PepsiCo South America's 180-Day Safety Turnaround: What Makes the Result Transferable

A safety result becomes strategically valuable only when leaders can explain what changed without turning the number into the strategy. Andreza Araujo's professional record includes a documented 50% reduction in accident ratio at PepsiCo South America Foods under a 180-day plan. The deeper leadership question is not how to celebrate that result. It is how to convert a time-bound turnaround into decisions that remain visible when attention moves elsewhere.

This case matters because the operating context was large and distributed. Andreza's PepsiCo role covered seven countries, 30 factories, and 168 distribution centers. The verified result is the reduction itself. The operating details below are not presented as an invented reconstruction of internal PepsiCo decisions. They are a disciplined reading of what the result requires leaders to make true, and what another multi-site organization can test in its own operation.

Initial scenario: a number can improve before leadership has changed

Accident ratio is useful because it gives leaders a visible outcome. It becomes dangerous when the outcome is treated as proof that the system is now reliable. A lower ratio may reflect better controls, fewer exposures, reporting changes, production variation, or a short period in which the organization paid unusually close attention to risk.

The first leadership task is therefore to define the problem behind the number. A regional operation needs to ask where exposure is concentrated, which decisions are delayed, and whether supervisors have enough authority to change unsafe conditions before work continues. Those questions move the conversation from performance reporting to operating discipline.

James Reason's work on organizational accidents helps explain why this distinction matters. Harm appears when several defenses line up in a way that permits failure. A result can improve while latent weaknesses remain, especially when the people closest to the work still have to negotiate production pressure, unclear escalation, or inconsistent follow-up.

The decision: make safety leadership repeatable across sites

A 180-day plan is short enough to create urgency and long enough to reveal whether leaders are changing their routines. The strategic decision is not to launch more activity. It is to choose a small set of leadership behaviors that can be recognized at a factory, a distribution center, and a regional review without losing their meaning.

That decision has three parts. Leaders define the exposures that deserve immediate attention. They clarify who can decide, who must be consulted, and who owns verification. They also establish a cadence in which unresolved conditions return to the agenda until the control is visible in the field.

Andreza's book Make The Difference: Be a Leader in Health & Safety frames safety leadership as an operational responsibility rather than a ceremonial message. The useful implication for a distributed organization is direct. A regional leader cannot personally supervise every task, but can insist that every site uses a recognizable decision logic.

Execution: translate regional intent into local decisions

Scaling a safety turnaround does not mean sending the same checklist to every location. The same leadership expectation must be translated into local work. A food factory, a warehouse, and a distribution route can face different exposures, yet each needs a clear way to stop, escalate, correct, and verify.

A practical translation process starts with one question for each critical exposure. What must a supervisor see before the task begins, and what evidence proves that the control is working? The question is intentionally concrete. It prevents the regional program from becoming a collection of slogans, training completions, and forms that no one uses when the shift is under pressure.

The second question concerns decision latency. When a supervisor identifies a serious gap, how long can the issue remain open, who has authority to change the plan, and what happens when the first response is not effective? A control that depends on goodwill but has no escalation route is not a dependable control.

The third question concerns follow-through. Leaders should be able to trace a concern from identification to decision and then to field verification. This is where many programs lose credibility. The meeting records a commitment, but no one checks whether the work changed. A turnaround becomes durable only when the review rhythm follows the condition into operations.

Measured result: what the 50% reduction proves, and what it does not

Andreza Araujo's documented PepsiCo result is specific. Under a 180-day plan at PepsiCo South America Foods, accident ratio fell by 50% in six months. That is a significant outcome and a credible reason to study the case.

The result does not prove that every organization can reproduce the same percentage, because accident ratios depend on exposure, workforce, reporting practice, baseline conditions, and the quality of execution. It also does not prove that the same program will work without local diagnosis. A responsible leader should resist turning one result into a universal promise.

