90 DEI survey questions to ask employees in 2026
Questions across five dimensions: diversity, equity, inclusion, belonging, and leadership accountability. Each item includes scale and format guidance, plus a legal self-ID guide and a 60-day rollout plan.
- DEI survey questions measure how fairly and inclusively employees actually experience work across five areas: diversity, equity, inclusion, belonging, and leadership accountability.
- A usable DEI survey pairs experience questions on a five-point agreement scale with a small set of optional self-ID demographic questions. Compare scores between groups rather than against an external benchmark, and publish one owned action per confirmed gap within 30 days.
- 90 questions, grouped into five categories: diversity, equity, inclusion, belonging and psychological safety, and leadership and career access. Near-duplicate items have been removed, so no construct is asked twice under two labels.
- What you can and cannot ask employees to self-identify, and the legal checks to run before you collect any demographic item.
- A minimum group-size rule that keeps small demographic segments genuinely anonymous.
What are DEI survey questions?
DEI survey questions are the items an employer uses to measure how fairly and inclusively employees experience work. They fall into two types: optional self-identification questions that capture who someone is, and experience questions that capture how work treats them. The experience questions carry the insight. The self-ID questions only exist to let you segment it.
- Diversity questions ask about representation, whether the workforce reflects a range of backgrounds at every level.
- Equity questions ask whether pay, promotion, and access to opportunity are distributed fairly.
- Inclusion questions ask whether employees experience fair treatment and can voice a contrary opinion day to day.
- Belonging questions ask whether employees feel fully themselves at work and genuinely connected to their team.
A DEI survey question is not the same as a diversity statistic. Representation numbers, headcount by demographic group, come from HRIS data. Perception comes from the survey. The survey tells you whether people feel treated fairly; the HRIS tells you whether the numbers back that up. Both matter. For a full explanation of what DEIB means and how the four pillars work, see what DEIB means.
DEI survey question: a single item on an employee survey that measures either who a respondent is (self-identification) or how fairly and inclusively they experience work (experience). Only experience items are scored. Self-identification items are used only to segment experience scores by demographic group.
Why run a DEI survey in 2026, when DEI programs are under scrutiny?
Because scrutiny raises the evidentiary bar rather than removing it. A DEI survey is now the most direct way to answer a board that asks whether the program does anything. Public support for corporate DEI has fallen to its lowest level since tracking began, which means opinion no longer carries the argument. Your own employee data has to.
Inclusivity scores 8.00 on the 0 to 10 engagement scale, ranking 15th of the 22 engagement drivers. It fell 0.41 year on year, the steepest single-driver decline in the set. Diversity scores 7.91, 17th of 22.
Inclusivity is 8.61 in Healthcare and 6.00 in Creative and Media, the widest gap among the 8 industries that publish an inclusivity score. In between sit Professional Services 8.57, Construction and Heavy Industry 8.23, Technology 8.14, Manufacturing 8.05, Finance 8.01, Education 7.34. An inclusion score only means something read against your own sector.
Those two figures are the shape of the problem a DEI survey exists to find: perceived inclusion is sliding while it sits well below the drivers around it, and the sector you operate in moves the baseline by more than two and a half points. A DEI survey gives a leadership team three things a headcount report cannot: how fairly employees perceive promotion and pay decisions, whether that perception is evenly distributed across groups, and whether the gap is moving. Without the survey, the board conversation is about values. With it, it is about evidence. For context on how the landscape is shifting, see how DEI programs are changing.
DEI survey vs employee engagement survey vs climate survey: which one do you need?
Run DEI questions inside your engagement survey if you have fewer than about 500 employees, and run a standalone DEI survey above that. The engagement survey tells you how people feel about work. The DEI survey tells you whether that feeling is distributed evenly. The climate survey sits between them and measures the current mood of the culture.
| Employee engagement survey | DEI survey | Employee climate survey | |
|---|---|---|---|
| What it measures | How committed and connected employees feel to their work | Whether that experience is distributed fairly across groups | The current mood and tone of the culture right now |
| Typical length | 30 to 50 items | 20 to 30 items | 10 to 20 items |
| Typical cadence | Annual plus quarterly pulse | Annual plus one or two pulses | Ad hoc, or after an event |
| Demographic segmentation | Optional, often coarse | Central to the whole instrument | Optional |
| Who reads the result | Managers and execs | CHRO, Legal, exec team | Exec team |
| Run standalone when | Always | Above roughly 500 employees | After a merger, restructure or incident |
If you are choosing between instruments, the employee climate survey questionnaire and the workplace culture survey both sit closer to the culture end, and neither replaces demographic segmentation.
