Running an employee survey: Checklist for getting started

Start with a specific concern. Do you want a broad overview of working conditions? Should a changed allocation of tasks be reviewed? Or do you lack a regular picture of how the team experiences collaboration?

Frame the desired outcome of preparation as a decision, not a request for data. ‘We want to get lots of feedback’ does not yet describe a benefit. ‘We want to decide whether our new handover needs adjusting’ narrows the subject and helps with choosing questions.

Also clarify what has already been decided. If a decision is no longer open, the survey should not suggest participation in it. You can still ask about experiences with implementation. That purpose must then be recognisable in the invitation.

A comprehensive one-off survey can systematically cover many topics. Short recurring check-ins can accompany developments over time. A workshop can help people understand a specific situation together. The formats do not need to be played off against one another, but they serve different purposes.

Choose the format according to the purpose and capacity available to work through results. If no one can interpret results or take responsibility for actions, a larger survey does not solve that bottleneck. A regular rhythm also needs a place in everyday work.

The article Check-in or annual survey compares the differences. For very small groups, also look at Pulse surveys in a small team, particularly because of possible inferences about individuals.

At a minimum, name the person responsible for the purpose, those responsible for organisation and analysis, and the people who can decide on follow-up actions. Depending on the company, IT, data protection, the works council or other responsible bodies may also need to be involved.

Participation rights may be relevant to an employee survey. Section 94 of the German Works Constitution Act (BetrVG) concerns personnel questionnaires, among other things; Section 87 BetrVG regulates various matters subject to co-determination. Which provisions apply to your project depends on its specific design. Not every short survey or survey described as anonymous is therefore automatically exempt from review.

Compile the essential information early: purpose, questions, participants, technical access, analyses, permissions and handling of data. The co-determination page and the existing works council guide provide a starting point. This is general guidance, not legal advice.

Only include questions whose answers can support a useful decision. Check each question: What exactly is being captured? Which period should people refer to in their answers? Is the group being asked actually able to assess the subject?

Avoid combining several subjects in one statement and using unclear terms. ‘Our processes are good and fair’ leaves open which process and which understanding of fairness the response refers to. More specific wording can help without making the questionnaire extensive.

Test the draft with suitable people. Ask them to explain how they understood a question and which situation they had in mind when answering. The GESIS guideline on question wording provides methodological background. If you want to observe a trend, then set the wording and scale deliberately.

Describe specifically what data is collected and who can see what. State possible limitations instead of relying only on the word ‘anonymous’. In small teams, with open comments and additional knowledge, inferences may remain possible.

Explain whether personal access is required and, if so, how it is separated from responses. Also decide which groups may be analysed and how long data is needed. The necessary arrangements belong before sending the survey, not just before presenting results.

A note such as ‘Please do not write names or identifying details in comments’ can be useful. But it does not replace an overall protection plan. Clear information for participants is available under What happens to your responses. In your own communication, only make commitments that your actual method fulfils.

An invitation should briefly explain why you are asking, who should participate, how access works and what happens after the survey. State the planned period and a contact person for organisational questions. Estimates of completion time should be based on your specific questionnaire.

An original text template for preparation could look like this:

We want to better understand how our current workflows are experienced within the team. The survey relates to the announced period. Before it starts, we will explain access, analysis and safeguards. Afterwards, we will discuss the summarised results and decide which change to address. For personal concerns, please use the designated confidential contact.

Add the missing details before sending, particularly access, dates and actual responsibilities. This text is a template for you to adapt, not a finished privacy notice. If participation is voluntary, communicate that information clearly and consistently too.

Review access from different employees' perspectives. Does it work on the intended devices? Do people without regular computer access have a suitable option? Is enough time set aside in the workflow? Technical availability does not automatically mean practical accessibility.

A reminder should state the purpose and deadline without singling people out. If you do not know who has already participated, the message must not suggest that you do. A shared team link is also no proof that each response comes from a different person.

Test the invitation and access route with sample data or a test procedure intended for that purpose. Also check the help available if a link does not work. A clear contact person prevents small access barriers from being misinterpreted as lack of interest.

Reserve a time for the first interpretation and name the responsible person. Decide how the team will learn what follows from the feedback. If several parties are involved, passing on open questions also needs a clear route.

Decide in advance which comparisons are useful and permissible. Extra filters applied only out of curiosity can weaken the information's value and reduce protection. Take particular care not to make small subgroups visible retrospectively.

The guide Analysing an employee survey describes the next steps. A useful endpoint for the first round is a reasoned action with responsibility and a review. The next survey should then answer a relevant question, not just fill the calendar.

Go through the entire process once from a participant's perspective. Do they understand the purpose? Can they assess the questions? Do they know who receives results and what limitations apply? Is it clear who they can contact with problems?

Then check the perspective of those responsible: Are analysis and reporting back planned? Is there scope to make decisions? Is it clear how urgent or personal concerns will be handled? If one of these questions is open, resolve that specific point before sending. Technical setup can continue in parallel, as long as the survey is not launched with unclear information for participants.

EngageBeat is designed for short weekly team check-ins. The 28 days for EngageBeat Team or EngageBeat Fokus start when the first survey opens; setup and preparation do not count. Details are available under trial and how it works. Your chosen paid subscription begins unless you cancel before the trial ends. Your own review remains necessary to assess the legal aspects of your use.