How PepsiCo South America cut accidents 50% in 6 months

During Andreza Araujo's EHS leadership tenure at PepsiCo South America, the accident ratio fell 50% in six months. That result is not useful because it creates a universal benchmark. It is useful because it exposes a practical rule many companies still avoid: safety culture improves when leaders change the operating system around risk, not only the language used to describe it.

The Headline Podcast audience knows this tension well. Andreza Araujo and Dr. Megan Tranter often bring safety leaders back to real conversations, especially when an organization has plenty of safety activity but weak proof that risk decisions are changing. The PepsiCo case belongs in that conversation because it shows how a visible result can emerge when field routines, management attention and cultural expectations begin to pull in the same direction.

Why does this case matter beyond PepsiCo?

The PepsiCo South America case matters because it shows that safety culture can move within an executive timeframe when the work targets decision quality, not corporate inspiration.

Many culture programs fail because they start with communication. Leaders launch a slogan, refresh a poster, run a training week and ask the safety team to prove engagement afterward. Those actions may support a change, although they rarely create it. If the supervisor still rushes pre-task review, if corrective actions age without consequence, or if a production manager treats risk escalation as an interruption, the culture remains intact under new vocabulary.

As Andreza Araujo argues in Safety Culture: From Theory to Practice, culture becomes visible in repeated decisions, not in declared values. That distinction explains why a six-month improvement is possible only when the organization changes the routines in which those repeated decisions happen.

The case also connects with safety culture diagnosis, because diagnosis should reveal which operating habits keep exposure alive. A culture survey alone can describe sentiment, but field evidence shows whether leaders know where risk is actually being normalized.

Initial scenario

The initial scenario in a fast-moving consumer-goods operation is usually a mixture of disciplined systems, high production pressure and many repeated exposures that look normal because they happen every shift.

Manufacturing, logistics and maintenance teams often carry risks that do not appear dramatic in isolation. A rushed changeover, a pallet movement near pedestrians, a minor lockout shortcut, a missing guard during troubleshooting, or a manual-handling workaround may look manageable to a supervisor who has seen the same routine a hundred times. The danger is that familiarity can reduce attention before it reduces exposure.

Andreza Araujo's 25+ years in multinational EHS roles support a practical reading of that environment. The first task is not to tell people that safety matters. The first task is to find where the organization has learned to tolerate weak controls because operations still run, customers still receive product and the monthly dashboard still looks acceptable.

This is where many leaders misread maturity. They see a trained workforce, a certified management system and a calendar full of safety events, so they assume the culture is ready for a campaign. In reality, the organization may need a sharper operating diagnosis, especially where training is not the answer because the work system keeps rewarding the old behavior.

Decision

The decisive move was to treat accident reduction as a leadership and operating-rhythm problem, not as an isolated EHS communication problem.

That decision changes the center of gravity. The EHS team still provides method, analysis and technical support, but the line organization must own the daily proof of control. Supervisors need to know which behaviors and conditions deserve immediate interruption. Managers need to review whether action closure actually removed exposure. Executives need to ask about weak signals before injury numbers become the only evidence in the room.

During the PepsiCo South America tenure, where the accident ratio fell 50% in six months, Andreza Araujo learned that speed comes from narrowing the field of change. A transformation cannot ask every site to improve everything at once. It has to identify the few routines where leadership attention will change the most risk fastest.

The market trap is assuming that visible activity equals visible felt leadership. A leader who attends a safety day is visible, but felt leadership is tested when the same leader delays a shipment, questions an aged action, funds a control, protects a worker who escalates risk, or refuses a shortcut that the operation has quietly accepted for years.

Execution

Execution depended on changing the recurring moments where leaders saw risk, discussed risk and decided what would happen next.

A practical transformation starts with field routines because culture is reproduced there. Supervisors need a rhythm for pre-task conversations, high-risk work verification and fast correction of unsafe conditions. Managers need a rhythm for reviewing patterns rather than isolated events. Executives need a rhythm for asking why the same exposure remains open after it has already appeared in observations, near misses or corrective-action reports.

In more than 250 cultural transformation projects, Andreza Araujo observes that organizations often underestimate the friction between a new safety expectation and the old operating reward system. If a site praises speed, tolerates incomplete closure and treats dissent as delay, the new safety message loses authority before the month ends.

