Seven driver themes, each item shown with the 5-point scale it was written for. Pick 25 to 35, keep the outcome items fixed, and send the rest to match this cycle. Every rated answer rolls up through its driver into one engagement score, so the wording you freeze today is what your trend line rests on for years.
Get the 55 questionsKriti Mishra
AI Content Marketing Specialist
Covers employee engagement measurement at CultureMonkey: instrument design, driver and eNPS scoring, and action planning.
People Science Team
Research team analyzing engagement across 15+ industries globally.
First published Nov 2025
Updated · 14 min read · Fact-checked
Engagement is low, it moves quickly, and gut feel cannot show which team or driver is slipping. A fixed questionnaire turns sentiment into a score a CHRO can track by manager, location, and quarter, and defend in a budget review: which managers need coaching first, which driver gets next quarter’s budget, and whether a reorg or policy change actually landed. For the wider picture, read our guide to employee survey questions.
31%
Only three in ten US employees were engaged at work in 2025. A number that low shifts from quarter to quarter, and no amount of walking the floor tells you which way.
Source: Gallup, 2025 data70%
Over seven in ten employees say they must feel engaged to stay with an employer long term, which makes this the earliest retention signal you have.
Source: SHRM, 2025 State of Company Culture ReportAn employee engagement questionnaire is the scored set of items. An engagement survey is the full program that delivers it, usually once or twice a year. A pulse survey is a short, frequent check that reuses 5 to 10 of the questionnaire’s items. Most HR teams run all three together.
| Aspect | Engagement questionnaire | Engagement survey | Pulse survey |
|---|---|---|---|
| What it is | The scored set of rated items and open questions | The full program: invites, questionnaire, reporting, action | A short recurring check on a few items |
| Items | 25 to 35 rated, 1 to 3 open | Uses the questionnaire | 5 to 10 |
| Cadence | Core items reused every cycle | Once or twice a year | Monthly or quarterly |
| Scale | One scale throughout | Same as the questionnaire | Same scale as the core items |
| Output | Engagement index and driver scores | Company and team reports, action plans | Trend on a few items |
| Owner | People Analytics or HR lead | CHRO or Head of People | HRBP or manager |
For short-cycle items, see these pulse survey questions.
Build an employee engagement questionnaire in six steps: set the decision it must inform, pick your layers, choose one scale, write or select items, pilot with a small cross-section, then lock the core items so every future cycle trends against the same wording.
Write the two or three decisions the results must inform, such as which managers get coaching or which driver gets budget. Owner: HR lead.
Use the Three-Layer Questionnaire Model: 4 to 6 outcome items, 18 to 28 driver items, 1 to 3 open questions.
Use a 5-point scale for every rated item, matching the label set to the verb in the statement, and keep 0 to 10 only for eNPS.
One idea per item, first person, present tense, no jargon. Start from the 55 items below.
Test with 10 to 20 employees across roles, locations, and shifts (5 to 10 in companies under 200 people). Cut anything they found unclear.
Freeze the outcome item wording, record the version, set your minimum reporting group size, and schedule the results date before launch.
Every item belongs to one layer. Fix the counts before you write anything.
Pride, advocacy, intent to stay, discretionary effort. Never reworded, because these are the items you trend on.
Keep 2 to 4 per theme, and map every question to exactly one driver so results roll up cleanly.
Always optional and carries no score, so read it by theme to explain the numbers.
These 55 employee engagement questionnaire items cover seven driver themes, and each one shows the scale it was written for. Pick 25 to 35 for an annual questionnaire, keep the outcome items fixed, and rotate driver items to match this cycle’s priorities.
For more ideas, see this broader list of engagement survey questions.
Selection rule: pick 25 to 35 items, and keep the outcome items fixed. Everything else is yours to rotate.
These items measure how employees feel about the work itself: the tasks, the workload, and whether the role fits what they are good at.
These items separate the manager relationship from senior leadership. Score them apart, because the fix for one is coaching and the fix for the other is communication.
These items ask whether employees can be themselves and speak up. They carry the most social desirability bias, so they need a stated anonymity rule to produce honest answers.
These items measure whether information reaches people in time to be useful, and whether feedback travels back up. Communication scores move fastest during a change.
