How PepsiCo South America Cut Accident Ratio 50% in Six Months Without Making the Number the Strategy

When a safety result improves quickly, leaders often celebrate the percentage and move on. That is exactly when the most important question begins. What changed in the way decisions were made, supervised, and verified so the number could move?

PepsiCo South America reduced its accident ratio by 50% in six months under a 180-day plan led by Andreza Araujo. The documented result is important, but the percentage is not a strategy that another operation can copy. The useful leadership lesson is to treat the result as evidence that a defined operating commitment can redirect attention, ownership, and action before the next incident occurs.

Initial scenario: why the percentage was not the starting point

An accident ratio is a lagging signal. It tells leaders that harm has already occurred or that the conditions producing harm have already been tolerated. It is therefore a poor starting point for a leader who wants to change the operation, even though it remains relevant for measuring whether the overall direction is improving.

The starting point is the decision environment around the number. Who can pause work when a control is missing? Which supervisor owns the exposure during a production change? How quickly does an unresolved concern reach a person with authority to change staffing, sequencing, equipment, or resources? These questions expose the management system that the ratio alone cannot show.

That distinction matters because a company can reduce incidents through chance, lower exposure, or a temporary production change without strengthening its safety leadership. A lower number deserves attention, but it does not prove that the organization has become more capable of preventing serious harm. Leaders need to test the mechanism behind the result.

Andreza Araujo's experience across 25+ years of executive EHS work supports a direct position. Safety improvement becomes durable when leadership decisions reach the conditions in which work is planned and performed, rather than stopping at communication, training attendance, or a campaign launch.

Decision: make the result a leadership test, not a reporting target

The PepsiCo result can be read as a leadership test with three parts. First, the organization had to define a time-bounded commitment that made delay visible. Second, leaders had to connect the commitment to operational choices instead of treating safety as a specialist agenda. Third, they had to verify whether the new expectations were present in the work.

A 180-day horizon is long enough to expose weak follow-through and short enough to keep leaders from hiding behind an indefinite improvement program. The timeframe does not create the result by itself. It creates a boundary in which executives can ask whether decisions are moving, whether ownership is clear, and whether field evidence supports the story being told in review meetings.

The leadership decision is therefore not “reduce accidents by 50%.” That sentence describes an outcome. The decision is to create the conditions under which supervisors can identify exposure early, escalate barriers without delay, and confirm that controls remain available after the meeting ends.

This is consistent with the argument in Make The Difference: Be a Leader in Health & Safety, where leadership is treated as a practical responsibility for the work system rather than as a ceremonial presence at safety events. The executive question should be, “What decision will change the exposure this week?”

Execution: move the conversation from activity to control

Many safety plans become crowded with activities because activities are easy to count. Training sessions, inspections, communications, and meetings may all be necessary, yet none of them proves that a critical control works during a real task. The leadership discipline is to connect each activity to a decision and a verification point.

For example, a supervisor briefing has more value when it identifies a task whose sequence, equipment, or staffing needs to change. A field observation has more value when the concern receives an owner and a response date. A management review has more value when leaders examine whether the response changed the work, not merely whether someone closed an action in a system.

James Reason's work on latent conditions helps explain why this matters. The visible unsafe act is only one part of an accident pathway. Design choices, production pressure, maintenance constraints, weak supervision, and unclear authority can align behind the final event. A leadership system that focuses only on the last action leaves the deeper pathway intact.

The execution question is simple but demanding. When the operation says that a risk is controlled, what can a leader observe at the point of work? If the answer is only a document, a dashboard, or a training record, the control has not yet been verified.

Measured result: what the 50% reduction proves

The verified result is specific. During Andreza Araujo's tenure at PepsiCo South America Foods, an accident ratio reduction of 50% was achieved in six months under a 180-day plan. That outcome gives the case real evidentiary value because it connects a named leader, a named business, a defined period, and a measured change.

