How 250+ safety culture projects changed the dashboard

Across more than 250 cultural transformation projects supported through ACS Global, Andreza Araujo has seen a pattern that does not appear in most executive dashboards. Companies often measure safety activity with discipline, although the measures still do not tell leaders whether serious exposure is being reduced. The dashboard looks busy, the meeting feels controlled, and the operation continues to teach people that unresolved risk can wait.

That is why this case belongs in safety indicators and metrics. The verified fact is the scale of the experience: 250+ transformation projects, 25+ years in multinational EHS leadership, and work that has reached teams in more than 30 countries. The practical lesson is sharper than a success story. A dashboard changes culture only when it changes what leaders ask, verify, fund and stop.

Initial scenario

The initial scenario in many organizations is a safety dashboard that appears complete because it contains injury rates, training completion, inspection counts, audit scores and observation volume.

Those numbers are not useless. They can show whether the system is active, whether reporting exists and whether the EHS calendar is being executed. The problem is that activity does not prove control. A plant can finish every planned safety conversation and still leave a bypassed guard, a weak isolation practice, a fatigued shift pattern or an aged corrective action untouched for another week.

In projects supported by Andreza Araujo, this gap usually becomes visible when leaders compare declared culture with operating evidence. The company says people can speak up, but weak signals are rare. The company says supervisors own safety, but supervisor field checks are not tied to serious exposure. The company says action closure is disciplined, but closure quality is not verified where the risk actually sits.

The market trap is treating the dashboard as proof of maturity. A dashboard can become a ritual in which leaders review what is easy to count while avoiding what is uncomfortable to decide. That is why the shift from activity volume to control evidence matters.

Decision

The decision behind stronger safety metrics is to stop asking whether the safety system is busy and start asking whether the organization is reducing exposure before harm occurs.

That decision changes the executive conversation. Injury frequency remains visible, although it no longer owns the room. Leaders begin to ask which critical controls are weak, which corrective actions are aging, which high-risk work has not been field-verified, and which repeated deviations are being normalized because production still succeeds.

As Andreza Araujo argues in Safety Culture: From Theory to Practice, culture becomes visible through repeated decisions. A metrics system that cannot show those decisions will keep culture abstract. A stronger system shows where leadership attention is present, where it is absent, and where the organization has learned to tolerate a gap between procedure and work.

This is the same reason safety KPI weighting matters. When every indicator has equal weight, the dashboard rewards volume. When high-potential exposure, control weakness and action aging carry more weight, the dashboard begins to reflect risk rather than administrative effort.

Execution

Execution starts by rebuilding the dashboard around decisions that leaders can actually make during the next operating cycle.

A practical dashboard should show control evidence close enough to the work that managers cannot hide behind averages. For example, a monthly rollup of inspections may look acceptable while one distribution center has repeated pedestrian conflict, one maintenance team has recurring isolation gaps, and one night shift has a pattern of fatigue-related near misses. The average calms the room precisely when leaders need friction.

The stronger approach is to connect four layers of evidence: serious potential, control health, corrective-action quality and leadership response. Serious potential tells leaders which weak signals deserve priority. Control health shows whether barriers are present and respected. Corrective-action quality shows whether fixes removed exposure or only closed a ticket. Leadership response shows whether escalation changed decisions about staffing, maintenance, scheduling or stop-work authority.

In more than 250 projects, Andreza Araujo has observed that organizations often resist this shift because it makes leadership visible. A training number flatters the system. A list of high-potential actions older than thirty days challenges the system, because someone has accepted the delay.

Measured result

The measured result in this case is not a fabricated percentage improvement. It is the repeated finding across 250+ projects that safety dashboards become more useful when they move from lagging counts to control evidence.

That distinction protects the integrity of the article. Some organizations can reduce injuries quickly when leadership rhythm changes, as the PepsiCo South America case shows through Andreza Araujo's verified 50% accident-ratio reduction in six months. Other organizations first need to repair reporting credibility, rebuild trust or separate low-severity noise from serious exposure. A single promised reduction would be false precision.

What can be stated responsibly is that the dashboard starts serving prevention when it shows the condition of controls whose failure could lead to serious harm. James Reason's work on latent failures supports this view, because serious events rarely begin at the final human action. They build through planning gaps, design weaknesses, supervision drift, maintenance pressure and management tolerance, which a late injury metric cannot see in time.

What changed in the dashboard?

The dashboard changed when safety indicators were reorganized around what leaders needed to decide, not around what EHS could easily collect.

Before the shift, the dashboard usually emphasized recordable injuries, lost time, observations, training completion and audit closure. After the shift, leaders still saw outcomes, but they also saw the health of controls tied to serious potential. They saw whether actions were closing fast enough, whether closure was verified in the field, and whether the same risk returned after a supposed fix.

This connects directly with control health versus TRIR and SIF exposure. TRIR can improve while critical controls decay. SIF exposure can rise while total injury rates remain quiet. Control health gives leaders a better view of whether the work is becoming safer or merely less visibly injured.

The table is intentionally practical. If a leader cannot make a decision from the metric, the metric may still be useful for EHS administration, but it should not dominate executive review.

Generalizable lessons

The first lesson is that safety culture cannot be managed through sentiment alone. Surveys can reveal useful perceptions, although repeated operating decisions reveal whether people believe the system will protect them when safety creates inconvenience.

The second lesson is that action aging is a cultural metric. A high-potential action that remains open while production continues teaches the organization that exposure is tolerable. This is why corrective-action aging belongs in the leadership dashboard, especially when actions are linked to serious potential.

The third lesson is that metrics should expose tradeoffs. If the dashboard does not show when staffing, schedule pressure, maintenance backlog or budget delay is weakening a control, the executive team is receiving a partial story. A partial story can be comforting, but it does not govern risk.

The fourth lesson is that EHS should not own the dashboard alone. EHS can design method and maintain technical discipline, but line leaders must own the decisions revealed by the indicators. When the dashboard points to a degraded control, the accountable leader is the one who can change the work.

What to apply in your operation

Start by selecting one high-risk process and rebuilding its dashboard view from the work outward.

Choose a process such as lockout, confined-space entry, working at height, forklift-pedestrian separation or contractor interface. For that process, list the controls that must be healthy before the work begins. Then decide what evidence proves those controls are present, respected and effective. A procedure is not enough evidence. A completed checklist may not be enough either, especially where the same deviation returns after every audit cycle.

Next, test the dashboard with three questions. Which metric would warn us before a serious event? Which metric would force a leader to make a decision this week? Which metric would reveal that our culture is tolerating a gap because the operation still runs? If the current dashboard cannot answer those questions, it is not yet a prevention tool.

For board-level review, connect the same logic to material risk. The article on safety as material risk explains why executives need indicators that show exposure, governance quality and decision accountability. A board does not need more safety noise. It needs fewer, sharper signals that show whether the company is accepting risk without naming it.

Every month spent reviewing activity volume without control evidence allows weak risk signals to age inside the business, while leaders continue to believe the dashboard is telling the whole story.

Conclusion

The lesson from 250+ safety culture projects is that the dashboard is not a neutral reporting object. It teaches leaders what deserves attention, what can wait and what the organization is willing to tolerate.

If the dashboard rewards volume, leaders will ask for volume. If it rewards credible control evidence, action quality and timely decisions, the safety conversation becomes harder and more useful. Headline Podcast exists for that harder conversation, where leadership and safety meet around the evidence that can change real work.