These items test whether the path forward is visible and whether people get the time and training to walk it. Low growth scores show up in attrition before they show up in the index.
These items measure sustainable workload rather than mood. Keep them worded about work conditions, never about an employee's health, so the questionnaire stays appropriate to ask.
This theme holds the outcome layer. The five marked items are the ones you trend on cycle after cycle, so their wording is frozen and never edited for style.
Use a 5-point scale for every rated item in an employee engagement questionnaire, and match the label set to the verb in the statement: Agreement for “I am recognized for my work,” Satisfaction for “How satisfied are you with your role.” Keep 0 to 10 only for the single eNPS item, and never change a scale between cycles, because a new scale breaks every trend line you report.
Point 1 is always the most negative and point 5 the most positive, with a genuine neutral in the middle. That fixed direction is what lets answers be scored the same way whichever label set you pick.
| Label set | The five points | Write it for | Watch out for |
|---|---|---|---|
| Agreement (default) | Strongly disagree, Disagree, Neither agree nor disagree, Agree, Strongly agree | Statements: “My manager gives me feedback I can act on.” | The neutral midpoint lets unsure employees answer honestly |
| Satisfaction | Very dissatisfied, Dissatisfied, Neither satisfied nor dissatisfied, Satisfied, Very satisfied | Questions that ask how satisfied someone is | Do not mix with Agreement inside one driver |
| Frequency | Never, Rarely, Occasionally, Frequently, Very frequently | Questions about how often something happens | Measures behavior, not opinion, so read it separately |
| Quality, Performance, Importance, Focus, Happiness | Five more shipped label sets, each 5 points, negative to positive | Match the set to the verb in the statement | A smiley display is available and records the same 1 to 5 value |
| 0 to 10 likelihood | 0 Not at all likely to 10 Extremely likely | The single eNPS item only | Scored as promoters minus detractors, not as favorable % |
| Open text | Free response | The Voice layer | Needs theme analysis to be useful |
For wording examples, see these Likert scale question examples, or the help center on rating scale questions.
An annual employee engagement questionnaire should have 25 to 35 rated items and 1 to 3 open questions. A pulse should have 5 to 10. Go shorter for frontline teams answering on a phone, and cut any item you cannot act on.
Length is a trust decision as much as a data one. Every item you add past the point of usefulness tells employees the questionnaire is for HR rather than for them, and the cost shows up as rushed answers in the second half. Size the instrument to your headcount and to how finely you intend to report.
Report by department only where the group meets your minimum size.
Report by manager.
Core items stay identical everywhere, plus up to 5 module items per business unit.
The evidence on length
Respondents told a survey would take 10 minutes started it far less often than those told it would take 8, and drop-off rose with every extra minute announced.
Score an employee engagement questionnaire by averaging the ratings. Each question scores as the mean of every rating it received, each driver scores as the average across its questions, and the overall engagement score is the average of every rating answer in scope. Beneath each average sits a favorable, neutral, and unfavorable split that shows the shape of opinion.
Report the average and the split together. The average is what trends and benchmarks cleanly; the favorable share is the sentence a manager can act on. A driver sitting on a comfortable mean can still hide a large unfavorable group, and that split is the thing worth the conversation.
| Benchmark | Figure | How to use it | Source |
|---|---|---|---|
| Global median engagement score (5-point mean) | 3.92 of 5 | Reference for mean scores on 5-point driver items | CultureMonkey Employee Engagement Benchmarks by Industry, Q1 2026 |
| Employees engaged worldwide | 20% | Gallup’s stricter “engaged” category, not a favorable %; do not compare directly | Gallup, 2025 data |
| US workers highly satisfied with their job | 50% | Satisfaction runs higher than engagement; never mix the two | Pew Research Center, Oct 2024 |
Mean scores and favorable percentages are different measures; convert before comparing. The CultureMonkey figure covers 10.2M anonymized responses across 1,247 companies. The Pew figure covers 5,273 employed US adults.
Next, learn how to interpret engagement survey results.
Use a custom, driver-based employee engagement questionnaire when you need to act by team, license Gallup Q12 when you want Gallup’s benchmark and program, and use UWES-9 when you need a research-grade measure of individual work engagement. Q12 items are proprietary to Gallup. UWES is published by its academic authors.