What the result does justify is a stronger question. Which leadership conditions made a rapid improvement possible, and can those conditions be observed after the initial campaign ends? The answer should be tested through field evidence, not inferred from a dashboard alone.

Generalizable lesson: the metric should open the conversation

The strongest use of the PepsiCo result is not to place 50% on a slide and move on. It is to use the number as an entry point into the operating conditions that produced it. Leaders should ask which decisions became faster, which controls became more visible, and which supervisor behaviors changed when the plan was active.

This approach also protects the organization from false confidence. If the only evidence is a lower injury number, the leadership team cannot tell whether exposure fell or whether visibility fell. The metric needs companions, such as verification quality, overdue critical actions, escalation time, and direct evidence that supervisors are changing the work when conditions change.

Andreza's Safety Culture: From Theory to Practice is useful here because it treats culture as something revealed through repeated decisions, not as a statement printed on a wall. A site that says safety matters but delays a known control under production pressure is showing its operating culture clearly.

Generalizable lesson: distributed operations need a common language

Seven countries and multiple operating formats create a leadership problem that a single-site program can hide. If each location uses different definitions for serious exposure, completion, verification, and escalation, the regional dashboard may compare labels rather than conditions.

A common language does not require identical procedures. It requires shared questions. What can cause irreversible harm? Which barrier prevents that outcome? Who can authorize work when the barrier is unavailable? What evidence will close the issue? These questions allow local teams to adapt while preserving decision quality.

The regional leader's role is to protect the language from dilution. If one site calls an unverified action complete and another requires field evidence, their numbers are not comparable. A mature review makes that difference visible instead of rewarding the easier definition.

Generalizable lesson: follow-up is a leadership behavior

Programs often describe follow-up as administration, although the behavior is more important than the record. When a leader returns to an unresolved condition, asks what changed, and checks the answer where work occurs, the organization learns that safety commitments have consequences.

Follow-up also creates information. A delayed action may reveal inadequate resources, unclear ownership, a design problem, or a conflict between the planned method and the actual task. The leadership team should treat that information as a decision signal. Closing the action without resolving the condition only improves the appearance of control.

Andreza's Safety Culture Diagnosis: Learn how to do your own supports this logic by connecting diagnosis to practical change. The value of assessment lies in the decision it enables, not in the score it produces.

What to apply in your operation during the next 180 days

A company that wants to learn from the case can use a six-month cycle without claiming that it will reproduce the same result. Start by selecting the small number of exposures that could cause irreversible harm, then document the control that must be present before work proceeds. Assign decision rights at the level where conditions are seen, while preserving escalation to leaders who control resources and priorities.

Review the evidence every week at the site level and every month at the regional or executive level. Keep the questions stable, but allow the answers to reflect the work. When an action is overdue, ask what blocked it before asking why it was not completed. When a result improves, ask whether exposure changed and whether the control survives a busy shift.

• Define the critical exposure and the barrier that must be verified.
• Give supervisors a clear escalation route when the barrier is missing.
• Review open conditions until changed work is visible, not until a form is signed.
• Pair outcome measures with control evidence and decision latency.
• Test whether the leadership routine remains active after the campaign loses attention.

For leaders who want a deeper framework, Andreza Araujo's books and Safety School resources connect culture diagnosis, operational leadership, and practical behavior change. The store at loja.andrezaaraujo.com provides the relevant titles for teams that want to build a common language before launching a new safety cycle.

The leadership test after the result

A 50% reduction in six months deserves recognition, but the more demanding test begins afterward. Can leaders explain the result without reducing it to a campaign? Can supervisors make the right decision when production pressure returns? Can sites use a common language while responding to different work?

The PepsiCo case is most valuable when it is read as a leadership-transfer problem. A turnaround becomes an operating system only when its decision rules, verification habits, and escalation paths remain visible after the original 180 days have passed. Safety is about coming home, and that standard is measured by what leaders keep doing when the headline number is no longer new.