Employee DEI Survey Template
People will answer honestly once. If nothing visible follows, the second survey gets polite answers and a lower response rate, and you have spent your credibility to learn nothing. This structure covers the questions worth asking, the anonymity rules that make honest answers possible, and the reporting that has to follow.
See the coursesEverything on this page is yours to copy and adapt. The courses cover the decisions behind it: what to commit to, how to measure it, and what to put in writing.
What yours needs to cover
- ✓ Belonging and psychological safety questions, asked about specific situations rather than in general
- ✓ Questions on fairness in hiring, pay, promotion and development
- ✓ Manager and team-level questions that separate a system problem from a local one
- ✓ Demographic questions kept optional, with an explanation of why each one is asked
- ✓ A minimum reporting threshold so no group is small enough to identify
- ✓ Two open questions, which is where the useful material almost always is
How to use it
- Decide what you will do with the answers, including what you cannot act on, before the survey opens.
- State plainly who reads the responses and how they are anonymized.
- Set the minimum group size for reporting before you see any results, never after.
- Report back within three weeks, including the uncomfortable findings.
- Say clearly what you are not going to do and why, in the same message.
What a survey can tell you and what it cannot
A survey is good at showing where a pattern sits, which team, which level, which part of the employee lifecycle. It is poor at explaining why. That is the job of focus groups and listening sessions, which is why the usual sequence is to survey first and then take the findings into smaller conversations with the people concerned.
Decide before you write a single question which decisions the results could change. If the answer is none, because the budget is spent or the priorities are fixed, run a shorter pulse survey instead, or wait. Asking people for honest answers that cannot change anything teaches them not to answer honestly next time.
Writing questions people can answer honestly
Ask about experiences and specific processes rather than general attitudes. "In the last twelve months, I have been given a fair chance to be considered for stretch assignments" produces a result a manager can act on. "This organization values diversity" produces agreement that tells you very little, because most people agree with it wherever they work.
Use the same agreement scale throughout, keep the core set short enough to finish in about ten minutes, and add two open questions at the end. The open answers usually contain the most useful material in the whole exercise, so budget time to read every one of them rather than only a keyword summary.
Protecting anonymity when groups are small
Anonymity is a design decision, not a promise in the introduction. Set a minimum group size below which no result is shown, and apply it to every cut, including combinations such as department by ethnicity, where two small groups can quickly identify one person. Many organizations use a threshold somewhere between five and ten responses. Agree yours with your privacy lead in advance.
Keep every demographic question optional, explain in one line why each is asked, and include a prefer-not-to-say option. Have the raw file held by someone who does not manage any of the respondents, whether an HR analyst or an external survey provider, and tell employees exactly who that is.
Closing the loop within weeks
Report back to everyone who was invited within about three weeks of the survey closing, while people still remember taking it. Include the uncomfortable findings, what you will do about two or three of them, and what you will not act on this year and why. That last part is what earns the next response rate.
Unit 2 of the course covers choosing between surveys, focus groups and listening sessions and running a session that produces usable evidence. The results then feed the baseline in your strategic plan and the measures on your dashboard.
A filled-in example
| Survey window | Two weeks in October, with time allowed during paid shifts for clinical staff |
|---|---|
| Core experience item | My manager applies the same standards to everyone on the team |
| Fairness item | I understand how decisions about promotion are made here |
| Open question | What is one thing that would make it easier to do your best work here? |
| Minimum reporting group | Eight responses, applied to every cut and every combination of cuts |
| Who sees raw responses | An external survey provider only; the center receives aggregated results |
| Report-back date | All-staff summary within three weeks of the survey closing |
Mistakes to avoid
- Asking about attitudes instead of experiences. Attitude questions produce high agreement and no direction. Ask whether specific things happened to the respondent, in a stated time period, so the answers point to a process someone can change.
- Setting the reporting threshold after seeing results. Choosing the minimum group size once the data is in looks like, and often is, a way to hide or show particular findings. Fix it in the survey plan and publish it in the introduction.
- Making demographic questions compulsory. Forced answers lower trust and increase dropouts, and some respondents will simply choose an answer at random. Make every demographic question optional and explain why each is asked.
- Surveying more often than you can act. A full survey every quarter with no visible change is worse than an annual one with a clear response. Match the frequency to your capacity to act, and use short pulse checks in between.
Employee DEI Survey Template FAQs
What questions should a DEI survey ask?
Ask about experiences and specific decisions rather than about opinions on diversity. "I understand how promotion decisions are made here" tells you something you can act on. "I value diversity" does not.
How do you keep a DEI survey anonymous?
Set a minimum group size below which you never report a figure, keep demographic questions optional, and have the analysis done by someone who cannot connect responses to individuals. Confirm the rules with your legal and privacy teams first.
How often should we run it?
Annually is enough for a full survey, with shorter pulse checks in between if something specific has changed. Running it more often than you can act on the results is how a survey becomes an insult.
More DEI templates
Get every template, plus the training
Ten documents, one course that covers the thinking behind all of them, from the first quote to the contract you hand a commercial client.
Choose your course