The choice comes down to what you owe the result: a licensed instrument buys an external comparison and costs you the ability to reword anything, while a custom one gives managers items they recognize and puts the burden of validity on you.
Gallup Q12
Proprietary benchmark instrument
Measures
Access
Gallup states the items are proprietary and delivered through its platform (Gallup Q12).
Best forOrganizations buying Gallup’s program and benchmark
UWES-9
Academic, peer-validated scale
Measures
Access
Published by its academic authors; validated with 14,521 respondents in 10 countries (Schaufeli, Bakker and Salanova, 2006). Check the authors’ terms before commercial use.
Best forResearch-grade measurement of individual work engagement
Custom questionnaire
The Three-Layer model
Measures
Access
You own the wording, and the burden of validity sits with you.
Best forHR teams that need to act by team and manager
For item-level detail, see Gallup Q12 survey questions explained.
Adapt an employee engagement questionnaire by changing delivery and length, not the core items. Frontline teams need QR or SMS access and fewer items. Remote teams need items on connection and clarity. Global teams need professional translation of every scale label, reviewed by a local People lead.
The line to hold is the outcome layer. Once those five or six items read identically in every language and on every channel, the rest of the instrument can flex to the audience without costing you the comparison.
A QR code or shareable link for shop floors, break rooms, and shared kiosks, or SMS; 15 to 20 items; launch at shift start. See how to survey factory workers without email.
Add driver items on clarity of expectations and connection with the team; deliver in Slack or Microsoft Teams.
Translate items and every scale label professionally, have a local People lead review them, and keep the core items identical. See how to run one survey across multiple countries.
Brief worker representatives on purpose and anonymity rules before launch.
Employees complete an engagement questionnaire when they trust it is anonymous, can answer on the channel they already use, and have seen action on the last one. Announce it from their manager, open it for one to two weeks outside peak workload, and share participation daily.
Participation is a lagging measure of the previous cycle. If the last questionnaire produced no visible change, no reminder schedule will fix this one.
Start from this engagement survey announcement sample.
Email, Slack, Microsoft Teams, WhatsApp, SMS, or QR.
Open for one to two weeks; avoid quarter-end, peak season, and shift changes.
Who sees what, and the two thresholds: the per-survey response threshold that hides a group until enough people answer, and the dashboard threshold that governs the heatmap.
Post participation by team daily and remind people what changed after the last cycle.
I actually care about that more than I care about what people say in the survey. I care about are they speaking up? Because when they’re speaking up, they care.
L. David KingsleyChief People Officer, IllumioCultureClub X S06 E19Watch episode Seven mistakes ruin employee engagement questionnaire data: leading items, double-barreled items, jargon, hypotheticals, personal questions, vague items, and rewording a tracked item between cycles. Each one produces a score you cannot act on or cannot compare.
The first six are write-time errors and cost you one cycle. The seventh is a reporting error and costs you every cycle that came before it, because the trend line silently stops meaning what it did.
| Mistake | What it sounds like | Write it this way |
|---|---|---|
| Leading item | “You enjoy working with your manager, don't you?” | My manager gives me useful feedback. |
| Double-barreled item | “I feel valued and fairly paid.” | Split it into “I feel valued at work.” and “I am paid fairly for my work.” |
| Jargon | “Leadership communication efficacy meets my needs.” | Senior leaders explain the reasons behind decisions. |
| Hypothetical | “If we doubled in size, would you stay?” | I see myself working here in two years. |
| Personal question | Any item on religion, politics, family, or health. | Remove it and keep items on work conditions. |
| Vague item | “Do you like your job?” | My work gives me a sense of accomplishment. |
| Rewording a tracked item | Changing “I am proud to work here” to “I feel proud of this company.” | Keep outcome wording fixed and add new items as new rows. |
In the 30 days after employee engagement questionnaire results arrive, share the company summary, give each manager their team report, have every team pick one driver to fix, and tell employees what will change and when. Silence after a questionnaire costs more trust than never asking.
Thirty days is not arbitrary. It is roughly the window in which employees still connect a change to the questionnaire they answered. Past that, the same change reads as something the company was going to do anyway, and the next cycle’s participation pays for it.
