PepsiCo Safety Turnaround: 4 Moves in Six Months

F5 narrative case study for EHS directors, operations leaders, and safety-culture owners

A safety turnaround is not proven by a campaign launch or a new dashboard. It is proven when leadership changes the conditions that shape daily decisions, then tracks whether the result holds. Andreza Araujo's documented PepsiCo South America Foods result gives that claim a concrete test. Under a 180-day plan, the accident ratio fell by 50% in six months.

The number matters, but the sequence matters more. A reduction can disappear when it depends on attention from one executive, a temporary reporting push, or a short burst of field activity. The useful question for leaders is therefore not how to copy a percentage. It is what kind of management moves can make a six-month improvement credible, visible, and repeatable.

This case study does not invent a plant story, a confidential dashboard, or a list of interventions that has not been published. It uses the verified result as a leadership lens, then separates what the evidence supports from what an operation must still test for itself.

Initial scenario: why a good result is not enough

A 50% reduction in accident ratio is an important outcome, yet the ratio alone cannot explain whether the operating system became safer. A lower number may reflect fewer exposures, better controls, changed production volume, improved reporting discipline, or a combination of factors. Leaders who celebrate the result without examining its mechanism risk turning a successful period into a repeatable myth.

ISO 45001:2018 requires organizations to manage hazards, operational controls, worker participation, and improvement as connected parts of an occupational health and safety system. That structure helps interpret the PepsiCo result because it shifts attention from the outcome to the decisions that could have changed the conditions behind the outcome.

Andreza Araujo's book Safety Culture: From Theory to Practice makes a related distinction between declared values and operated values. A company can describe safety as important while its decisions still reward speed, tolerate weak control verification, or leave supervisors without authority to resolve competing demands. The turnaround question begins at that gap.

For a senior leader, the first task is to establish a baseline that includes more than injury counts. The baseline should show where critical exposure exists, which controls are owned, how quickly concerns are resolved, and whether supervisors can change the work when the plan no longer matches reality.

Decision: treat the turnaround as a management system

The central decision in a 180-day turnaround is to stop treating safety improvement as a communications project. Posters, training refreshers, and leadership messages may support a change, but they cannot substitute for decisions about work design, control ownership, supervision, and escalation.

A leadership team that chooses a system approach asks four questions before selecting activities. Which exposures could cause the most serious harm? Which decisions keep those exposures controlled? Who owns each decision in the field? What evidence will show that the control is working under production pressure?

Those questions are more useful than a general request to improve safety culture because they connect the aspiration to authority. The recent Headline analysis of leadership quality before metrics improve develops the same point, namely that leadership quality becomes visible in the choices made before a dashboard turns green.

The decision also needs a time boundary. A 180-day plan creates enough time to change routines and review evidence, while six months remains short enough to expose whether the organization is relying on enthusiasm rather than management discipline.

Execution: four moves that make the result testable

The published PepsiCo result does not provide a public intervention-by-intervention case file, so the four moves below are not presented as an invented reconstruction. They are the management moves that a leader should test when translating a documented turnaround result into an operating plan.

Move 1: put critical exposure before activity volume

Start with the work that can produce fatal or life-changing harm, not with the activity that is easiest to count. A site may report thousands of observations and still lack proof that energy isolation, mobile-equipment interaction, work at height, or contractor interfaces are controlled.

Build a short exposure register that names the task, the credible consequence, the critical control, the control owner, and the evidence required before work continues. The register should be reviewed with operators and supervisors because a control that exists only in a procedure is not yet an operating control.

Move 2: give supervisors a decision they can actually use

Supervisors often receive responsibility without enough authority. They are expected to maintain production, enforce procedures, respond to changing conditions, and escalate unresolved problems, even when the organization has not clarified which tradeoffs they may make.

Define the conditions that require a pause, a redesign, a second verification, or senior escalation. Then review whether the supervisor has the time, staffing, competence, and support to act. A recent article on turning a safety walk into one control decision shows why visible leadership becomes more credible when the walk ends with an owned action rather than a collection of observations.

