4 Myths About Safety Culture Diagnosis That Leaders Still Believe

Safety culture diagnosis is the discipline of comparing what leaders say, what workers experience, and what the field shows in routine decisions. It matters because a survey score, a slogan, or a single workshop can look convincing while the real culture still decides under pressure.

Across 25+ years leading EHS in multinational operations and more than 250 cultural transformation projects, Andreza Araujo has seen the same pattern repeat. Teams overtrust the easiest signal, then discover too late that the score was only one slice of the truth. During the PepsiCo South America tenure, where the accident ratio fell 50% in six months, the change came from field decisions, not from prettier language.

In Safety Culture Diagnosis: Learn how to do your own, Andreza treats diagnosis as comparison work. That is the thesis here too. Culture becomes visible when the organization checks whether the same rule survives shift change, production pressure, and an uncomfortable conversation.

What leaders need before they diagnose

Before any diagnosis starts, leaders need three evidence streams: perception data, field observation, and decision history. The survey shows how people answer. The walk shows what people do. The decision history shows what leaders reward, tolerate, and fix. If one of those streams is missing, the diagnosis becomes a guess with better formatting.

This is why a good culture review starts by comparing current perception with what the work area actually looks like. A team can give positive answers on a survey and still carry weak handovers, silent contractors, or slow response to bad news. The article on safety climate surveys and their blind spots is useful here because it shows why a score is only the opening question.

A diagnosis also needs a practical owner. If nobody can say who will change the next routine, the conversation becomes descriptive instead of operational.

Myth 1. The survey score is the diagnosis

This myth feels true because numbers are tidy. A percentage looks objective, portable, and easy to compare across sites, so leaders start treating it like a verdict. The problem is that the score only captures what people were willing to say in that moment, in that group, under that social pressure.

In a plant with one strong supervisor and one silent contractor crew, the average can hide the real story. The night shift may answer differently from the day shift, and the maintenance team may answer differently from operations. A company that reads only the average is often looking at the most comfortable slice of the culture, not the most exposed one.

The correction is simple and hard at the same time. Segment the data by shift, function, contractor status, and exposure area, then compare it with field proof. That is how diagnosis becomes evidence. It also keeps leaders from confusing a single favorable score with the real condition described in Safety Culture: From Theory to Practice.

Myth 2. One workshop can fix the culture

This myth survives because workshops are visible. They create energy, a calendar entry, and a photo. They also feel like action, which is why organizations reach for them quickly when the diagnosis is uncomfortable. Yet culture changes in repeated decisions, not in a single event that makes people feel heard for an afternoon.

Across more than 250 cultural transformation projects, Andreza Araujo has seen that the durable shift appears when one routine changes and keeps changing. A handover gets tighter, a briefing names the live risk, a supervisor closes the loop on bad news, or a leader stops rewarding speed when verification is missing. The workshop may help, but it is only the opening move.

If leaders want proof, they should pick one routine and watch it for 30 days. The article on diagnosing compliance theater before the next audit shows the same logic. If the field does not change, the event was ceremony.

Myth 3. One site can stand for the whole company

This myth is attractive because executive time is limited. A strong site or a flagship business unit can look like the company, especially if its leaders are confident and its score is high. The danger is that culture is local. One crew may have direct challenge, while another learns to soften every concern before it reaches management.

That is why diagnosis should happen at the level where risk is actually managed. A safety climate survey at corporate level can help with trend visibility, but it cannot replace observation in the shift where the work happens. The article on the Bradley Curve helps here because it reminds leaders that maturity is not a slogan. It shows up in behavior under pressure.

In practice, leaders should compare one strong area with one weak area, then ask what the stronger area does differently. The answer is usually not a poster. It is a routine, a response, or a leadership habit that the other area has not made normal yet.

Myth 4. Diagnosis is a yearly event

This myth sounds efficient because annual surveys fit planning cycles. The problem is that culture can change faster than the calendar. A supervisor change, a contractor surge, an acquisition, a shutdown, or a period of intense production pressure can shift the way people speak within weeks.

Andreza Araujo's experience at PepsiCo South America is a good reminder. When the accident ratio fell 50% in six months, the pace of change came from disciplined leadership rhythm, not from waiting for the next annual review. If the organization only diagnoses once a year, it usually finds drift after it has already settled in.

A better pattern is to diagnose the high-risk routines monthly and the broader culture quarterly. That rhythm keeps the conversation close to the field and keeps leaders from mistaking stale comfort for stability.

What leaders should do now

Start with one question in three places. What do people say, what do they do, and what do leaders reward? If those three answers do not line up, the diagnosis has already found something important.

Then use the following comparison to keep the next leadership review grounded in evidence rather than impression.

In The Illusion of Compliance, Andreza warns that formal evidence can look clean while the field remains exposed. The safest next step is to put a date on a real field review, use the survey as input rather than verdict, and ask one leader to own the next change. For a deeper diagnostic, talk to Andreza Araujo and ACS Global Ventures.