Safety Dashboards: 5 Executive Questions That Expose Metric Theater

A safety dashboard should help leaders read control, exposure, and drift. It becomes metric theater when the slide answers the meeting but not the worksite, because the number is treated as proof while the field evidence that should challenge it stays out of view.

This article is for executives, EHS leaders, and board members who need the monthly review to show whether risk moved, not just whether the graph looked clean. If you already feel that gap, the companion articles on safety dashboards and blind spots and metric ownership show how the same number can look disciplined while it still fails to govern the work.

Across 25+ years in executive EHS and more than 250 cultural transformation projects, Andreza Araujo has seen the same pattern repeat. In Safety Culture: From Theory to Practice and A Ilusão da Conformidade, the message is steady. Repeated decisions define reality more than polished reports, and a dashboard only matters when it pushes leaders back to the field.

Why the monthly dashboard is not enough

The monthly dashboard is useful, but it is still only a filtered view of the work. A site can show green numbers while contractors bypass steps, supervisors rely on habit, and weak signals never reach the room. The problem is not the existence of the report. The problem is the belief that a report can replace verification.

James Reason's work on latent failures matters here because the dangerous conditions usually sit below the visible score. The organization sees the number first, then builds a story around it, yet the real risk often lives in planning, supervision, maintenance, or permit discipline. That is why a good dashboard has to trigger a second look rather than a quick celebration.

This is also where safety indicators explained becomes a useful bridge. If the team does not separate lagging, leading, precursor, and control checks, the review collapses into one blurred signal. Once that happens, the meeting feels informed while the work continues to drift.

Andreza Araujo learned this in practice during the PepsiCo South America turnaround, where the accident ratio fell 50% in six months under a 180-day plan. The lesson was not that numbers are useless. The lesson was that numbers only matter when leaders ask better questions than the chart itself can answer.

Question 1: what changed in the field since the last review?

The first question is the simplest and the hardest. What changed in the field since the last review? Not what was reported, not what was closed in the spreadsheet, and not what looked good in the deck. The question is about the worksite itself, because that is where risk either moved or stayed put.

If the answer is nothing, the organization has to decide whether that means control is stable or learning is absent. A mature review can name which one is true. A weak review pretends both are the same, which is how static routines stay invisible for months.

A board or executive team should ask for three forms of proof. First, a field observation that shows the control in use. Second, a record that shows the control owner followed the rule. Third, a short explanation of what would have been different if the control had failed. Without those three elements, the review is describing activity rather than change.

That is why the article on safety dashboards and blind spots matters to leaders who want more than a clean trend. The metric may improve, but the worksite must be able to show why. If it cannot, the number is only a shadow of the operation.

Question 2: which signal disappeared when the target turned green?

Zero-looking targets often create the illusion that nothing is wrong. In practice, the most important question is which signal disappeared when the line turned green. Did near misses stop being reported? Did first aid stay informal? Did supervisors stop writing down the awkward cases because they did not want to break the streak?

Andreza Araujo addresses this risk in Muito Além do Zero, or Far Beyond Zero. The thesis is not that zero should be abandoned as a moral ambition. The thesis is that zero becomes dangerous when leaders mistake silence for control and reward people for protecting the line instead of revealing the truth.

That is why the executive review should compare the dashboard with first aid, clinic visits, anonymous concerns, and precursor events. A green line that sits beside a quiet reporting channel deserves suspicion, especially in maintenance, driving, lifting, working at height, and other high-energy work. Silence can mean maturity, but it can also mean fear.

The companion article on precursor indicators board members still overlook helps here, because weak signals are often the first thing to disappear when teams want a perfect score. When that happens, the dashboard has not improved. It has only become less honest.

Question 3: who owns the decision after the slide closes?

A number without an owner is a decoration. It can be discussed, admired, and filed, but it does not govern anything. The next executive question is therefore blunt. Who owns the decision after the slide closes? Not who owns the report, but who owns the action that should follow if the metric moves in the wrong direction.

This is where metric ownership becomes non-negotiable. A good dashboard needs a data owner, a process owner, and a decision owner. Those roles may sit in different people, but they cannot be vague. If nobody can say who acts, then the metric will drift into theater even when the data is accurate.

Patrick Hudson's maturity model is useful as a comparison because mature systems do not just count events. They convert signals into decisions and decisions into field change. That is the difference between a calculative routine and a genuinely managed one. One produces slides. The other produces control.

