Bradley Curve vs Hudson Maturity Model vs Safety Culture Diagnosis: which one helps leaders decide?

Leaders often ask which safety-culture tool is best, although the real question is simpler. Are you trying to explain where the organization is, map how far it can move, or uncover what must change in the field? Bradley Curve, Hudson Maturity Model, and Andreza Araujo's Safety Culture Diagnosis answer different questions, so using them as if they were interchangeable usually creates elegant confusion.

Across 25+ years leading EHS in multinational operations, and in more than 250 cultural-transformation projects, Andreza Araujo has seen this mistake many times. A board likes a compact maturity story, a plant manager wants something that leads to action, and an EHS team needs evidence that survives the next production pressure. If the chosen tool cannot support the decision that the audience actually needs to make, the model becomes decoration.

During the PepsiCo South America period, where the accident ratio fell 50% in six months, the useful move was not to make culture language more sophisticated. The useful move was to choose the right level of lens for the decision in front of the team. That is the thesis of this comparison. The Bradley Curve is the fastest language, the Hudson Maturity Model is the most nuanced roadmap, and Safety Culture Diagnosis: Learn how to do your own is the tool that most directly turns culture talk into field evidence.

This comparison fits the same family as safety climate survey versus culture diagnosis versus management review, although the decision here is narrower. That earlier article asked which process fits which governance moment. This one asks which maturity lens helps leaders decide, because a model that cannot guide action quickly becomes a label people repeat without change.

The warning also appears in management review, where leadership can confuse a neat ritual with real control. Culture language has the same risk if the chosen model looks complete but never touches the field.

What decision are you actually trying to make?

Before comparing the three tools, define the decision. A board member usually wants to know whether the organization is still reactive, a site leader wants to know what to change first, and a safety team wants a method that can survive skepticism from operations. Those are related questions, but they are not the same question, and the wrong instrument blurs that difference.

Patrick Hudson's maturity model is useful when leaders need more granularity than a simple four-step story can provide. The Bradley Curve is useful when the organization needs a compact map that people can remember in one conversation. Andreza Araujo's Safety Culture Diagnosis is useful when the leader needs to see the underlying pattern, not just the label, because the next decision depends on what the field is actually doing.

If you use the wrong tool for the decision, two bad things happen. Either the model is too simple and the team feels seen but not guided, or the model is too complex and the room spends half an hour admiring the language instead of changing the work. James Reason's work on latent failures helps here, because the culture conversation only matters when it reveals the decision and condition that keep the gap alive.

Bradley Curve: the fast story leaders remember

The Bradley Curve is a useful starting point because it is simple enough to explain quickly. It groups the culture conversation into four broad stages, which makes it easy for directors, senior managers, and supervisors to remember where the site seems to sit. That simplicity is the strength. People can talk about the curve without needing a long workshop.

The weakness is the same as the strength. Four stages are easy to remember, but they compress reality so much that leaders can start acting as if culture were a neat staircase. Real organizations do not move in a straight line, and a site can look proactive in one area while staying reactive in another. Bradley is a good language for orientation, although it is not deep enough to explain the specific behaviors that keep the system stuck.

In Safety Culture: From Theory to Practice, Andreza Araujo uses maturity language as a discussion starter, not as the final answer. That distinction matters. If the curve is used to open a conversation, it works. If it is used to close the conversation, it becomes a label that hides work, supervision, and design problems.

The Bradley Curve is strongest in board decks, leadership workshops, and culture narratives that need clarity fast. It is weaker when the team must diagnose contractor interfaces, supervisory drift, or the gap between what the site says and what the field actually does.

Hudson Maturity Model: the ladder with more resolution

The Hudson Maturity Model gives leaders more detail than Bradley because it uses five stages instead of four and names the movement from pathological through reactive, calculative, proactive, and generative. That extra resolution helps when a company wants to compare plants, regions, or business units with more precision. It also helps when senior leaders need to see that two sites can both be "better than reactive" while still sitting at very different levels of maturity.

Hudson is especially useful when the organization has already outgrown a one-slide explanation. If leaders are asking why one site has strong procedures but weak voice, or why another site has good metrics but poor field discussion, the model helps separate the conversation into a more workable sequence. It gives structure without pretending that culture is finished just because the labels are neat.

The limitation is that maturity models can still tempt leaders to treat the label as the outcome. A site can call itself proactive and still fail to verify critical controls. It can call itself generative and still punish bad news. That is why the Hudson model should be read as a map of tendencies, not as proof that the system is safe. The map is only useful if it changes what leaders inspect next.

Across 25+ years of executive EHS work, Andreza Araujo has seen Hudson-type language work best when it is connected to real decisions, such as escalation rules, leadership routines, and field verification. Without that connection, the model sounds sophisticated but remains detached from the work that actually creates exposure.

