The values poster on the wall is not the culture

A building-materials distributor, 260 people, a mid-size Brazilian company (figures in Brazilian reais). At reception, a panel with five values; the first is quality of service. On the last Thursday of the month, with the sales target three percent short of closing, the manager authorizes billing two orders still pending review. Everyone gets the message, and no one needs to say a word.

The following month, the company hires a consultancy to work on culture. It runs a workshop, revises the values, prints a new panel. The last Thursday of the month stays exactly the same.

01The idea: three levels, and only one is visible #

Edgar Schein published Organizational Culture and Leadership in 1985, after formulating the model in an article in Sloan Management Review the year before. The central contribution is a three-level split, which remains the most useful way to talk about the topic without falling into generalities.

Artifacts are what can be seen and heard: the office, the panel on the wall, the org chart, the way people write email, the time people arrive. They're easy to observe and hard to interpret — two places with the same artifact can mean opposite things.

Espoused values are what the organization says about itself: the mission, the code, the founder's speech, the answer given in an interview. They may describe reality or only aspiration, and there's no way to tell from the text alone.

Tacit assumptions are the shared beliefs no one debates anymore, because they've become obvious. These are what actually guide choices under pressure. Schein argued that this level forms from what worked repeatedly in the past: the solution that solved the problem became the right way, and then stopped being perceived as a choice at all.

The gap between the second and third levels already had a name before Schein. Chris Argyris and Donald Schön, in Theory in Practice (1974), separated the espoused theory — what a person says they do — from the theory-in-use, which is inferred from observed behavior. The two often diverge without anyone lying: the divergence is invisible to those inside it.

02Why it still holds up #

Because almost every culture effort happens at the first two levels. The artifact gets swapped, the espoused value gets rewritten, and the tacit assumption stays intact — partly because no one ever put it into words to begin with.

The assumption is learned by watching what the company does when something has to give. At a distributor that bills pending orders at month-end, the assumption is that the target beats quality. At a company where the partner's project never enters the priority queue, the assumption is that the queue is for other people. In neither case did anyone write it down — and in both cases every employee with three months on the job knows it.

It's worth guarding against the opposite reading, which also shows up: concluding that espoused values are worthless. They're not, when used as a verifiable criterion in an actual decision. What's worthless is an espoused value that never costs anything — because what costs nothing informs no one.

03The cost of the unspoken assumption #

At the scale of the 260-person distributor. Explicit assumptions, to rework with your own numbers:

Want to see how this looks inside a real operation? Explore the platform.

The math: 18 orders × 31% gives 5.6 occurrences per month; × R$ 1,400 gives around R$ 7,800 monthly, or R$ 94,000 a year. That's visible money. What's invisible is the lesson the operation learns: the rule holds until the 28th. After that, it's a suggestion. That lesson applies to every other rule too, including the ones no one has tested yet.

A second part of the cost shows up in hiring. New hires take the reception panel as a description of the company and take a few months to discover the real criterion. That mismatch is one of the causes of first-year departures that rarely surfaces in exit interviews — it's a hard subject to say out loud.

04Where this gets stuck in practice #

1. Culture is treated as communication. The poster and the presentation change; Friday's decision stays the same.
2. No one writes down the conflict criterion. When deadline and quality collide, each manager decides alone — and the differing decisions teach that there's no criterion.
3. The exception from the top isn't logged. The order that jumps the queue because it came from leadership shows up nowhere, but everyone saw it.
4. The climate survey measures mood, not assumption. It asks whether people feel valued; it doesn't ask what they'd do with the pending order on the 28th.
5. The espoused value never cost anything. Until there's a case where the company gave up revenue to uphold it, it's decoration.

05What data-driven management answers #

A tacit assumption is hard to change through speeches and relatively simple to make visible through records. When decision, criterion, and exception live in the same place as the work — with date, owner, and reason — the gap between what the company says and what it does stops being a matter of perception and becomes a list anyone can read.

This doesn't replace leadership; it gives it raw material. It's the same logic as teams that report more errors make fewer of them, where what gets measured is observable behavior, not declaration. It also connects to the assumption about people described in Theory X and Theory Y and to the sequence a change needs to sustain itself, in why process change dies in the middle.

At Relevanti, processes, approvals, and decisions get recorded inside the workflow itself — see the modules on the platform and the breakdown by area in solutions. If you want to test it against your own operation, a conversation is worth having.

Your company's culture is already written down. Not on the wall: in the last hard decision someone made without asking for permission.