It does not prove that every 180-day plan will produce the same percentage. It does not identify a universal formula for every country, plant, hazard profile, or workforce. It also does not justify replacing serious-risk controls with a target for total incidents. The result should be used as a leadership reference, not as a promise.

The most defensible interpretation is that a focused management intervention can change safety performance when it alters the operating rhythm around risk. That interpretation remains stronger when leaders separate the outcome measure from the evidence that explains it. They should review exposure, control availability, escalation speed, supervision quality, and the disposition of recurring concerns alongside the accident ratio.

A mature review asks whether the organization is safer or merely quieter. If reporting falls while unresolved concerns, weak controls, or informal workarounds remain, the percentage can create false confidence. The number becomes useful when it is challenged by operational evidence.

Why leaders should not copy the percentage

Copying the result is attractive because it gives a board, a plant manager, or an EHS team a clear headline. It is also dangerous because a percentage can become a pressure mechanism that encourages underreporting or narrow optimization.

Leaders should resist three shortcuts. The first is treating the accident ratio as the sole definition of success. The second is converting the 180-day horizon into a campaign with a closing ceremony. The third is assigning the outcome to EHS while leaving production, maintenance, engineering, procurement, and frontline supervision outside the decision.

These shortcuts matter because safety is shaped by choices made outside the safety department. A maintenance delay can change exposure. A staffing decision can remove recovery time or competent supervision. A production sequence can make a control impractical. A procurement decision can determine whether the safer equipment is available at all.

Leadership becomes visible when the people who control those choices are accountable for the risk they create or remove. The EHS function can provide expertise, challenge assumptions, and verify evidence, but it cannot substitute for operational ownership.

Generalizable lessons from the PepsiCo transformation

The case offers several lessons that can travel without pretending that every operation is identical.

• Use a defined horizon. A time-bounded commitment gives leaders a reason to review movement, remove barriers, and decide what will not wait for the next annual plan.
• Translate the target into decisions. The percentage belongs in the result column. The action column should identify what leaders will change in design, staffing, sequencing, supervision, or control verification.
• Keep lagging data in conversation with field evidence. The accident ratio shows direction, while observations, escalations, and control checks help explain whether the direction is credible.
• Place ownership where work is controlled. If the supervisor cannot change the exposure, the organization must define the escalation route and the leader who can act.
• Protect reporting quality. A target that makes bad news costly can reduce visibility instead of reducing risk. Leaders must reward early disclosure and close the loop on concerns.

These lessons fit Andreza Araujo's broader position in Safety Culture: From Theory to Practice. Culture becomes visible through repeated choices, especially when production, cost, schedule, and safety appear to compete in the same moment.

What to apply in your operation during the next 30 days

Start by selecting one serious exposure that leaders already discuss but have not changed. Do not begin with a broad campaign. Define the task, the people who control its conditions, and the evidence that will show whether the control is present during work.

During the first week, map the decision path. Identify who can pause the task, who can authorize a change, who supplies the required resource, and who verifies the result. If any step depends on an unnamed person or an informal favor, the path is not reliable.

During the second week, review the last three concerns, deviations, or near misses connected to that exposure. Trace what happened after each signal was raised. A concern that received no clear owner is not closed, even when the record says otherwise.

During the third week, ask line leaders to demonstrate the control in the field. The demonstration should include the conditions that make the control difficult, because a control that works only when the operation is ideal is not a dependable barrier.

During the fourth week, review the evidence with operations, maintenance, engineering, and EHS together. Keep the accident ratio in the dashboard, but place it beside the decisions and field checks that make the trend interpretable. Then decide what the next 30-day commitment will change.

A 50% improvement in six months is a meaningful result. Its greatest value is not that it supplies a number to repeat. Its value is that it shows why leaders need a defined commitment, operational ownership, and evidence from the work before they claim that safety performance has changed.

For a deeper leadership framework, explore Andreza Araujo's books and safety resources, including Make The Difference: Be a Leader in Health & Safety.