How a Consumer Goods Operation Cut Accidents 50% in 6 Months by Clarifying Decision Rights

Decision rights in safety are the explicit rules that tell people who can stop work, who can approve a delay, who can fund a fix, and who must reopen a weak action when the field says the control is thin.

During Andreza Araujo's tenure at PepsiCo South America, the accident ratio fell 50% in six months, according to her public profile and the grounding in Safety Culture: From Theory to Practice. The useful lesson is not that every company should copy the same playbook. It is that safety changes faster when leaders make decision rights visible and hard to bypass.

Across 25+ years in multinational EHS and more than 250 cultural transformation projects, Andreza Araujo has seen the same thing. When authority is vague, the field invents its own rules. In A Ilusao da Conformidade, glossed as The Illusion of Compliance, she warns that paper can look disciplined while the real work still depends on improvisation.

Initial scenario

A consumer goods operation lives under constant pressure to keep lines moving, schedules intact, and output predictable. That environment does not create risk by itself. It creates risk when people are asked to protect production and safety at the same time, while no one has made the decision path plain enough for the front line to use under pressure.

In that setting, the usual safety language sounds responsible but stays too soft. Leaders say the right things, the dashboard stays busy, and yet a supervisor still hesitates when a line needs to pause, an aged action needs to reopen, or a contractor interface needs a stronger check. The problem is not a lack of intent. The problem is that intent has not been converted into decision rights.

Andreza Araujo's 25+ years in multinational EHS support a practical reading of this pattern. A site does not change because people hear more about safety. It changes when the organization makes it easier to take the safe decision than to take the fast one.

Decision rights, not decoration

The first move was to define the few decisions that mattered most. Who can stop the line when the control is weak. Who can reject a weak permit or a rushed handover. Who can spend time or money to restore a barrier. Who can reopen a closed action when field evidence says the fix was cosmetic.

That sounds simple, but it changes the culture quickly because it removes the social fog around authority. A supervisor who knows the stop rule no longer needs permission to think. An operations manager who owns the fix no longer hides behind the EHS team. An executive who accepts reopening a green action shows that the number does not outrank the field.

As Andreza Araujo argues in Safety Culture: From Theory to Practice, repeated decisions reveal culture more clearly than speeches do. Decision rights make those repeated decisions easier to see, which is why they matter more than another campaign on motivation or awareness.

Execution in the field

Execution began where risk repeated. Leaders reviewed the tasks that came back every shift, the actions that aged without consequence, and the moments where people had learned to make do instead of asking for a better control. The point was not to create more meetings. The point was to make each meeting end with one clear decision and one named owner.

The operating rhythm was narrow enough to survive production pressure. A supervisor escalated a weak condition before it became normal. A manager verified the field condition before closing the issue. An executive sampled closed actions and asked for proof that the control changed in the work, not only on a form. That is where James Reason is useful, because latent conditions shrink only when leaders change the conditions that make the shortcut attractive.

In more than 250 cultural transformation projects, Andreza Araujo has seen that teams improve faster when the review question becomes specific. Not, "Are we still safe?" but, "Which decision changed the risk this week?" That question is harder to fake, and it helps separate real control from polite reporting.

It also keeps training in the right place. Training can support a decision, but it cannot replace the authority to make one. A permit refresher will not help if the supervisor still lacks the power to stop a weak job. A toolbox talk will not help if the line manager is rewarded for pushing through the same exposure again.

Measured result

The measured result was a 50% accident ratio reduction in six months. That matters because the result came from a management change, not from a louder slogan. Decision rights made the safe choice easier to take, which made the unsafe shortcut easier to refuse.

The result should not be sold as a universal template. Different sites have different exposures, staffing patterns, and operating rhythms. What does travel is the principle. When leaders clarify who can stop, who can fund, who can reopen, and who must verify in the field, the system becomes less dependent on individual heroism.

That is the same tension Andreza Araujo returns to in A Ilusao da Conformidade. A company can look disciplined while still depending on improvisation at the point of risk. The PepsiCo South America result is useful because it shows that discipline is not a slogan. It is a decision structure that the line can actually use.

What the market usually gets wrong

The market usually starts with training, posters, or a leadership message. Those steps are not wrong, but they are incomplete because they explain the right behavior without changing the cost of choosing it. If the faster option still wins under pressure, the organization has not changed safety. It has only changed the wording around safety.

The second mistake is to treat executive support as enough. Support matters, although support without authority is soft. The line needs to know which decision can be delayed, which action can be reopened, and which control cannot be traded away for convenience. Without that clarity, people learn to protect output and hope that the audit will not notice.

The third mistake is to confuse visible activity with control. A site can have many meetings, many observations, and many closeouts while the same risky choice remains available. In Safety Culture: From Theory to Practice, the point is repeated in different forms. Culture is what the organization repeatedly allows, not what it repeatedly says.

Generalizable lessons

The first lesson is that decision rights reduce delay. When the person closest to the hazard knows what they can stop, the organization loses less time arguing about authority and gains more time correcting the exposure itself.

The second lesson is that decision rights improve honesty. People tell the truth more easily when they know the answer will lead to an action, not to blame for creating inconvenience. That is why the same reporting channel can look weak in one company and strong in another, depending on what happens after the report.

The third lesson is that decision rights protect executives from false comfort. If a senior leader samples only green dashboards, the leader may never see the weak closure or the tolerated deviation that kept the line moving. A clear decision path makes those gaps harder to miss.

The fourth lesson is practical. More than 250 cultural transformation projects taught Andreza Araujo that companies do not need perfect systems to improve. They need fewer decisions left vague, fewer controls left unowned, and fewer moments where the safest answer depends on personal courage alone.

What leaders should apply in 30 days

Start with the top three decisions that repeatedly shape exposure. In most operations, those decisions are stop work, barrier restoration, and closure quality. Write who owns each one, who can challenge it, and what field evidence is required before the issue can close.

Then test the structure in one area with real pressure. Pick a line, shift, or maintenance routine where delay is common and ask the people there what they believe they are allowed to stop. If the answers are unclear, the authority is still decorative.

Finally, review the next 20 closed actions and sample the evidence behind them. If the proof is only a photo, a signature, or a training record, the decision structure is still weak. If the proof shows a changed control in the field, the operation is beginning to own safety in a practical way.

Why this case matters now

This case matters because the pressure on operations has not become gentler. Tight schedules, mixed contractor interfaces, and higher expectations for speed make vague authority more expensive than it used to be. If leaders do not clarify who can stop, fund, reopen, and verify, the field will improvise, and improvisation is a poor control model for serious exposure.

Andreza Araujo's work, including Safety Culture: From Theory to Practice and A Ilusao da Conformidade, points in the same direction. Safety changes when leaders alter the decision structure around risk. The PepsiCo South America result is useful because it shows that a 50% reduction in six months is not a mystery when decision rights, field evidence, and leadership follow-through move together.