How a consumer-goods leader cut its accident ratio 50% in 6 months by changing the weekly leadership cadence

On paper, the consumer-goods network already had dashboards, campaigns, and tidy meetings. In the field, the accident ratio still told the harder truth. Across 25+ years of executive EHS work and more than 250 cultural transformation projects, Andreza Araujo has seen that the first lever is rarely another message. It is usually the weekly decision pattern that leaders keep repeating under pressure.

During Andreza Araujo's PepsiCo South America tenure, where the accident ratio fell 50% in six months under a 180-day plan, the result came from changing the cadence that governed the work. The operation did not become safer because people heard more slogans. It became safer because leaders reviewed the right work, named the right owner, and pushed the field to show evidence instead of reassurance. That is the same thesis she returns to in Safety Culture: From Theory to Practice and The Illusion of Compliance.

Initial scenario

The case sits inside a large consumer-goods operation that worked across 7 countries, 30 factories, and 168 distribution centers. That scale mattered because monthly reassurance could not keep pace with the number of moving parts. A score on a slide was never going to explain whether a supervisor in one site had already drifted from the standard while another site was quietly doing better than the report suggested.

The visible problem was not a shortage of information. The operation already had numbers. The real problem was that the weekly rhythm did not force leaders to confront the field often enough. As Andreza Araujo argues in Safety Culture: From Theory to Practice, culture is not defined by what leaders say in the meeting. It is defined by the decisions that survive production pressure. When the rhythm is weak, the site learns to look calm before it learns to be controlled.

That is why the first question was not how to add another campaign. The first question was which weekly leadership habit had to change so the field could no longer hide behind a clean summary. In practical terms, the answer was a shift from reassurance to verification.

Decision

The decision was to change what leaders reviewed, when they reviewed it, and what they had to prove before leaving the room. Instead of treating the monthly report as the main event, the team moved the real discussion to a weekly cadence where field evidence had to show whether a control was still alive. That made the work less polite, but far more useful.

The review had to answer three questions every time. What changed in the field since last week? Who owns the exposure that still matters? What would force a stop if the control started to drift? Those questions sound simple, yet they are uncomfortable because they expose the gap between a green summary and a real barrier.

This is also where the book The Illusion of Compliance matters. A site can look disciplined while the work itself keeps moving under the old habit. Compliance language often sounds complete, but it does not prove that the next shift inherited a safer condition. Andreza Araujo's point is blunt. If the field did not change, the leadership decision did not go far enough.

Execution

The first move in execution was cadence. Leaders stopped using the monthly meeting as the place where they discovered the truth, and they began using a weekly field review to read the truth sooner. If a control was slipping, the people closest to the exposure had to explain it before the drift hardened into routine. That rhythm is the operational logic behind How to Run a Weekly Critical-Control Review in 30 Minutes.

The second move was ownership. Recurring exposures no longer belonged to the room in general. Each issue had one named owner, one due date, and one verification point in the field. That sounds obvious until production pressure arrives, because shared ownership is usually delayed ownership. Once the leader named the owner, the site could no longer pretend that everybody was responsible and therefore nobody was.

The third move was field proof. The leader had to see the control where the work happened, not where the presentation looked neat. A report that says a problem is closed is not the same thing as a worksite that shows the control still in place. The weekly review therefore asked for evidence that a supervisor, a crew, or a contractor could recognize without reading a slide.

That execution style also fits Andreza Araujo's broader experience in more than 250 cultural transformation projects. The sites that moved faster were rarely the sites that spoke best. They were the sites whose leaders changed the rhythm enough for the field to speak first. That is a practical difference, not a rhetorical one.

Why the result held

The result held because the intervention changed habits rather than slogans. Once the weekly rhythm became normal, leaders had less room to treat drift as a background issue. They had to look at the work, ask for evidence, and name the owner before the issue could slip into next month. The system got less comfortable, which is usually what makes it stronger.

The same pattern also scales better than most campaigns. Andreza Araujo later saw the same logic in multi-country work, including her 19-country Unilever experience, because a cadence can travel across sites while a campaign often stops at the site gate. What repeats across a network is not the poster. It is the routine that line leaders are willing to keep.

James Reason's work on latent conditions is useful here. What looks like a sudden improvement is often the visible result of a hidden pattern being interrupted earlier. When the team starts asking harder questions every week, the weak condition is less likely to survive long enough to become an incident. The site becomes less dramatic because the leadership rhythm is doing its job sooner.

Measured result

The measured result was a 50% reduction in accident ratio in 6 months during the PepsiCo South America tenure. That number matters, but it should not be read alone. The number only became credible because the organization had also changed the way leaders looked at the work. The reduction was the output, not the entire method.

What makes the result useful for other leaders is not the copy of the number. It is the logic that produced it. The site stopped allowing a neat summary to substitute for field truth. Once that happened, the operation had a better chance of catching drift before it turned into injury.

Generalizable lessons

Lesson 1. Cadence is the intervention

The first lesson is that cadence is not a supporting detail. It is the intervention. If leaders only talk about the work once a month, the system teaches people to live with monthly truth. If leaders ask for weekly proof, the system learns to answer sooner. That is why cadence belongs next to control, not beside communication as an afterthought.

Lesson 2. Ownership beats shared concern

Shared concern is easy. Shared accountability is harder. Once one person owns the exposure, the organization can check progress, correct drift, and escalate when necessary. Without ownership, the same issue keeps circulating between meetings while everyone agrees it matters.

Lesson 3. Field proof beats a polished report

A polished report can describe order while the field still carries risk. Leaders need the habit of testing what they were told against what the work shows. That habit is central in The Illusion of Compliance, because the report is only useful when it leads to a change in the work.

Lesson 4. Scale comes from repetition

The PepsiCo South America case matters because the same logic can travel. Andreza Araujo's work across 30+ countries and more than 250 projects shows that a repeatable leadership rhythm is more portable than a one-off campaign. Repetition is what turns one result into a management system.

What to apply in your operation

If you lead a plant, a distribution network, or a regional EHS team, the first step is to identify three recurring exposures that still rely on memory instead of proof. Then assign one owner to each, move the main review to a weekly cadence, and require the field to show whether the control is still in place.

The second step is to stop asking only whether the report is current. Ask whether the next shift can recognize the same control without a reminder from EHS. If the answer is no, the work is still too dependent on the safety function and not enough on line leadership. That is the difference between a managed site and a managed narrative.

The third step is to close the loop in the field. When a leader finds drift, the correction should be visible in the work area, not only in the action log. The operation will change faster when people can see that the leader expects proof, not performance.

If you want the broader book-level logic behind this case, Safety Culture: From Theory to Practice is the closest fit, and Andreza Araujo can help turn the same cadence into a diagnostic and a field routine.

The practical thesis is simple. When leaders change the cadence that governs the work, the accident ratio stops being a surprise and starts becoming a management result.