CultureMonkey runs the employee engagement questionnaire end to end: it delivers it over email, SMS, WhatsApp, Slack, Microsoft Teams, or a link and QR code, holds anonymity to a response threshold, rolls every rating up through drivers into one engagement score, and shows each manager what moved since the last cycle.
AI breaks each open-text comment into topics and tags the sentiment and emotion behind it, sized by how many people raise each theme. Any comment’s sentiment can be corrected by hand. It is an account-level feature enabled on eligible plans.
Compare Survey puts an earlier cycle beside the latest one and shows the difference in every cell, green where a score rose and red where it fell. It is one of the engagement survey tools teams lean on most between cycles.
One master survey, auto-translated into the languages you add and editable by hand. Each employee answers in their own language and it all reports as one survey.
Share a survey link or QR code for on-site teams. Respondents still verify an email or employee ID and a one-time passcode, so nobody answers twice. A QR-only survey sends no notifications.
Scope a sub-admin to chosen teams, locations, or business units, and hide free-text responses, the heatmap, or strengths and weaknesses from managers where anonymity needs it.
Each manager sees their team’s engagement score, participation, and eNPS with trends, plus their highest and lowest drivers, and works actions through To do, In progress, and Completed. The same data feeds the wider employee experience platform.
The Benchmark tab compares your standard drivers against an external benchmark for your industry and employee size, with a separate eNPS benchmark and an optional regional filter. The reference data behind it comes from our people science work.
Comparing engagement survey software? Use this vendor evaluation checklist alongside a shortlist of survey vendors.
An employee engagement questionnaire is the scored core of any engagement program: a fixed set of rated statements and a few open questions that shows how committed, motivated, and supported employees feel, in a form you can compare across teams and cycles. This guide gave you 55 items with their scales, the Three-Layer Model for picking 25 to 35 of them, a scoring method with benchmarks, a comparison of Gallup Q12 and UWES against a custom set, and a 30-day plan for acting on results.
CultureMonkey helps HR teams run that questionnaire without the manual work. It sends one survey in every language you add, reaches frontline sites by link or QR code, holds anonymity to a response threshold, reads open comments into topics and sentiment, and puts an earlier cycle beside the latest one on a team heatmap so managers can act within the month.
Eight questions HR teams ask most often when they are putting an engagement questionnaire together.
A satisfaction survey asks whether employees are content with pay, benefits, and conditions. An employee engagement questionnaire asks whether they are committed, motivated, and willing to give extra effort, and which engagement drivers explain that. Satisfied employees are not always engaged, so HR teams usually track engagement as the primary score on an employee engagement survey.
Design it so an employee finishes in one short sitting. With 25 to 35 rated items and one to three open questions, most HR teams aim for under 15 minutes. Pulse versions of 5 to 10 items should take a few minutes, which matters most for frontline staff answering on a phone.
Gallup states that its Q12 items are proprietary and available through its own platform. If you want the Q12 benchmark, license it from Gallup. If you run your own questionnaire, write original items that measure similar drivers, such as clarity of expectations, resources, recognition, and development.
Yes, for most organizations. Employees answer more honestly when no individual response can be traced back to them. Keep demographic questions to the minimum you need, report team results only above your response threshold, and explain in the invitation exactly who sees what. Anonymity is set by the survey, not the channel, so it holds whether someone answers by email, Slack, or a QR code.
Run the full questionnaire once or twice a year and short pulses in between, usually quarterly. Only survey as often as you can share results and act. A questionnaire that repeats before employees see any change from the last one lowers participation.
Compare against your own last cycle first, then against an external benchmark that uses the same scale and scoring method. Scores only compare fairly when the items, scale, and favorable definition match. See the benchmark table in the scoring section for current reference points.
Set the threshold before launch and never show a team result below it, even to senior leaders. In CultureMonkey an anonymous survey defaults to a response threshold of 3 and cannot go lower, while the dashboard heatmap uses a separate anonymity threshold that defaults to 5 and can never be set below 3. Roll small teams into their parent group rather than hiding them, so every employee’s answers still count.
Yes, if the core outcome items stay identical and only delivery changes. Reach frontline staff with a QR code or shareable link they can scan at a shared kiosk, or by SMS, trim driver items for shift workers, and translate every scale label. Identical core items let you compare the two groups fairly.