Move 3: make follow-up part of the control

A concern is not controlled when it is logged. It is controlled when the organization closes the loop, verifies the change, and checks whether the new condition survives the next shift, contractor handover, maintenance window, or production surge.

Assign an owner and due date, but also define the verification evidence. The evidence may be a field check, a test record, a revised isolation boundary, a completed engineering change, or a decision log that explains why the exposure was accepted temporarily. The appropriate evidence depends on the hazard, which is why a single closure percentage is not enough.

Move 4: review the metric with the people who can change the work

Senior leaders should still examine the accident ratio, yet they should not let it become the only conversation. Pair the outcome with leading evidence that explains whether the system is becoming more capable, such as overdue critical-control actions, quality of field verification, escalation age, and repeated exposure at the same interface.

The article on changing the meeting with a better safety metric is relevant because a metric earns its place when it changes a decision, not when it merely fills a reporting slot. If no owner changes course after reviewing the number, the number is descriptive rather than managerial.

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

The verified result is specific. Under a 180-day plan at PepsiCo South America Foods, Andreza Araujo led a 50% reduction in accident ratio in six months. That is a meaningful before-and-after signal, and it is the strongest quantitative fact available for this case.

The result does not, by itself, prove that every site using the same plan would achieve the same reduction. It does not identify a universal percentage for training, observations, inspections, or leadership visits. It also does not remove the need to validate exposure, reporting quality, severity potential, and production context.

The table is a management interpretation, not a claim that the unpublished PepsiCo dashboard contained these exact fields. That distinction protects credibility. Leaders can learn from a verified outcome without pretending to possess evidence that has not been released.

Generalizable lessons from the turnaround

Four lessons travel well across industries, although their application must be adapted to the exposure and authority structure of each operation.

Lesson 1: speed comes from decision clarity

A turnaround slows down when every concern must climb an unclear hierarchy. Clarify who can stop the work, who can accept a temporary deviation, who must fund a redesign, and who verifies that the decision remains safe.

Lesson 2: culture becomes visible under competing pressure

Safety culture is not measured by how people speak during a campaign. It is revealed when output is late, staffing is thin, a contractor is waiting, or a senior customer wants the job finished. The decisive evidence is what leaders permit, protect, and correct in those moments.

Lesson 3: a result needs a mechanism

When a metric improves, ask which mechanism changed. Did the organization reduce exposure, strengthen a barrier, improve competence, shorten escalation, or change the production decision? If the answer is unclear, the improvement is fragile because the organization cannot reproduce what it cannot explain.

Lesson 4: learning must reach the next decision

An investigation, audit, or review creates value only when its finding changes the next work package, permit, staffing decision, design choice, or leadership review. This is why four safety-culture decisions under production pressure matter more than a broad statement of intent.

What to apply in your operation during the next 180 days

Begin by selecting one serious exposure and one management routine that currently shape it. Document the baseline, including the control owner and the evidence required for continuation. In the first 30 days, test whether supervisors can act on the boundary. During days 31 to 90, review recurring deviations and remove the organizational causes that keep producing them. During days 91 to 180, compare the outcome with the control evidence and decide whether the change is stable enough to scale.

Keep the measurement set small enough to use. Pair one outcome metric with three or four explanatory signals, then require every review to end with an owner, a decision, and a verification date. If the meeting produces only commentary, the measurement system is not yet connected to risk control.

Andreza Araujo's A Ilusão da Conformidade, translated as The Illusion of Compliance, is useful for leaders who need to distinguish completed requirements from operating capability. A turnaround is credible when the work becomes safer even after the campaign language fades.

Explore the Headline Podcast for conversations about the leadership decisions that shape safer work.

Conclusion: make the percentage explainable

The PepsiCo result is valuable because it gives safety leadership a measured outcome, a 50% reduction in accident ratio in six months under a 180-day plan. Its deeper lesson is that a number should invite an inquiry into decisions, controls, authority, and follow-up.

Leaders do not need to copy a percentage. They need to build a system in which serious exposure is visible, control ownership is clear, supervisors can act, and metrics change the work. When the result can be explained in those terms, a turnaround becomes more than a successful period. It becomes evidence of management capability.