Safety Culture Diagnosis: the instrument that creates work

Safety Culture Diagnosis is different because it is not just a maturity story. It is an instrument that helps leaders gather evidence, compare patterns, and decide what to do next. The point is not to produce a flattering label. The point is to show where leadership routines, field habits, and organizational pressure are reinforcing each other in ways that keep risk alive.

That is why Andreza Araujo's book Safety Culture Diagnosis: Learn how to do your own is the most operational of the three lenses. It is designed for leaders who need a practical method, not just a conceptual map. If a site wants to know why people keep bypassing controls, why supervisors normalize shortcuts, or why surveys sound positive while incidents repeat, diagnosis is the better instrument because it reaches into behavior, evidence, and decision context.

Diagnosis also fits the logic of Safety Culture: From Theory to Practice because culture is not a slogan. It is repeated behavior under pressure, and repeated behavior leaves traces. Those traces show up in meetings, permits, field observations, escalation paths, contractor interfaces, and the way people talk when production is late. A diagnosis looks for those traces instead of assuming the answer from a single score.

If your team wants a practical starting point, Safety Culture Diagnosis: Learn how to do your own is the right anchor because it gives the leader a way to move from opinion to evidence. That matters when the site has already heard too many culture speeches and needs a method that can survive the next review.

Decision matrix

Use the matrix below as a practical way to choose the lens. The goal is not to crown a universal winner. The goal is to match the instrument to the decision, because a board, a plant manager, and a safety team do not need the same amount of resolution.

The pattern is clear. Bradley is best when you need a shared mental picture fast. Hudson is best when you need a deeper maturity map. Diagnosis is best when you need to decide what must change in the system, because it turns culture into something leaders can inspect, question, and act on. That is exactly why a single score is rarely enough for a serious safety conversation.

Which lens fits the board, the plant, and the line?

For the board or executive committee, Bradley or Hudson can help create a common language, although the board should not stop at language. If directors want to know whether the organization is becoming safer in real terms, they need a diagnosis that connects maturity claims to evidence, especially when the business is exposed to serious injury and fatality risk.

For the plant manager or site director, Hudson is often the better intermediate lens because it gives more detail than Bradley without losing the strategic view. But if the plant is already struggling with repeat incidents, hidden shortcuts, or weak escalation, the manager needs diagnosis quickly. A maturity label does not fix a drift problem unless it names what is drifting.

For supervisors and frontline leaders, Bradley is sometimes enough to start the conversation because it is easy to explain. Yet supervisors usually benefit more from the diagnosis method once the conversation begins, because they need to see where leadership routines, work pressure, and field habits are reinforcing each other. The line does not improve because the curve looks good. It improves when the work changes.

If the organization is already confused about whether it needs a survey, a review, or a diagnostic process, the earlier article on safety climate survey versus culture diagnosis versus management review is the right companion. This piece is narrower. It helps leaders choose the maturity lens first, then decide whether they need a broader process around it.

How to combine them without confusing the team

The best sequence is usually diagnosis first, maturity language second, and communication last. Start by identifying what the field is doing and which leadership routines are shaping that behavior. Then use Hudson or Bradley to explain the current position and the direction of travel. The cultural story becomes credible only after the evidence is clear.

That sequence protects the organization from two common traps. The first trap is to start with a beautiful maturity story and then force the field to fit it. The second trap is to run a long diagnostic process and never translate the findings into a narrative that leaders can remember. Andreza Araujo's work across more than 250 projects has shown that the strongest change happens when the instrument and the story support each other, instead of competing for attention.

James Reason's model of latent conditions is still useful here because the important issue is not whether leaders can repeat a maturity label. The important issue is whether the label helps them see the hidden decision that keeps the hazard alive. A culture tool is only serious when it improves judgment under pressure.

What to do in the next 30 days

Start with one site, one audience, and one decision. If the audience is the board, use Hudson or Bradley to frame the discussion, then attach a diagnosis plan that shows how the site will gather evidence. If the audience is the plant team, use diagnosis first so the room sees the actual work, then use Bradley or Hudson to explain where the site sits.

Then do three things. First, write the question the tool is supposed to answer in one sentence. Second, define the evidence that will prove the answer is real. Third, name the decision that follows if the answer is uncomfortable. Those three moves keep the process honest, because a culture tool that cannot lead to a decision is only a presentation aid.

1. Choose one audience and one decision, not three agendas at once.
2. Use Bradley when the organization needs a fast shared picture, Hudson when it needs more resolution, and diagnosis when it needs real action.
3. Test the chosen lens against field evidence, not against how neat the slide looks.
4. Convert the result into one leadership decision, one owner, and one next review date.

Andreza Araujo's own advice is simple. Do not ask the model to do the job of leadership, because leadership is what gives the model meaning. If you want help turning the culture conversation into a method that changes decisions, start with Safety Culture Diagnosis: Learn how to do your own and then move to the broader advisory work at Andreza Araujo.