How PepsiCo South America Reduced Accident Ratio by 50% in 6 Months

PepsiCo South America Foods did not move because someone launched a louder campaign. It moved because a real operating system, spread across 7 countries, 30 factories, and 168 distribution centers, had to show the same discipline in every site, shift, and leadership meeting.

In that context, Andreza Araujo led a 180-day plan that produced a 50 percent reduction in accident ratio in six months. The lesson is not that one company found a magic trick. The lesson is that culture changes when leaders change the cadence of work, the way weak signals are reviewed, and the point at which a manager must stop treating risk as a background issue.

Across 25-plus years in multinational EHS and more than 250 cultural transformation projects, Andreza Araujo has seen the same pattern repeat. Programs fail when they ask workers to care harder, and they succeed when leaders make the operating rhythm visible enough that the field can no longer hide behind habit, convenience, or local excuses.

Initial scenario at PepsiCo South America

The scale mattered because scale creates variation. A site that looks stable on a slide can behave very differently once one shift works under time pressure, another shift inherits a backlog, and a third shift accepts old shortcuts as normal. At that point, the problem is no longer whether people know the rules. The problem is whether the rules survive contact with production.

As Andreza Araujo argues in Safety Culture: From Theory to Practice, culture is not what the company declares in a campaign. Culture is what the organization repeats under pressure. That is why a multi-site business can have one policy and many real cultures, because each plant, warehouse, and distribution center turns the policy into daily practice in its own way.

This is also where James Reason still helps. The visible event is rarely the whole story. Behind it sit latent conditions, weak supervision, bad handoffs, and controls that drift quietly until someone finally collides with the system. The PepsiCo case mattered because it treated those hidden conditions as the real target, not the injury count as an isolated number.

The decision that changed the operating rhythm

The first useful decision was to work inside a 180-day horizon. That was short enough to force accountability and long enough to change habits that had become normal. A long transformation program often becomes a storage room for intentions, while a six-month plan forces leaders to decide what will actually change on the floor, in the meeting rhythm, and in the way follow-up happens after a weak signal.

The second decision was to make leadership behavior part of the intervention. Training still had a role, but it could not be the whole answer, because a class does not change the way a supervisor responds when a job is already late and a shortcut looks tempting. The operating question was more concrete: what do managers inspect, what do they close, and what do they repeat every week?

That is where the own thesis becomes clear. The accident-ratio reduction came from leadership cadence, not from safety messaging volume. The company did not need more words. It needed fewer excuses and more visible decisions, because the field only changes when leaders make the same standard unavoidable across sites.

What execution looked like in the field

Execution in a case like this is rarely dramatic. It is usually repetitive, which is exactly why it works. Leaders had to show up in the places where work was actually happening, ask what had changed since the last visit, and force the conversation toward controls rather than opinions. When a manager does that consistently, the field starts to learn that risk will be examined, not merely admired from a distance.

Across more than 250 cultural transformation projects, Andreza Araujo has observed that consistency matters more than intensity. A single powerful speech rarely survives a bad week. A weekly rhythm, which the workforce sees and begins to expect, can outlast the mood of the moment. That is one reason the PepsiCo result is useful to study: it shows that real transformation depends on repetition that everyone can recognize.

The execution also depended on making managers own the follow-up. If a weak signal was raised, someone had to decide what changed, by when, and how the team would know the change was real. This is where many organizations fail, because they confuse attention with action. A meeting that collects concerns but does not alter work only trains people to speak carefully and expect little.

What the 50 percent result really proves

A 50 percent cut in accident ratio in six months is significant, but it should be read correctly. It proves that the operating system can move quickly when leadership changes the way it manages risk. It does not prove that the work has become safe by default, and it does not prove that one lagging indicator can now stand in for the whole reality of the site.

The point matters because accident ratio is a lagging measure. Bird and Heinrich help explain why the visible event never arrives alone, while James Reason shows why the useful work starts before the event, at the latent failures and organizational choices that make the event possible. If leaders only celebrate the metric, they risk missing the precursors that would have mattered even more.

Inline check. If your operation needs a way to turn impressions into evidence, the process in How to Run a Safety Culture Evidence Review in 90 Minutes shows how to do that without drifting into opinion contests.

Why training alone would not have produced it

Training creates language. It does not create a stable operating rhythm if managers still reward speed over verification. That is why Andreza Araujo is blunt in A Ilusao da Conformidade, or The Illusion of Compliance: an organization can look aligned while the work itself remains unchanged. The forms are signed, the classes are attended, and the system still bends toward convenience when pressure rises.

This is the practical difference between compliance and culture. Compliance says the rule exists. Culture asks whether the rule survives the shift handoff, the production delay, the contractor interface, and the day when a supervisor is tempted to let one exception slide. In a real transformation, leaders stop assuming that more training will fix a management problem that lives in the routine.

That point is easy to miss because training is visible and budget friendly. It also lets organizations feel active without confronting the harder question of authority. If a site keeps sending people to class while the same shortcuts remain untouched, it is not changing the system. It is decorating it.

Three traps that copycats miss

Trap 1. Copying the number without copying the management rhythm. The 50 percent result becomes useless if another company treats it as a slogan. What mattered in the PepsiCo case was not the statistic itself, but the way leaders reviewed weak signals and forced action on a fixed timetable.

Trap 2. Turning a lagging result into proof that the job is done. A lower accident ratio can hide unresolved precursors, especially when field verification is weak. If the next injury appears after the celebration, the company has only learned how to tell a better story, not how to manage risk better.

Trap 3. Replacing leadership with communication. Posters, town halls, and launch events do not change the way a supervisor reacts when production is late. In serious-risk work, the real signal is what the line manager does when the standard collides with schedule pressure. That is the moment culture becomes visible.

These traps are common because they are comfortable. They allow the organization to admire change without making it operational. Andreza Araujo's experience across multinational sites shows the opposite pattern. When leaders keep the same routine, the field keeps the same habits. When leaders change the routine, the field changes with it.

What another operation should apply

The practical lesson for a plant manager, EHS manager, or regional director is simple enough to test in the next quarter. Start with one hard result, one leadership sponsor, and one weekly review that nobody can skip. Then connect the result to precursor events so the team knows it is not only chasing a lower number, but also a safer operating system.

1. Pick one lagging metric that matters, then pair it with 2 or 3 precursor indicators that show whether the work is drifting before harm appears.
2. Assign one senior owner who must review the field evidence every week, because delegated safety without ownership turns into paperwork.
3. Standardize the question supervisors ask when a weak signal appears, so every site reacts to the same risk with the same seriousness.
4. Close each month with a field check that confirms the change is real, not only recorded.

That discipline is what turns a case study into a working model. Across 25-plus years in safety culture, Andreza Araujo has seen that the best programs are the ones that become boring in the right way, because the right questions, the right follow-up, and the right leadership behavior happen so often that nobody needs to improvise them anymore.

For leaders who want that kind of consistency, Andreza Araujo's consulting and keynote work at andrezaaraujo.com can help translate evidence into a routine the operation can actually keep.