How a Consumer-Goods Company Cut Accident Ratio by 50% in 6 Months

This case study is for plant managers, EHS directors, and regional leaders who need a result they can repeat, not a slogan they can frame. In PepsiCo South America, the accident ratio fell 50% in six months, and the useful lesson was not the number itself. The useful lesson was that leadership routines, field verification, and weak-signal discipline changed what the organization accepted as normal.

The question behind the result was simple. What has to change when a company already has metrics, committees, and audits, but still sees exposure behaving as if no one owns it? The answer is visible in the way Andreza Araújo describes culture in Safety Culture: From Theory to Practice, because culture appears in repeated decisions under pressure, not in declared values. That is also why this article sits best beside safety culture surveys and leading indicators, which only matter when they change the next decision.

Across 25+ years leading EHS at multinationals, Andreza Araujo has seen the same pattern in consumer goods, supply chain, construction, mining, and industrial operations. The companies that improve fastest do not chase noise. They force a shorter path between weak signals, leadership attention, and field correction.

Initial scenario: why the number was visible but the cause was not

The starting point was a visible problem with an invisible mechanism. The accident ratio gave executives a hard number, but the number did not explain which decisions were feeding it, which routines were failing, or which sites were quietly normalizing risk. That is the trap of many lagging indicators. They show outcome, but they do not reveal the operating rhythm that produces the outcome.

During the PepsiCo South America tenure, where the accident ratio fell 50% in six months, the first insight was that the company did not need a prettier report. It needed a tighter learning loop. As Andreza Araújo argues in Make The Difference: Be a Leader in Health & Safety, the leader's job is to turn care into visible action, and that means deciding what gets checked, what gets challenged, and what gets escalated.

The field reality was more important than the dashboard. If a team can close a month with a good-looking ratio while the same weak patterns remain in permits, handovers, supervision, and corrective-action follow-up, the metric is only describing the past. It is not governing the next shift.

That is why the case is useful for leaders who read the 19-country safety culture scale-up case. In both situations, the real work is not celebrating the score. The work is deciding what the score must force the organization to do next.

Decision: what changed first

The first decision was to stop treating the accident ratio as a monthly commentary item and start treating it as a signal that had to be translated into field action. That sounds small, but it changes the entire governance model. Once leaders agree that the ratio is a consequence indicator, they stop asking who can explain it best and start asking what must change in the work that feeds it.

Andreza Araujo has seen this pattern in more than 250 cultural-transformation projects supported by her team. The strongest change begins when executives become more demanding about how they review the work, not just how they review the number. A manager who only asks for the final slide gets a story. A manager who asks for weak signals, open controls, and field proof gets a decision.

The decision also required a cleaner division of attention. Leaders could no longer let every issue sit at the same level. A minor housekeeping comment did not deserve the same treatment as a weak critical control, a rushed permit, or a recurring deviation in a high-risk task. That distinction matters because the organization only changes when leaders show that some issues are operational noise while others are exposure.

As Andreza Araújo explains in A Ilusão da Conformidade, translated as The Illusion of Compliance, a clean file does not prove a safe field. The PepsiCo lesson matched that view. Compliance artifacts were useful, but they became meaningful only when they triggered a sharper conversation about actual control quality.

Execution: how the routine was made harder to ignore

The execution phase depended on rhythm. The company needed a repeatable pattern that forced leaders to see the field before the month ended. The useful shift was not a campaign. It was a set of review habits that made weak signals impossible to postpone until the next meeting cycle.

One part of that rhythm was leadership presence. Another part was field verification. A third part was follow-up on the actions that were already open. That combination matters because accident reduction rarely comes from one dramatic fix. It comes from many small decisions that make unsafe shortcuts harder to repeat.

The same logic appears in safety culture survey blind spots. If leaders only look at the score, they stay abstract. If they compare the score with the way permits are issued, supervisors respond, and actions are closed, the organization starts to see where the real drift sits.

Execution also had to protect voice. People have to believe that a weak signal will not disappear into a slide deck. When workers see that a supervisor question leads to a field check, and that a field check leads to a correction, reporting becomes more honest. That is the point where a metric begins to change the organization rather than decorate it.

Measured result: what the 50% reduction actually proved

The measured result was a 50% reduction in accident ratio in six months, but the deeper proof was not mathematical. The deeper proof was that the leadership system had become more responsive to the work itself. The number mattered because it showed that the company was no longer waiting for harm to justify action.

The case did not prove that any operation can copy the same percentage on command. It proved something narrower and more useful. When leaders shorten the distance between weak signals and correction, they can change the trajectory of harm faster than they can change culture through slogans. That lesson is consistent with Andreza Araujo's experience across 30+ countries, where the strongest gains appear when field discipline and executive discipline move together.

Before and after tells the story better than the headline alone:

That shift is also why the case belongs near the 250-company culture diagnosis pattern. The value is not the score itself. The value is whether the score forces leaders to do something different with the work.

Generalizable lessons: what leaders can copy without copying the context

The first lesson is that a metric only helps when leaders treat it as a consequence of decisions, not as the decision itself. If managers use the accident ratio to describe performance without looking at permits, supervision, planning, and follow-up, they are reading the rear-view mirror and calling it steering.

The second lesson is that field verification beats executive comfort. In more than 250 cultural-transformation projects, Andreza Araujo has repeatedly seen that leadership teams move faster when they verify what the work looks like, not when they debate how the dashboard should feel. A clean report can coexist with unsafe habits, which is why the field has to stay in the loop.

The third lesson is that compliance language can disguise drift. The company did not need more polished words. It needed a way to see whether the routine was actually changing how people worked. That is why a book like Safety Culture: From Theory to Practice is useful here. It keeps the focus on repeated behavior, not on declarations.

The fourth lesson is that leaders should never confuse one good result with a finished system. The 50% reduction was a strong outcome, but it was only durable because the organization kept testing its own habits. If the routine had slipped back into commentary mode, the result would have lost its value quickly.

What to apply in your operation

Plant leaders can adapt the lesson in 30 days if they keep the scope narrow. The goal is not to redesign the whole management system at once. The goal is to give one important metric a direct route to field action, so the next review can show whether the work changed or only the commentary changed.

Start with one metric, one review cadence, one field check, and one owner. Then make the connection visible. If the metric is accident ratio, the field check might be permit quality, supervisor response, or critical-control verification. If the weak signal is action aging, the field check should show whether the correction touched the actual job or only the document trail.

The practical sequence can stay simple:

If leaders want a stronger companion read, critical control verification shows how to test whether the field is actually protected. That is the right next step when the operation wants a metric to do real work instead of symbolic work.

Conclusion

When a consumer-goods company cuts accident ratio by 50% in six months, the headline is interesting, but the operating lesson is what matters. The change came from leaders forcing a shorter path between weak signals, field verification, and correction. That is the part other organizations can copy.

For practitioners who want the broader playbook behind this kind of result, the book Make The Difference: Be a Leader in Health & Safety offers the leadership lens, while Andreza Araújo's store gives access to the full set of practical books and tools.