How a Decision-Ownership Map Cut Accidents 50% in 6 Months

In one PepsiCo South America program, accident ratio fell 50% in 6 months. The result mattered because the change was not a louder campaign or a bigger dashboard, but a decision-ownership map that moved responsibility closer to the field.

On Headline Podcast, the strongest safety conversations keep returning to the same question, which is what actually changes after the meeting ends. Across more than 250 cultural transformation projects, Andreza Araujo saw that leaders can sound committed and still leave the work untouched when nobody owns the next move.

This case study is for leaders who already know that sponsorship matters but want a cleaner test for whether it is real. In A Ilusão da Conformidade, Andreza argues that systems are judged by the behavior they produce when nobody is watching, and that principle matters here because the field only respects ownership that can be seen, checked, and repeated.

1. Why decision ownership is the real lever

A safety program can have meetings, dashboards, and training without changing risk if the next decision remains vague. A decision-ownership map makes the person, deadline, and proof explicit, which is why the companion article on executive sponsorship into decision discipline is the right place to start if you want the sponsor side of the same mechanism.

What most safety blogs miss is that ownership works because it shortens latency. When a concern has to cross too many desks, the work proceeds before the control does, and the field keeps the old exposure even when the room feels aligned.

That is the practical reason the map matters. It is not a culture poster, and it is not an HR exercise. It is a way to force a real role to answer for a real decision whose effect can be checked where the work happens.

The field only changes when the decision has a name attached to it, and that name belongs to someone who can actually move the condition rather than merely comment on it.

2. The initial scenario looked healthy but still carried the same exposure

The program had executive attention, recurring reviews, and enough documentation to look mature. The issue was not the absence of concern. The issue was that the concern kept arriving in the room as a topic rather than leaving the room as a decision.

The field, which is where the risk either changes or repeats, still saw the same exposures because the open item list did not tell supervisors who could change what. That is the gap a dashboard cannot close by itself, which is why the article on board safety dashboard review belongs in the same reading path.

When leaders confuse visibility with control, they create a polite system that keeps the right language and the wrong exposure. The meeting looks orderly. The work remains ordinary.

Andreza has seen this pattern across many projects, and the pattern is always the same: if the next move is not named, the organization quietly decides that nobody really owns it.

3. The decision that changed the work

The decision was to stop treating safety follow-up as a list and start treating it as ownership. Every significant risk item had to leave the review with one named owner, one due date, and one field proof that would show whether the condition actually changed.

That sounds simple until you try it in a real operation, because the first question is not whether the risk matters. The first question is which role can change it without waiting for another meeting, and that is where many well-intentioned programs lose time.

In more than 250 projects, Andreza Araujo saw that one vague owner invites a chain of interpretation, and every handoff creates another place where the signal can fade. By the time the item returns, the energy is gone and the condition is still there.

The better test is whether the owner can point to the control change in the field. If the answer stays abstract, the decision has not yet become operational.

4. How the change was executed

The first move was to translate every open concern into a role that could act, which meant the sponsor stopped asking for updates and started asking for a field change. The second move was cadence, because a weekly review only counted when the decision landed in the work area, where the proof can be seen rather than argued.

The third move was verification. The team did not consider the item closed when someone said it was done. They looked for the altered sequence, the revised authorization step, the removed obstacle, or the changed supervisor decision that showed the exposure had actually moved.

That is also why the companion article on safety meetings into control decisions fits here, because the meeting only matters when it produces a control change that the next shift can feel.

At this point, the work stops being ceremonial. It becomes measurable, which is exactly what leaders want when they say they want accountability, although what they usually need is not more pressure but a clearer decision path.

5. Ceremonial sponsorship and decision ownership are not the same thing

This is the comparison that matters, because the problem is rarely a lack of words. The problem is the distance between the words and the work. The table below shows why one model changes exposure while the other mainly changes how the meeting feels.

That gap also explains why dashboards can mislead leaders. The article on LTIFR distortions shows how a cleaner number can hide a weaker control story, and that is exactly the trap a sponsorship-only culture falls into.

Andreza's own books point in the same direction. In A Ilusão da Conformidade, compliance is only the floor, and in Sorte ou Capacidade, an accident is the late result of failed layers rather than a random event. The ownership map matters because it puts the next layer in the hands of the role that can still change it.

6. What the measured result actually says

In the PepsiCo South America example, accident ratio fell 50% in 6 months. That number matters, but only because it followed a change in the way decisions were owned and checked, not because someone improved the narrative around the work.

The result cannot be reduced to a campaign or a lucky quarter. It is the kind of shift that appears when leaders stop treating concern as a topic and start treating it as a control decision, which is why the number belongs beside the mechanism rather than beside a slogan.

The same logic appears in the article on dashboard staleness, because boards often notice the score before they notice the lag that made the score possible. A faster number is not the same thing as a safer system.

This is the point that boards and plant leaders both need to hear. The metric moved after the control moved, and that sequence is the only one that should calm a leader down.

7. What leaders should copy next

C-level leaders and supervisors can copy the mechanism without copying the exact case. Pick one stubborn risk, name the person who can change it, set the date, and define the proof that will show the field changed. If any one of those four parts is missing, the item is still only a discussion.

That is why a leader should use the same question every week. What changed in the field because someone owned the decision? If the answer is vague, the system is still spending energy on motion rather than movement.

This is also where Safety Culture: From Theory to Practice stays useful, because culture shows up in repeated decisions and not in one-off declarations. A decision-ownership map is simply a way to make that repetition visible enough to manage.

For a board audience, the companion article on board safety dashboard review is the next read, because the board should ask which decision changed, not just which number improved.

Conclusion

The case is simple once you strip away the ceremony. Safety improved when ownership moved closer to the field, because the next decision stopped hiding inside a meeting and started living in a role that could change the work.

If your leadership team needs more conversations like this, stay with Headline Podcast and keep reading the linked articles above. That is the shortest path from a good discussion to a safer operation.