DEI survey demographic questions: what to ask, what to avoid, and what needs careful handling
You may ask employees to self-identify where every item is voluntary, anonymous, and carries a prefer-not-to-say option. Skip any demographic item you cannot name a reporting or analysis use for. Several data-protection regimes treat race, health, religion and sexual orientation as protected or special-category data that needs a documented lawful basis before collection, so confirm the rules for every location you field in before you add one of those items.
| Demographic item | Verdict | Guidance |
|---|---|---|
| Gender identity | Ask | Voluntary, anonymous, free-text or expanded options plus prefer not to say |
| Race or ethnicity | Ask | Voluntary, anonymous, prefer not to say. Use the category set your region reports on, and only if you will report on it |
| Age band | Ask | Bands, never exact date of birth |
| Tenure band | Ask | Safe and highly useful for segmentation |
| Disability status | Ask with care | Voluntary and anonymous. Treated as protected or special-category data in several regimes, so confirm your lawful basis first |
| Sexual orientation | Ask with care | Voluntary and anonymous. Treated as protected or special-category data in several regimes. High re-identification risk below the threshold |
| Religion | Ask with care | Treated as protected or special-category data in several regimes. Only ask if you have a named analysis use |
| Caregiver status | Ask with care | Useful, but easily re-identifying in small teams |
| Veteran status | Ask with care | Region-specific. Confirm the item exists and is lawful where you are fielding before you use it |
| Exact date of birth | Do not ask | Directly identifying, no analytical gain over a band |
| Health condition or diagnosis | Do not ask | Not a DEI survey item. Belongs to occupational health under separate handling |
| Immigration or visa status | Do not ask | Creates legal and trust exposure with no analytical gain |
| Employee name or ID on an anonymous survey | Do not ask | Ends the anonymity claim entirely |
Self-ID on a survey is optional, anonymous and used for segmentation. Required government reporting, such as EEO-1 component data, is a separate legal obligation collected through HR systems, not through an anonymous survey. Never merge the two.
This is general guidance and not legal advice. Have employment counsel review your final demographic item list before launch.
The 12-question starter DEI survey
Run these 12 items if this is your first DEI survey or you are rebuilding a baseline. They cover all five dimensions, complete in under four minutes, and leave room for the demographic self-ID block without pushing the whole survey past the point where completion rates fall. Each row names the dimension the item measures and what a low score actually tells you, which is the part most question lists leave out.
| Question | Dimension | What a low score signals |
|---|---|---|
| Q01People from a wide range of backgrounds are represented at every level of this organization. | Diversity | Representation thins out above a certain level and the whole workforce can see it. Treat it as a talent planning problem rather than a communications one. |
| Q03I see people who look like me in leadership positions at this organization. | Diversity | A visible ceiling. Expect it to track with low scores on career access, because people calibrate their own prospects against who is already there. |
| Q20People at this organization are paid fairly for comparable work, regardless of their background. | Equity | Either pay outcomes are uneven or the reasoning behind them has never been explained. Check the pay data before you decide which. |
| Q22Promotion decisions at this organization are based on merit and performance, not on who you know. | Equity | Advancement is running on networks. Read it next to the sponsorship items, which usually fall in the same direction. |
| Q25Performance review criteria are applied consistently, regardless of demographic background. | Equity | The review process is the mechanism carrying the gap. Calibration sessions move this item faster than any training programme. |
| Q35I am treated fairly here, regardless of my background or identity. | Inclusion | The single broadest inclusion signal in the set. A low score here with healthy engagement scores means the average is hiding a group. |
| Q36My ideas and contributions are taken seriously at this organization. | Inclusion | Voice is not evenly distributed. Meeting structure and who gets asked first are usually the practical cause. |
| Q43I can raise a concern about unfair treatment without fear of retaliation. | Inclusion | Your reporting channel is not trusted, which means the absence of complaints is not evidence that nothing is happening. |
| Q52I feel like I belong at this organization. | Belonging | The headline retention signal. Segment it before reading it, because a company-level belonging score is almost always the dominant group's score. |
| Q56I feel safe admitting a mistake at work without fear of it being held against me. | Belonging | Low psychological safety. Errors are being hidden rather than fixed, which carries a quality cost as well as a people cost. |
| Q71My manager actively creates an environment where everyone is heard and respected. | Leadership | Segment by manager before briefing anyone. This item varies more between managers than between demographic groups. |
| Q75Managers at this organization are held accountable when their teams show poor DEI outcomes. | Leadership | The programme reads as symbolic. Nothing else on this list will move for long while this item stays low. |
All 12 use the same five-point agreement scale, so they carry forward between cycles unchanged and the trend line stays readable. Once you have a baseline, add the full set for whichever two dimensions scored lowest rather than lengthening the core 12. Every item is drawn from the 90-question bank below.
