Your First Culture Survey: What to Decide First

As a company grows, what its people experience at work changes too. Teams that once talked directly with the founders now have managers in between, and agreements that used to be worked out at the next desk now have to go through several departments. By the time a company first thinks about a culture survey, several of these changes are usually happening at once. There’s a lot you want to know, but it can be hard to tell what to ask first, or what to do once the results come in.

Your first survey doesn’t have to be a major project that tries to examine every problem in the company at once. Decide which problem you need to look at now and which decision the results will inform, then start with the questions those call for. Here’s what a growing company of 100 to 500 employees needs to decide, step by step, as it prepares its first culture survey.

1. Start With the Decision the Results Will Inform

A culture survey uses questions to examine the rules, relationships, and ways of making decisions that employees experience in their work. You can use it to check whether the values your company promotes actually show up in everyday work, and where people run into trouble collaborating or speaking up. A survey can cover every area broadly or focus on a specific change or problem.

For example, if work keeps bouncing back between teams since you added departments, go one step further than “We want to know how satisfied people are with collaboration.” Decide whether you’ll use the results to realign roles across departments or to clarify who has decision-making authority. A specific decision also narrows down whose experiences you need to ask about, and what kind.

In the planning meeting, it helps to write down one sentence: “Who will decide what, based on these results?” If executives will set priorities and team leaders will change how work gets done, build time into the survey schedule for reading the results together. Gallup also emphasizes setting a clear purpose when preparing an employee survey and following through with action on the results. [1]

When I design a first survey, I check two things up front: whether the current concern started with a specific change, and who can change what once the results are in. If both questions have concrete answers, it’s enough to start with a survey that asks only about experiences tied to that change, without a full company-wide assessment. But if no one can say what the results will change, the report is likely to end up in a drawer no matter how you scope the survey.

2. Scope the Survey to What You Need Now

Satisfaction, engagement, and culture are connected, but they aren’t the same question. Gallup links satisfaction to how well people’s expectations or needs are met, and engagement to how involved and enthusiastic people are about their work and workplace. Culture relates to how work gets done at a company. If you sort out what you actually want to know before choosing a name for the survey, it’s easier to cut the questions you don’t need. [2]

Results on how satisfied people are with their benefits can’t, by themselves, explain how people collaborate. A question like “When someone raises a different opinion, does it get enough discussion?” comes much closer to how employees experience conversations at work. One survey can serve several purposes, but if you combine scores from different questions into a single “culture score,” first define what that number means.

If this is your first survey, you can choose a few topics the company is in a position to take up right away. If your main concerns are collaboration and roles as you grow, or the information leaders provide, build the survey around those experiences. If you need a broad read on where the organization stands overall, look at how validated instruments are structured and what their terms of use are. Even if you write a short set of your own items, the link between each question and the concept you mean to measure should be clear.

Choose respondents to fit the purpose, too. To see experiences shared across the company, ask all employees; to see how a group that went through a particular policy or program responded, ask that group. If you run the survey with a few teams first, read the results as those teams’ experience. Be careful not to treat results from the teams that were easiest to enlist as a reading of the whole company.

3. Keep One Experience per Question

If someone gives a low score to “We communicate well and collaborate effectively,” you can’t tell what was difficult. Results are easier to interpret when you separate information flow from collaboration and ask about situations people can actually picture. Pew Research Center’s guidance on writing survey questions also recommends clear, specific wording and asking about one concept at a time. [3]

A draft tailored to your company might split that item into statements like “I get the context behind decisions in time to do my work” and “It’s clear which department owns what.” These are examples of how to make an item more specific. Before you use items like these, have respondents read them first to check whether they can actually observe the experience in question and what they’ll be thinking of when they answer.

You can pair rating questions with a short open-ended question about what’s behind the score. Setting a time frame and a focus, as in “Over the past month, what moment in cross-department collaboration most needed improvement?”, makes it easier to gather experiences you can use in follow-up conversations. For people with no relevant experience to draw on, offer “Not applicable” and “Unable to judge” as separate options.

4. Explain Up Front Who Will See Which Results

How you handle responses matters as much as the questions. Before you send the survey, explain what information you’ll collect, who will have access to the raw data, what the reports for leaders will cover, and when results will be shared. If you collect information that can link responses to individuals, such as employee IDs or HR data, you must not call the survey “fully anonymous.” Tell people what you collect, why, and how long you’ll keep it, and get their consent. You can also explain the difference between identifying individuals to administer the survey and disclosing individual responses in reports.

In small teams, people can guess who responded even with names removed. A table broken down by department, job level, and tenure, or a comment that mentions a specific project, can be enough of a clue. For example, you might decide to roll groups with fewer than five respondents into the next level up, and to check verbatim comments for identifying details before sharing them. Setting a single minimum reporting threshold doesn’t mean you can assume anonymity holds for every combination of filters.

If your company is small and you’re hesitant to share results widely, you can start by sharing company-wide results and the main themes. Expand to detailed team-level reports once you know how far they can be shared safely and have a clear use for them. If you’re asking for candid answers, these commitments have to show up both in the survey invitation and in the actual reports.

5. Decide Who Reads the Results and What Will Change

When results come in, it’s tempting to start at the bottom of the scores and turn every item into an action item. First, check what experiences the scores and comments actually point to. Even a response like “Roles are unclear” calls for different actions depending on whether the problem is too many channels for work requests or no clear final approver. Work out the questions to check with the relevant departments, then narrow down the action items.

In a first round, what matters is what the company can actually carry out. For instance, if you choose to streamline requests between departments, you can put someone in charge of setting the request channel and the final decision-maker, then try it out in day-to-day work. Separate items that need budget or policy changes from the ones you’ll act on now, and let people know when those will be reviewed. You can draw a line between acknowledging what you heard and promising to meet every request right away.

Low scores aren’t the only thing to look at when choosing what to work on. When I analyze results, I always take a second look at teams with a high average but responses split in two, and at teams with an unusually low response rate. Quite often, the open-ended comments from those teams carried important signals. So I sort next steps not by score ranking but by one question: “Can this team solve it on its own, or does the company need to step in?” Leaders decide the first kind together with their teams; HR gathers the second kind and puts them on the executive team’s agenda for a decision.

You Can Widen the Scope in Later Rounds

Tell employees what you heard, what you’ll change first, and what still needs a closer look. After that, you can briefly ask again whether people have actually experienced the changes you chose. This doesn’t have to wait until your next comprehensive assessment is on the calendar.

As the organization grows and you have more questions, you can expand the areas you measure and how finely you report results by team or department. What your first survey should leave behind isn’t a record of how many questions you asked, but which problem you looked into and what you decided as a result. As that experience builds, the reasons and the scope for asking more broadly next time become clearer too.

References

1. Gallup · Employee Surveys: Types, Tools and Best Practices — Purpose, administration, and follow-up action for employee surveys.
2. Gallup · Employee Engagement vs. Employee Satisfaction and Organizational Culture — How the three concepts differ. Updated July 29, 2025.
3. Pew Research Center · Writing Survey Questions — Clear wording, one concept per question, and keeping wording the same when comparing over time. Sources accessed Nov. 3, 2026.