TRIR, LTIFR, DART, and SIF rate also belong in this discussion, because each one should lead to a different decision. If the same executive answer fits every number, then the dashboard is too broad to be useful. One metric cannot be both a warning and a verdict.

Question 4: what proof shows the control still exists?

Control cannot be assumed. It has to be shown. That is why the fourth question asks for proof that the control still exists, still works, and still reaches the task at the moment work begins. A paper sign-off is not proof. A clean slide is not proof. Field evidence is proof.

ISO 45001 requires monitoring and review, which means the system has to test whether the planned controls are present in practice. The useful executive habit is to ask for the exact evidence that would let a skeptic agree. If the team cannot produce that evidence, then the control is being described rather than verified.

This is the point where a Bow-Tie, a control register, and a field walk should meet. The model says the barrier exists. The register says someone owns it. The walk says it still functions. When those three views disagree, the executive should trust the field first, because the field is where exposure either exists or does not.

Andreza Araujo's experience across multinational operations is relevant here because she has seen how easily a site can confuse compliance with control. A Ilusão da Conformidade captures the trap well. The form gets signed, the review gets finished, and the barrier is still weak. That is the moment when the dashboard should become more demanding, not more reassuring.

Question 5: which tail event the average still hides?

Averages are comfortable, but they are not enough for serious risk. The fifth question asks which tail event the average still hides. A site can improve the center of the distribution while one crew, one task, one shift, or one contractor group keeps carrying the exposure that can produce a serious injury or fatality.

Frank Bird and Heinrich are still relevant because precursor events matter more than a polished mean. The executive team should ask which high-energy tasks, line breaks, vehicle interactions, energized work, or working-at-height scenarios still sit outside the average. If the tail remains untouched, the dashboard is only telling part of the truth.

This is also why TRIR versus LTIFR versus DART versus SIF rate deserves a direct link in the review. A low injury rate can coexist with serious exposure, because TRIR does not measure the same thing as SIF risk. Leaders who keep those layers separate make better decisions than leaders who let one clean line stand in for everything.

The practical test is to ask which one percent of tasks would worry the most experienced supervisor even if the monthly average looked fine. That question is better than asking whether the chart stayed flat, because it forces the room to name the tail, not hide it.

What the review should do next

The review should end with a decision chain, not a discussion. For each top indicator, the team should name the owner, the field proof, the threshold, the escalation path, and the date of the next verification. If any of those parts is missing, the metric is not ready for executive use.

A dashboard that rewards motion instead of control can feel disciplined for months, then fail suddenly when a weak signal finally becomes visible. The safer move is to test ownership, decision rules, and field proof now, before the next clean slide hides the next ugly surprise.

That is the right moment for a tighter support system. If your leadership team wants help turning the monthly review into a real control test, Andreza Araujo's Safety School and ACS Global Ventures can support the redesign. If the team needs the books first, start with Andreza Araujo's books and tools.

For broader strategic work, Andreza Araujo's site remains the fastest path to the executive conversation, and the article on safety dashboards and blind spots can be used as the next internal read before the next board packet is built.

FAQ

What is metric theater in a safety dashboard?

Metric theater happens when the dashboard makes the meeting look disciplined while the field still runs on habit, silence, or weak controls. The chart looks active, but the worksite has not changed enough to justify confidence.

Why are executive questions better than more indicators?

More indicators do not help if the team still does not know what each one should change. Executive questions force ownership, field proof, and decision rules, which is what turns measurement into governance.

How does Andreza Araujo's work fit this topic?

In Safety Culture: From Theory to Practice and A Ilusão da Conformidade, Andreza Araujo shows that repeated decisions matter more than polished reports. That view helps executives test whether the dashboard reflects real control.

Should a company stop using TRIR?

No. TRIR still has value as a lagging indicator, but it should sit beside SIF exposure, precursor signals, and control verification. A board that sees only TRIR can miss the more serious risk layer.

What should leaders change first?

Leaders should name the owner, write the decision rule, and sample field proof for the top indicators already in use. If a metric cannot pass that test, it should be redesigned before the next review.

Andreza Araujo's best work starts when leaders stop asking the dashboard to reassure them and start asking it to prove something. That shift is small on paper and large in the field.

To keep building on this angle, move next to metric ownership and safety indicators explained, then bring the questions back to the monthly review. If the field cannot answer them, the dashboard is still too friendly.