90 DEI survey questions to ask employees, grouped by category
Grouped into five categories, de-duplicated so no construct is asked twice under two labels. Use a five-point agreement scale throughout and keep the same wording across cycles so scores stay comparable. The survey-length rule is set out once, in the question-writing section below; the starter set above is the 12 items to run first.
Diversity survey questions
These diversity survey questions measure representation, demographic fairness in hiring, and whether employees experience the workforce as genuinely mixed. Use a five-point agreement scale. Pick 5 to 7 of these per survey cycle.
How do you write DEI survey questions employees answer honestly?
Write DEI survey questions in the first person, about the respondent's own experience, on one consistent five-point agreement scale. Avoid double-barreled items, avoid asking people to rate the company's intentions, and never ask a demographic question you will not use.
- Write in the first person about experience, not about the company's intentions. Ask what the respondent perceives, not what the organization means to do.
- Use one five-point agreement scale throughout. Mixing scales in the same survey makes scores incomparable across items.
- Never ask a double-barreled question. One item, one idea. Split any question that uses "and" to join two concepts.
- Never ask a demographic question you will not analyze. If you cannot name the segmentation use before launch, cut the item.
- Include one open-ended question per category maximum. Open text is qualitative context, not a second scoring layer.
- Keep the survey under 30 items and under eight minutes. Completion rate drops sharply after eight minutes. For information on avoiding leading questions, see bias in the workplace.
How do you keep a DEI survey anonymous when some groups have fewer than 10 people?
You set a minimum reporting threshold and refuse to show any result below it. If a demographic segment has fewer than your threshold of respondents, roll it into a broader group or report it only at the total-company level. Anonymity in a DEI survey breaks through cross-tabulation, not through the response itself. For general guidance on anonymous surveys, see anonymous survey questions for employees.
- Set the threshold before the survey opens and publish it. Employees need to know the rule before they answer, not after results disappoint.
- Suppress any segment below the threshold. Do not show the number, do not show the trend, do not show that the segment exists.
- Suppress any two-attribute intersection below the threshold. The intersection of gender and department is almost always smaller than either attribute alone.
- Handle free-text separately. Verbatim comments are either suppressed for manager-level viewers or reviewed before release, because a comment can identify a person that a score cannot.
- Give every demographic question a declared analysis use before the survey opens. Publish the use alongside the threshold rule. An item you cannot explain is an item employees will assume the worst about.
- Tell employees the rule in the survey invitation, not in the results email. The reason employees do not answer demographic questions honestly is almost never the question itself. It is that they can count the number of people like them on their team and do the arithmetic themselves.
All six rules have to hold before any segmented result is published to managers. This is a governance rule set rather than a product claim, and it is the only place on this page where the anonymity rules are stated. Everything else that touches anonymity points back here.
How often should you run a DEI survey?
Run a full DEI survey once a year and a short DEI pulse once or twice between cycles. Annual gives you a stable trend line. The pulse tells you whether an action you took actually moved the item it was meant to move. Add an event-triggered survey after a restructure, an acquisition or a policy change.
| Cadence | Length | What it is for | Best fit |
|---|---|---|---|
| Annual full DEI survey | 20 to 30 items | The stable trend line and the year-on-year comparison | Every organization above 200 employees |
| Biannual | 20 to 30 items | Faster feedback when the program is new and changing | Organizations in the first two years of a DEI program |
| Quarterly pulse | 5 to 8 items | Testing whether a specific action moved a specific item | 2,000+ employees, rotating item subsets |
| Event-triggered | 5 to 10 items | Restructure, acquisition, policy change or a public incident | Any size, when something material changed |
Under 200 employees, run annual only because a pulse on a small population re-identifies people faster than it informs anyone. 200 to 2,000, annual plus one pulse. Above 2,000, annual plus quarterly pulses on a rotating subset of items. For related cadence guidance, see employee pulse check.
How do you run a DEI survey, step by step?
Run it in six steps across roughly 60 days: design and legal sign-off, stakeholder alignment, communication, fielding, analysis, and action publication. The step most teams skip is the one that decides response rate, which is telling employees in advance what will happen to their answers.
- 1Days 1-10Kickoff and survey designPick your items to the length rule in the question-writing section, set the reporting threshold, and get employment counsel sign-off on the demographic list.Owner: People Ops lead plus Legal.
- 2Days 11-25Stakeholder alignment and communication prepBrief the exec team on what the results may show before they see them. Write the pre-launch note that states the threshold rule.Owner: CHRO plus internal comms.
- 3Days 26-40Launch and field the surveyOpen for 10 to 14 working days. Send two reminders.Owner: People Ops.