The PepsiCo case therefore should be read as disciplined execution rather than charisma. A reduction of 50% within six months requires a weekly management cadence, sharper supervisor conversations, visible correction of tolerated deviations and a dashboard that distinguishes real risk reduction from administrative completion.

That distinction is close to the logic in safety KPI weighting. When all indicators have the same weight, leaders can hide behind volume. When the dashboard gives more importance to serious potential, action quality and control health, managers have fewer places to bury the signal.

Measured result

The measured result was a 50% accident-ratio reduction in six months, a pace that only makes sense when reporting quality, operational discipline and leadership response improve together.

A number like this should never be sold as a template promise. If the baseline includes underreporting, a short-term drop can be false comfort. If the baseline reflects real exposure and reporting remains credible, a reduction becomes meaningful evidence that the operating routines changed. The leadership task is to protect the integrity of the number while studying what produced it.

James Reason's work on latent failures helps explain why the result cannot be attributed only to worker attention. Injuries fall sustainably when upstream conditions change, including planning, supervision, maintenance, equipment condition, escalation rules and management tolerance. The final behavior matters, but it is shaped by a system whose signals leaders either correct or normalize.

What changed in the operating system?

The operating system changed when safety stopped living mainly in EHS artifacts and started appearing in management routines that carried authority.

Before a culture shift, many organizations keep safety in parallel channels. The EHS team owns the dashboard, supervisors own production constraints, managers own cost and executives own strategy. Workers quickly learn where the real power sits. If safety does not alter planning, budget, maintenance priorities, staffing and stop-work decisions, it remains a support function rather than an operating principle.

A stronger model puts safety evidence into the same rooms where operational tradeoffs are decided. That does not mean every issue becomes an executive crisis. It means the organization defines which signals require escalation, which controls cannot be degraded without authorization, and which action delays are unacceptable because they preserve exposure.

The link with safety decision rights is direct. Culture improves when people know who can stop work, who can accept residual risk, who must fund a control, and who is accountable when the same risk returns after a superficial fix.

Generalizable lessons

The generalizable lesson is not that every company can cut accidents by the same percentage, but that culture work becomes measurable when it changes how leaders make repeated decisions.

The first transferable lesson is focus. A company should identify the few work routines where risk, frequency and leadership discretion overlap. The second lesson is evidence. Leaders need to see whether controls are present, healthy and respected in real work, not only whether procedures exist. The third lesson is consequence. When managers ignore weak signals, the organization must respond with the same seriousness it would apply to missed production or quality commitments.

Antifragile Leadership (Araujo) is useful here because it treats pressure as a test of leadership quality. A fragile safety culture looks good when work is calm. A stronger culture reveals itself when a deadline is tight, a senior person is watching output, and the supervisor still has enough authority to stop a bad decision.

This is also why the PepsiCo result should not be reduced to a success story. It is a management lesson. The result came from making the organization less dependent on luck, more attentive to weak signals and more disciplined about the routines that translate values into action.

What should leaders apply in their own operation?

Leaders should apply the case by auditing six months of their own operating rhythm, then choosing three routines where changed leadership behavior would remove the most exposure.

Start with the routines that repeat often enough to shape culture. Review pre-task briefings, corrective-action closure, supervisor field time, risk escalation, stop-work response and executive dashboard review. For each routine, ask what leaders currently reward, what they tolerate, what they inspect and what they fund. The answer will show whether culture is being shaped by intent or by habit.

Then compare your internal story with field evidence. If leaders say workers speak up, check whether bad-news reports are increasing in quality. If leaders say supervisors own safety, check whether supervisors have time and authority to correct risk. If leaders say controls are verified, check whether verification happens before high-risk work starts or only after an audit asks for proof.

The table should be uncomfortable. If it only confirms what leaders already believe, the review is too soft. Culture transformation begins when the executive team can name the operating habits that keep exposure alive, including habits that senior leaders benefit from.

Each month without this review allows the old culture to keep teaching the organization what really matters, while workers continue to read leadership priorities through deadlines, funding decisions and the way bad news is received.

Conclusion

PepsiCo South America's 50% accident reduction in six months is not a slogan about culture. It is evidence that safety culture becomes operational when leaders change what they see, ask, verify and decide.

The practical challenge is to stop treating culture as atmosphere and start treating it as a pattern of decisions. If your company wants a different result, choose the routines where risk is reproduced every week, put leadership authority into those routines, and measure whether exposure is actually falling. Headline Podcast exists for that kind of real conversation, where safety and leadership come together to shape better workplaces and better lives.