Line item (annual) | 50 employees | 250 employees | 1,000 employees |
|---|---|---|---|
| Software (Basic + Surveys, $36 per user per year) | $1,800 | $9,000 | $36,000 |
| Completion rewards ($18 per employee) | $900 | $4,500 | $18,000 |
| Owner time | About 3 hours per survey | About 6 hours per survey | A half-time owner |
| Total spend | $2,700 (about $54 per employee) | $13,500 (about $54 per employee) | $54,000 (about $54 per employee) |
A program on a free form builder costs the time line only, and for a company under 50 that is a legitimate design for the first year. Paid software buys delivery inside chat (where the response rate comes from), anonymity handled by the tool rather than by whoever owns the spreadsheet, and trend lines nobody maintains by hand.
Common mistakes in employee survey programs
Surveying without follow-through
The most common failure and the most expensive. Every survey that produces no visible action lowers the response rate of the next one, because people learn that answering changes nothing. If you cannot commit to one action per survey, run fewer surveys.
Anonymity people cannot verify
“Your answers are anonymous” with no stated mechanism or group threshold is a promise, not a design. Publish how responses are separated from identities and the minimum group size for reporting, and never let a manager filter a result to one person.
Survey fatigue from an unmanaged calendar
Three teams sending ad hoc surveys in the same week is how a 90% response rate becomes 40%. One calendar, one owner, and a rule that nobody is asked twice in a week.
Only an annual survey
A 50-question survey once a year tells you what last spring felt like, three months after it closed. Start with a short monthly pulse and add the annual survey once the loop works.
Delivering by email link
A link in an inbox competes with everything else in the inbox. Deliver where people work, with the answer buttons in the message, and give frontline staff a path that does not need email.
Results that stop at HR
If managers never see their own team’s numbers, they cannot act, and the company-level score becomes a number leadership argues about. Report at team level above the threshold and make managers the owners of the action.
How to measure an employee survey program
Measure the program before you measure the company. The reference points are from Matter’s published benchmarks; your first pulse will be lower, and that is fine.
- Response rate: the share of invited people who completed. Above 70% for a short pulse is healthy; below 50% means the survey is in the wrong place or people have stopped believing it matters.
- Time-to-response: how quickly answers arrive. Over half of pulse responses on Matter arrive within the first hour and 87.5% within a day; if most of yours arrive after a reminder, the delivery channel is the problem.
- Completion by team and by channel: the view that shows which teams and which delivery paths are silent, so frontline groups are not missing from the average.
- Action rate: the share of surveys, and of teams, with a published action within 30 days of close. Target 90%; this is the number that predicts next quarter’s response rate.
- eNPS and core item trends: movement in the recommend question and the fixed core items over rolling quarters, at company and team level. A single reading is noise; three in a row is a signal.
- Retention: voluntary turnover by team and tenure band against the baseline from Step 1, and the exit-survey reasons over time. Movement here takes quarters.
Report program metrics monthly to managers and business metrics quarterly to leadership.
Software for employee survey programs
A company under 50 can run a good program on a free form builder, a recurring calendar post, and a spreadsheet for the trend line, and many should for their first year. Software earns its place when you want surveys answered inside chat, anonymity handled by the tool, a repeating calendar with reminders, and trend lines by team that nobody maintains by hand; that usually happens between 50 and 200 people. Our guide to employee survey software compares twelve tools, from free form builders to enterprise listening suites, on verified pricing and Slack and Teams delivery. We build Matter, which delivers pulse, eNPS, onboarding, and custom surveys inside Slack and Microsoft Teams with coin rewards for completion; dedicated listening suites go deeper on benchmarks and driver analysis, and the Slack survey apps and Microsoft Teams apps guides linked in Step 5 rank the options platform by platform. For a company that wants employee listening in Slack and Teams without a separate login, our employee surveys product page shows what that looks like.
Frequently asked questions about employee survey programs
What is an employee survey program?
An employee survey program is a planned, repeating set of surveys (typically an annual engagement survey, a monthly or quarterly pulse, an eNPS question, and onboarding and exit surveys) run on one calendar, with published anonymity rules, team-level reporting above a minimum group size, and a commitment that every survey produces a visible action.
How much should we budget for an employee survey program?
Plan for three lines: software from nothing (a free form builder) to about $2 to $5 per user/month, billed annually, for a chat-native tool, with enterprise listening platforms quote-based; optional completion rewards of roughly $15 to $20 per employee per year; and time, which is the largest line, at a few hours per survey for the owner plus an hour per manager per survey. That is roughly $40 to $80 per employee per year for a mid-size company with paid software and rewards.
How do we get managers to participate in a survey program?
Give managers their own team’s results above a minimum group size, a 30-minute routine (read, pick one action, tell the team, report back), and a template for the “you said, we did” note. Have leadership share their own results first, including the lowest score. Managers act when the result is theirs to act on and leadership visibly does the same.
What is the difference between an employee survey program and a pulse survey?
A pulse survey is one instrument: a short, frequent survey that tracks a few core items over time. An employee survey program is the whole system a pulse sits inside: the calendar that spaces it against the annual survey, eNPS, and lifecycle surveys, the shared question bank, the anonymity rules, and the follow-through loop.
How often should we survey employees?
A short pulse of three to five questions monthly, with eNPS quarterly, is a workable default for most companies; some teams run a one-question pulse weekly. The annual engagement survey comes once the monthly loop has proven you act on results. Keep it on one calendar and never ask anyone more than once in a week; fatigue comes from unmanaged surveys, not frequent ones.
Should employee surveys be anonymous?
Pulse, engagement, and exit surveys should be anonymous by default, with a published minimum group size (commonly five) below which results are not shown and a stated mechanism for separating responses from identities. In Matter’s benchmark data, 98.2% of pulse surveys run anonymous. Onboarding surveys are the exception: they are often attributed on purpose so a struggling new hire can be helped, and the survey should say so.
The bottom line
An employee survey program succeeds on three decisions: a cadence short enough to act on, anonymity people can verify, and a follow-through loop that shows what changed after every survey. Delivery decides whether anyone answers, which is why the survey belongs where people already work. Start with a monthly pulse, put every survey on one calendar, train managers on a 30-minute routine, and measure the response rate and the action rate before you measure engagement. A small team can do that on a free form builder. When you want pulses, eNPS, onboarding, and custom surveys delivered and answered inside Slack or Microsoft Teams, with a small reward for completing them and no separate login, that is what we built Matter to do.
