Understanding the unwritten rules of your organisation

To understand organisational culture, investigate what people repeatedly do, what they say matters and what they seem to take for granted. Look especially at moments when priorities collide: a deadline and a quality concern, a customer promise and a capacity limit, or a manager’s request and an employee’s judgement. A values statement is one piece of evidence, not the answer.

This is useful before a change that depends on people acting differently. A new reporting process, for example, will struggle if staff believe that bringing bad news damages their standing. But that belief should be investigated, not assigned to a whole workforce because one meeting felt uncomfortable.

Use Schein’s levels as questions, not labels

Edgar Schein’s cultural model distinguishes artefacts, espoused values and underlying assumptions. MIT Sloan’s account of his work explains these as visible organisational features, stated reasons or values, and deeper taken-for-granted beliefs. The distinction helps you avoid treating what is easiest to see as the whole culture.

Artefacts include meeting routines, workspace arrangements, reports, stories and observable behaviour. Espoused values include the principles people publicly describe as important. Underlying assumptions concern beliefs so familiar that people may not think to explain them. These levels are connected, but the meaning of an artefact is not self-evident.

An open-plan office might reflect collaboration, cost pressure or simply the available building. A silent meeting could indicate agreement, uncertainty, fatigue or fear. Do not jump from observation to diagnosis. Ask what the routine means to those involved and look for occasions when it works differently.

The model is a lens for inquiry, not a personality test or a scale on which one company receives a definitive score. It does not tell you that every difference between groups is a problem.

Begin with one work situation

Consider a fictional print studio preparing to introduce earlier reporting of production problems. Its stated priority is dependable quality. Yet concerns about print samples sometimes emerge only after a job has moved into the production queue. Leaders describe this as a reluctance to speak up.

A useful inquiry starts more narrowly: what happens when a designer or operator notices a possible defect before a deadline? Who sees it first? What is expected of that person? Who can pause the job? What happens to the schedule and to the person raising the concern?

Agree this focus with the sponsor and affected staff. Explain that the aim is to understand the work and improve conditions, not to identify disloyal employees. Our guide to change contracting can help clarify the purpose, boundaries and decision rights before collecting accounts.

Do not promise anonymity you cannot provide. In a small studio, a distinctive story may identify its speaker even without a name. Agree how notes will be handled and report patterns with enough context to be useful but without unnecessary identifying detail.

Collect more than one kind of account

Observe ordinary meetings and handoffs with appropriate agreement. Ask people to reconstruct recent events rather than describe “the culture” in general. “Tell me about the last time a job was paused” invites a more checkable account than “Are people empowered here?”

Look at the artefacts involved: schedules, approval steps, rework records and the information available at the time. Compare what the formal process asks for with what the working day makes possible. A rule requiring early escalation means little if nobody can reach the person authorised to decide.

Include different roles, shifts and lengths of service. A new designer may experience the studio differently from a senior operator with a long relationship with the owner. Those differences are findings to investigate, not noise to remove from an average.

CIPD’s employee-voice guidance connects expressing views with influencing matters at work. Applied here, counting suggestions is insufficient. Ask what happens after a concern is raised and whether people can see a response.

Keep observation and interpretation apart

Use a working evidence record with at least two columns: what was observed or reported, and what it might mean. Add alternative explanations and the next check. GOV.UK’s research-analysis guidance makes this separation explicit. It is equally helpful when examining workplace routines.

Fictional evidence	Possible interpretation	Alternative explanation	Useful follow-up
A sample concern is deferred until after scheduling	Keeping the queue moving has priority	The decision-maker lacks the sample details	Follow how information reaches scheduling
Senior operators pause work more often than newcomers	Informal status affects permission to act	Newcomers see different kinds of jobs	Compare similar situations across roles
Staff privately discuss defects before meetings	Public challenge feels costly	Private discussion helps verify uncertain observations	Ask what makes a concern ready to raise

Avoid converting the middle column into a set of facts. “People value speed over quality” may be a tempting headline, but the evidence could instead show unclear authority, incomplete information or different interpretations of quality. Each would suggest a different response.

Keep exceptions visible. If one team raises concerns promptly, investigate what makes that possible. Perhaps a supervisor explicitly accepts schedule changes, or an experienced colleague helps translate concerns into decisions. An exception can reveal practical conditions a broad cultural label would hide.

Explore the unwritten rule together

Bring a tentative interpretation back to the people who supplied the context. You might say: “We have heard that people are expected to arrive with a solution before raising a defect. When is that true, and when is it not?” This invites correction without presenting a hidden diagnosis as settled truth.

Use systemic questions to explore relationships and consequences. What does the scheduler need from the operator? What does a newcomer think the supervisor expects? What happens when someone raises an incomplete concern? Keep the discussion about observable work rather than presumed motives.

In our fictional studio, staff might explain that an earlier manager criticised interruptions without evidence. The current manager may believe that asking for early concerns has replaced that expectation. The mismatch becomes understandable without claiming that either side is irrational.

Historical explanations remain accounts to check. A persuasive story about the past is not proof of what currently drives every person’s behaviour. Compare it with recent decisions and with people whose experience differs.

Change a reinforcing condition

Choose an adjustment that addresses the identified pattern. The studio might create a brief sample review before scheduling and give a named role authority to pause a job while a concern is checked. The change should specify what information is sufficient, who responds and how the schedule is updated.

That is more concrete than asking everyone to “be courageous.” It changes the conditions under which a concern becomes useful. Still, it is an experiment in this setting, not a guaranteed culture intervention.

The Cambridge Institute for Manufacturing’s open-innovation research highlights organisational culture as an implementation challenge. Its setting differs from this studio, but it reinforces the need to examine organisational conditions alongside a new process. Do not infer a universal effect size from that research summary.

Agree how you will review the adjustment. Look at when concerns become visible, whether they receive decisions, whether avoidable rework changes and whether people experience new burdens. A rise in reported problems may initially reflect better visibility rather than deteriorating work.

Know the boundaries of a culture inquiry

This approach is useful when a change depends on habits, expectations and coordination. It is not a substitute for a formal investigation into harassment, discrimination or other serious misconduct. Use appropriate independent and qualified processes when those issues arise; do not reframe them as an interesting workshop topic.

Other limitations include selective participation, observer effects, political incentives and the difficulty of inferring assumptions. Interviews can miss what people cannot safely say. Surveys can reveal patterns but should not erase differences between groups or replace follow-up inquiry.

Revisit interpretations as conditions change. The point is not to publish a permanent cultural portrait. It is to make one important pattern more understandable, alter a condition that sustains it and see what follows. That connects cultural inquiry with double-loop learning: sometimes the rule behind the behaviour needs attention, not just the behaviour itself.