Employee pulse survey template: 12 questions and an action plan
Copy a short employee pulse survey, keep non-rating answers separate, and turn a small set of responses into one clear follow-up action.
An original question set for team improvement, not a validated employee-engagement instrument or an individual performance rating. The worked results are fictional.
Set a period, a purpose and a response policy
Use this pulse to choose a practical improvement to the way a team works. Set a specific period, such as the past two weeks, and name the person responsible for reviewing the results. Use the same wording and period length next time if you want to compare answers.
Before sending it, explain who can see individual responses, how comments will be shared, when records will be removed and when the team will hear back. A form without a name field is not necessarily anonymous: account settings, collection tools and identifying comments can reveal a respondent. Check the actual collection setup before promising anonymity. For a small team, avoid department or role breakdowns that identify people.
Suggested introduction: ‘We are asking about your work during [start date–end date] so we can choose one improvement. Every question is optional. [Reviewer] can access [describe the actual access]. We will share a summary and next action by [date]. Please leave names and sensitive personal details out of comments. For an issue that needs individual support, use [existing support route].’ Replace every bracket before use.
Copy these 12 employee pulse survey questions
For questions 1–10, use the same five choices: 1 Strongly disagree; 2 Disagree; 3 Neither agree nor disagree; 4 Agree; 5 Strongly agree. Also offer ‘Not applicable’ and ‘Prefer not to answer’ as separate, unscored choices. Leaving an answer blank is allowed. The opening instruction makes the reference period apply to every statement.
| Question | Statement or prompt | Answer format |
|---|---|---|
| 1 · Priorities | I knew which work to prioritize. | Agreement, 1–5 |
| 2 · Workload | My assigned workload was manageable within my scheduled working hours. | Agreement, 1–5 |
| 3 · Tools | I had the tools I needed to complete my work. | Agreement, 1–5 |
| 4 · Information | I could find the information I needed for my work. | Agreement, 1–5 |
| 5 · Help | I knew whom to ask when I was blocked. | Agreement, 1–5 |
| 6 · Feedback | The feedback I received helped me improve my work. | Agreement, 1–5; N/A if no feedback |
| 7 · Coordination | I knew who owned work that depended on another person. | Agreement, 1–5; N/A if no dependencies |
| 8 · Speaking up | I felt able to raise a concern about how our work was organized. | Agreement, 1–5 |
| 9 · Recognition | My contribution to the team's work was acknowledged. | Agreement, 1–5 |
| 10 · Focus | I had enough uninterrupted time for tasks that needed concentration. | Agreement, 1–5; N/A if not relevant |
| 11 · Keep | What is one thing about how we worked that we should keep? | Optional short text |
| 12 · Change | What is one change that would help you do your work? | Optional short text |
Use fewer questions when the decision is narrower
If the team is reviewing handoffs, keep priorities, information, help and coordination, then add the two open questions. A short survey should answer the decision in front of you. Try the draft with a colleague and ask what each statement means to them before sending it to everyone.
Keep response options in a consistent order. Ask one thing per statement; split a question such as ‘I had clear priorities and enough time’ into separate items. Pew and AAPOR's guidance supports clear wording, suitable answer options and pretesting. This particular set of questions is SimilarForm's editorial template, not a questionnaire endorsed by either organization.
Worked example: workload needs a closer look
In this fictional exercise, 24 employees were invited and 18 returned a survey. The survey return fraction is 18 ÷ 24 = 75%. It describes participation in this exercise; it does not establish that the respondents represent everyone.
For the workload question, the 18 responses break down as follows. Count each question separately because the number of valid ratings can differ.
| Workload answer | Count | How to treat it |
|---|---|---|
| 1 · Strongly disagree | 2 | Valid rating |
| 2 · Disagree | 4 | Valid rating |
| 3 · Neither agree nor disagree | 3 | Valid rating |
| 4 · Agree | 5 | Valid rating |
| 5 · Strongly agree | 2 | Valid rating |
| Not applicable | 1 | Keep outside the rating denominator |
| Prefer not to answer | 0 | Keep outside the rating denominator |
| Blank | 1 | Keep outside the rating denominator |
Show the denominator before the percentage
There are 16 valid workload ratings. Seven selected Agree or Strongly agree, so favorable responses are 7 ÷ 16 × 100 = 43.75%. Six selected Disagree or Strongly disagree and three chose the middle response. Report those counts with the percentage and keep the two excluded responses visible. If no one gives a valid rating, report ‘No valid ratings’ rather than zero percent.
Do not combine all ten questions into a single engagement score: this template has not been validated as a scale. One person's different answer changes a small group's percentage substantially. Read the distribution, compare the same question over time and use comments as questions to investigate, not proof of a cause or a reason to identify respondents.
Close the loop with one observable action
A useful fictional follow-up is: ‘Workload answers were mixed: 7 of 16 ratings were favorable, 6 unfavorable and 3 neutral. We will review incoming work against available capacity each Monday for the next two weeks. The team lead owns the trial. At the next pulse we will repeat the workload question and review whether priorities were actually removed when capacity was full.’
The worksheet keeps the signal, owner, action, due date and next review together. Agree on a minimum group size for reporting before collecting answers, based on your team's privacy context; there is no universally safe cutoff. Share an aggregated summary and avoid verbatim comments that would reveal their author. A later score change alone will not show that the trial caused it.