- 4Days 41-50AnalysisSegment, apply the threshold, look for gaps between groups on the same item.Owner: People Analytics or People Ops.
- 5Days 51-55Action planningOne owned action per confirmed gap, built to the five rules in the action section below.Owner: the function that owns the gap.
- 6Days 56-60Publish and close the loopShare what you found, what you will do, and what you will not do.Owner: CHRO.
How do you read DEI survey results and tell a real gap from noise?
Compare each group's score against the company average on the same item, not against an outside benchmark. A gap matters when it is large, consistent across related items, and stable across two cycles. A single-item difference in one group in one cycle is usually noise, especially when the group is small.
- Start at the total-company score for each item, not at a segment.
- Split by one demographic attribute at a time and compare each group's score against the company average on the same item.
- Look for the same direction across related items, not a single outlier. If one group scores lower on Q20, Q22, Q23 and Q25 all in the same direction, that is a pattern. One item diverging is noise.
- Check that the segment clears your minimum reporting threshold before treating it as a result.
- Check that the gap repeats in the next cycle before you build a program on it.
| Question | Item | Company | Group | Gap | Signal |
|---|---|---|---|---|---|
| Q20 | People at this organization are paid fairly for comparable work, regardless of their background. | 3.8 | 2.9 | −0.9 | Investigate |
| Q22 | Promotion decisions at this organization are based on merit and performance, not on who you know. | 3.6 | 3.0 | −0.6 | Consistent |
| Q25 | Performance review criteria are applied consistently, regardless of demographic background. | 3.7 | 2.8 | −0.9 | Pattern confirmed |
Once the internal gaps are settled, the sector baseline is the next question. CultureMonkey publishes the inclusivity and diversity driver scores, and the same scores split by industry, in the engagement driver benchmarks and the engagement by industry benchmarks, both Q3 2026 Edition.
For the metrics and KPIs that sit alongside survey data, see DEI metrics and KPIs to track alongside it.
How do you turn DEI survey results into action leaders will actually fund?
Publish your actions within 30 days of closing the survey, built to the five rules below. Funding follows accountability. An action with no owner and no measurable item is the reason most DEI survey cycles produce nothing the second year.
- Name the gap in the survey's own words, using the item text as the heading of the action.
- Assign it to the function that owns the relevant process, not to HR.
- Attach the specific survey item it must move, and the target score.
- Set the date it gets re-measured in the next survey cycle.
- Publish what you are not doing and why. Unexplained inaction erodes trust faster than inaction itself.
For ideas on what to do once a gap is confirmed, see DEI activities to run against a confirmed gap.
How should DEI survey questions change for frontline, hybrid and multi-site teams?
Change the delivery, the language and the reference point, not the constructs. A frontline worker with no company email needs a QR code or kiosk route and a shorter set. A hybrid employee needs items that separate proximity bias from manager behavior. A multi-site workforce needs each site's results segmented before any of it is averaged.
| Workforce type | What changes in delivery | What changes in the questions | What changes in reporting |
|---|---|---|---|
| Frontline and deskless, no company email | QR code or kiosk at the gate or break room; no login required | Shorter set, 10 to 15 items; plain language; no jargon | Segment by site and shift before averaging |
| Hybrid | Email or chat; no change needed | Add items separating manager behavior from physical presence | Split by work pattern, not just by team |
| Multi-site or multi-country | Local language per employee; local launch timing | Same constructs, locally reviewed wording | Never publish a single global average without the site split |
| Multilingual within one site | Per-employee language assignment | Reviewed translations, not machine output at send time | Check response rate by language before reading any score |
The most common mistake on a distributed workforce is averaging a multi-site result before segmenting. A company-level DEI score on a workforce spread across four sites and six countries is almost always a meaningless number.
Why do DEI surveys fail, and what are the warning signs?
DEI surveys fail for five reasons: no visible action from the last cycle, demographic questions employees cannot see a use for, groups too small to feel anonymous, wording that reads as a compliance exercise, and no owner for the result. Each has a warning sign you can spot before you launch.
Owner: CHRO
Owner: People Ops
Owner: People Ops
Owner: CHRO
How does CultureMonkey help you run a DEI survey?
CultureMonkey helps you run a DEI survey by enforcing anonymity thresholds automatically, delivering surveys in 100+ languages, supporting QR and kiosk participation for frontline workers with no company email, and surfacing demographic gaps through heatmaps that compare any group's scores against the company average on the same item.
Frequently asked questions
Run your first DEI survey in 60 days
CultureMonkey gives you the anonymity controls, multilingual delivery, and heatmap segmentation you need to turn DEI survey results into decisions leadership will fund.