How a consumer-goods company cut accident ratio 50% in six months by changing follow-up

At PepsiCo South America Foods, a 180-day plan cut accident ratio by 50 percent in six months. The point was not a bigger campaign or a new slogan, because the operation already had enough communication. The real shift was to treat follow-up as a control function, not as EHS paperwork that could wait until the next review.

Across 25+ years in multinational EHS and more than 250 cultural transformation projects, Andreza Araujo has seen that culture changes when leaders stop asking whether the message was sent and start asking whether the field changed. In Safety Culture: From Theory to Practice, she argues that culture is what the organization tolerates under pressure, while Make The Difference: Be a Leader in Health & Safety places the line manager inside the control loop instead of outside it.

This case matters for safety leaders, plant managers, and EHS managers who keep seeing the same weak signals after every meeting. If you want the diagnostic frame that sits behind this case, Safety Culture Diagnosis: 250-Company Case shows why measurement alone does not move behavior, and Leadership Cadence: How 180 Days Cut Accidents gives the leadership-side version of the same lesson.

Initial scenario

The operating context was large and uneven. The PepsiCo South America Foods scope covered 7 countries, 30 factories, and 168 distribution centers, which means the same weakness could repeat in different locations before the organization noticed the pattern. The problem was not a lack of data. It was that the data arrived faster than the decisions that should have followed it.

In that kind of system, teams can confuse activity with control. A report can show many closed points, yet the same unsafe condition returns because nobody owns the correction, nobody verifies it in the field, and nobody asks whether the line manager actually changed the work. That gap is why follow-up becomes a safety issue rather than a clerical one.

Readers who want the operational version of this gap can compare it with How to Run a Toolbox Talk That Changes Field Risk, because the same principle appears there in miniature. If the conversation does not change the next decision, the conversation was incomplete.

Decision

The key decision was to shorten the distance between signal and action. Instead of letting EHS absorb every issue and wait for a later committee, the leadership team made unresolved deviations visible to the people who owned the work. That matters because culture shifts when responsibility moves closer to the place where risk is created.

As Andreza Araujo argues in Safety Culture: From Theory to Practice, measurement matters only when it changes what the organization tolerates. The decision here was to stop treating follow-up as a courtesy and start treating it as a management routine that line leadership had to own, review, and close.

That decision also fits the leadership logic in Leadership Cadence: How 180 Days Cut Accidents, because cadence is not about looking busy. It is about creating a rhythm in which a weak signal cannot sit unanswered long enough to become normal.

Execution

The execution pattern was simple, which is one reason it worked. The operation tightened the loop from detection to ownership to verification. A deviation was not just recorded. It was named, assigned, tracked, and checked where the work happened. That is a small sentence, but in practice it changes who feels responsible when a risk stays open.

This is where many organizations fail. They build more forms, more dashboards, and more layers of review, yet they never make the supervisor or plant leader the visible owner of the next move. Andreza Araujo's point in Make The Difference: Be a Leader in Health & Safety is that leadership is present when someone with authority closes the loop, not when someone in EHS files the note.

The team also resisted the usual trap of turning the response into a campaign. Posters do not remove drift. Messages do not close a gap that still sits in the shift handover, maintenance queue, or supervision routine. That is why the article Safety Culture Scale-Up: 19-Country Case is relevant here too. Scale only works when each site carries the same decision discipline, not just the same vocabulary.

Measured result

The result was a 50 percent reduction in accident ratio in six months, which is the kind of number that tells you the loop changed instead of the language. The organization did not become safer because people talked about safety more often. It became safer because the weak signals had a shorter path to a real decision.

That before-and-after pattern is compatible with James Reason's latent failure logic. The visible event sits at the end of a chain, but the chain only weakens when the organization finds the place where responsibility was delayed or diluted. In this case, the delay was not technical. It was managerial.

Generalizable lessons

The first lesson is that follow-up is a control, not admin. When a leader allows unresolved issues to age, the organization teaches itself that the signal does not matter. Over time, that habit becomes culture.

The second lesson is that line ownership changes speed. Once the manager who controls the work also owns the closure, the issue stops drifting between departments. That is why this case reads less like a communication story and more like a decision-design story.

The third lesson is that the best safety systems reduce latency. A faster response does not guarantee perfection, but it keeps the organization from normalizing the same weakness across multiple shifts, which is where many serious events begin to grow.

The fourth lesson is that culture is visible in what gets closed, not in what gets said. Andreza Araujo's A Ilusão da Conformidade makes that point directly. Compliance can make the organization look orderly, while the real test remains whether the work changed when no one was watching.

What to apply in your operation

If you run a plant, region, or contractor-heavy site, start with the top ten recurring deviations from the last quarter and do not add new metrics until the existing ones have a named owner. Then assign one line leader to each item, set a closure date, and make the next review verify the field condition rather than the spreadsheet row.

That approach works because it forces the management system to answer a practical question. Who can move this today. If the answer is unclear, the organization has found a governance defect, not just a safety issue.

For a broader leadership lens, New Plant Manager in 90 Days: Safety Leadership Plan shows how to build the first operating rhythm, while How PepsiCo Cut Accident Ratio 50% in 6 Months gives the shorter version of this same case for readers who want the headline result first.

What this case proves

This case proves that safety culture changes when follow-up becomes leadership work. Andreza Araujo's PepsiCo South America experience did not succeed because the organization talked more about risk. It succeeded because the organization closed the distance between signal, ownership, and verification, which is the point where culture turns into control.

If you need help turning recurring deviation follow-up into a management routine that actually moves the field, talk to Andreza Araujo.