How to measure company culture comes down to one sequence: gather survey, conversation and behavioral evidence, then compare them. Disagreement between those sources is a finding to investigate, not an error to average away.
The steps below assume the basics are settled. For the underlying concept, start with what is workplace culture, then return here to put a number and a story behind it.
What you need before you start
- A defined question. Decide whether the goal is a baseline, a gap analysis against a desired culture, or a check after a change. The question decides the instrument.
- Access to employees across teams. A survey and conversations only describe the groups that take part, so confirm that each function and level can be reached.
- HR records you can pull. Turnover and absence figures, broken down by team where privacy rules allow, supply the behavioral evidence.
- A named owner and a promise to report back. Employees answer honestly only when they expect the results to be used and shared.
- Time to repeat the exercise. One snapshot has limited value. Plan the next round before the first begins.
The expected outcome is a short written picture of the current culture, the culture people want, and the behaviors that confirm or contradict both.
Steps to assess workplace culture
Step 1: Define what you are measuring
Organizational culture has layers, and each layer needs a different kind of evidence. Edgar Schein’s model, as summarised by Gray Group International, separates artifacts (what you can see and hear), espoused values (what people say they believe) and basic assumptions. Surveys capture artifacts and espoused values well. Underlying assumptions need deeper qualitative work.
That split explains why a lone questionnaire is not a complete workplace culture audit. Write down which layer matters for your question, and use the traits described in what makes a good workplace culture as a reference for the values you expect to find.
Step 2: Choose a culture assessment framework
Pick one framework and keep it for every later round, so results stay comparable. Three established options cover most needs:
- Organizational Culture Assessment Instrument (OCAI). Developed by Kim Cameron and Robert Quinn at the University of Michigan and built on the Competing Values Framework. It assesses six dimensions of organizational culture to identify an organization’s values and fundamental assumptions, and it produces four culture types: Clan (collaborative, people-focused), Adhocracy (creative, innovative), Market and Hierarchy. It captures current and preferred culture in one administration. The Canadian Collaborating Centre for Methods and Tools lists it as the Competing Values Culture Assessment, also known as the OCAI, with the Cameron and Quinn workbook as its source.
- Denison survey. Measures four traits across three indices each, producing a 12-index profile. It has been administered to hundreds of organizations across dozens of industries, which supports benchmarking against others.
- Barrett model. Its Cultural Transformation Tools measure personal values, current organizational values and desired organizational values. It suits a values-focused conversation.
Which should you use? Choose the OCAI when the aim is a compact comparison of the current culture with the preferred one. Choose the Denison survey when outside benchmarks matter. Choose the Barrett tools when values are the central topic. The OCAI has been criticised for oversimplifying cultures that blend multiple types, so read its four types as a starting profile, not a verdict.
Step 3: Run the survey
An employee engagement survey or a framework questionnaire gives the quantitative backbone. A published case study of PT. SGM distributed OCAI questionnaires to all employees to evaluate six cultural dimensions, which is a useful model of full-population coverage.
If you write your own items, Gray Group International recommends these design habits:
- Use Likert scales. Items typically run on five or seven points and capture attitudes and perceptions efficiently.
- Avoid double-barreled questions. An item should never ask about two things at once.
- Add reverse-coded items. They help detect acquiescence bias, the tendency for some respondents to agree with everything.
For ready-made wording, see culture survey questions for employees.
Step 4: Hold focus groups
Focus groups test what the survey numbers mean. Use them to ask why a dimension scored high or low and to hear how people describe daily norms in their own words. Keep each group to one team or level where possible, so that members do not hold back because a manager is present.
Step 5: Hold one-on-one conversations
Individual interviews reach what people will not say in front of colleagues. Because basic assumptions need qualitative work, ask about concrete episodes, such as how the last disagreement was settled or how a mistake was handled, instead of asking people to rate the culture. Take notes during the conversation and keep a consistent set of prompts, so answers can be compared across interviewees.
Step 6: Pull behavioral data from HR records
Turnover and absence figures show what people do, which complements what they say. Compare them across teams and over time, then set them beside the survey results for the same groups. A team that reports strong collaboration yet shows high departures is a pattern to explore in the next round of conversations.
Treat these figures as signals, not explanations. They tell you where to look, and only the conversations tell you why.
Step 7: Observe and audit the workplace
Observation covers the artifacts layer directly: what can be seen and heard in meetings, shared spaces and routine interactions. Use a short checklist built from the values your chosen framework measures, and record what happens, not what you infer. This observation step turns a survey project into a fuller culture audit.
Step 8: Combine the evidence and resolve disagreements
Lay the survey results, interview themes, focus group findings, HR figures and observation notes side by side, organised by team and by framework dimension. Where they agree, report the finding with confidence. Where they disagree, use these rules:
- Current versus preferred is a gap, not a conflict. In the PT. SGM study, the current culture was predominantly clan-oriented with significant hierarchical elements, while employees preferred a shift toward a more competitive, market-oriented culture. Report both and plan around the gap.
- Say versus do is a finding. If espoused values in the survey conflict with turnover, absence or observed behavior, trust the behavior as the stronger signal and use interviews to find the cause.
- Team differences are normal. Because a four-type profile can oversimplify blended cultures, report subgroup patterns instead of one company-wide label.
- Thin evidence needs more evidence. If only one source points to a problem, run a further focus group or a few more interviews before acting.
Step 9: Report and schedule the next round
Share a short summary with the same groups that took part, name two or three priorities, and fix the date of the next assessment. Gray Group International advises building regular assessment into the operational calendar so that cultural drift is caught early.
Checkpoint before you report
Use this table to confirm each evidence type is complete before you draw conclusions.
| Evidence type | What it shows | Check before moving on |
|---|---|---|
| Survey | Artifacts and espoused values, current and preferred culture | All teams and levels reached; items free of double-barreled wording |
| Focus groups and interviews | Reasons behind scores and underlying assumptions | Consistent prompts used; groups kept to one team or level |
| HR figures | Behavior shown in turnover and absence | Compared by team and over time, then matched to survey groups |
| Observation | Visible and audible artifacts | Recorded behaviors, not interpretations |
The evidence layers and survey design points in this table follow Gray Group International’s overview of culture assessment.
Frequently asked questions
How often should a culture assessment run?
Build it into the operational calendar as a recurring exercise, so that cultural drift is caught early instead of discovered after problems appear.
Why does the OCAI use points instead of ratings?
Respondents allocate 100 points across four descriptions for each of six organizational dimensions. This forced-choice design prevents respondents from rating everything as equally important.
What is the difference between a culture audit and an engagement survey?
An employee engagement survey is one evidence source, mainly reflecting artifacts and espoused values. A culture audit adds conversations, behavioral data and observation to reach the deeper assumptions a survey misses.
Do employees care about culture, or only executives?
Both do. A Deloitte finding cited by Gray Group International reports that 94 percent of executives and 88 percent of employees believe a distinct workplace culture is important.
