Your HSE Dashboard Is Green. But Are You Actually Safe?

There is something reassuring about a green dashboard.

Training compliance is 98%. Inspections are complete. Audits are on schedule. Actions are being closed. Injury rates remain below target. Leadership visits have been completed.

Everything appears to be under control.

Then something serious happens.

After the incident, the organisation discovers information that somehow never made it onto the dashboard.

A critical piece of equipment had been unreliable for months. Experienced personnel had left. Supervisors were stretched. Maintenance backlog was increasing. Contractors were adapting procedures to get the work done. Several minor events had occurred with similar characteristics.

None of these signals necessarily appeared red.

Some were not being measured at all.

This exposes one of the fundamental weaknesses in HSE performance measurement.

Organisations can become extremely good at measuring safety activity without becoming equally good at measuring whether risk is actually controlled.

A dashboard can therefore be completely accurate and still give leadership the wrong picture.

The problem is not the dashboard.

It is what we choose to put on it.

We Measure What Is Easy to Count

HSE has no shortage of data.

Incidents. Injuries. Training hours. Inspections. Audits. Observations. Toolbox talks. Leadership visits. Corrective actions. Permit compliance. Environmental events.

Modern digital systems allow organisations to collect enormous quantities of this information.

But quantity does not automatically create insight.

Many indicators survive because they are easy to measure.

Training completion is easy to calculate.

The number of inspections completed is easy to count.

The percentage of actions closed is easy to display.

Whether a workforce genuinely understands a critical risk is much harder to measure.

Whether supervision is becoming overloaded is harder still.

Whether a safety-critical control will function when required may require technical verification, field observation and professional judgement.

The danger is obvious.

What is easy to measure gradually becomes what management pays attention to.

And what management pays attention to gradually becomes what the organisation tries to improve.

That can create impressive numbers without necessarily creating better control of risk.

Lagging Indicators Tell Us Something Important

Lagging indicators have received considerable criticism in modern safety management.

Sometimes unfairly.

Injury rates, environmental events, spills, process-safety events and other outcomes remain important.

If people are being injured, leadership needs to know.

If containment is being lost, leadership needs to know.

If vehicle collisions are increasing, leadership needs to know.

The problem begins when absence of negative outcomes is interpreted as evidence that the system is healthy.

It may be.

Or the organisation may simply have avoided the circumstances required to expose its weaknesses.

This is particularly important with low-frequency, high-consequence events.

A company could operate for years without a fatality while carrying significant weaknesses in the controls intended to prevent one.

Historical success does not guarantee future control.

Lagging indicators tell us what happened.

They are much less effective at telling us what nearly happened, what is deteriorating or what may happen next.

Low Injury Rates Can Create Dangerous Confidence

Imagine two facilities.

Facility A reports several minor hand injuries during the year.

Facility B reports none.

Which is safer?

There is not enough information to answer.

Facility A may have excellent major-hazard controls but a recurring issue with minor manual tasks.

Facility B may have no minor injuries while operating with significant weaknesses in isolation, lifting, process containment or vehicle-pedestrian separation.

The injury rate alone cannot distinguish between them.

This is why organisations should be careful when using total recordable injury rates or lost-time injury rates as broad measures of HSE excellence.

These metrics tell us something about injury experience.

They do not provide a comprehensive measure of risk-control effectiveness.

A company can therefore achieve an excellent injury rate while remaining vulnerable to a catastrophic event.

The reverse can also be true.

A minor injury does not automatically mean the organisation's most significant risks are poorly controlled.

Different indicators tell us different things.

Good measurement starts by recognising that distinction.

Leading Indicators Are Not Automatically Better

The response to criticism of lagging indicators has been a growing emphasis on leading indicators.

That is sensible.

But there is another trap.

Not everything that happens before an incident is a useful leading indicator.

Consider:

• Number of inspections completed.

• Number of toolbox talks delivered.

• Number of safety observations submitted.

• Number of leadership visits conducted.

These are technically proactive activities.

But completing more of them does not automatically mean risk is better controlled.

One hundred poor inspections are not necessarily better than twenty excellent ones.

A thousand safety observations about housekeeping may tell leadership almost nothing about the condition of safety-critical barriers.

A toolbox talk completed by every worker may satisfy the target while creating almost no learning.

The distinction is therefore not simply between leading and lagging.

It is between activity and effectiveness.

Stop Measuring How Much HSE You Do

This is one of the most useful challenges an organisation can apply to its dashboard.

For every indicator, ask:

Are we measuring an activity, or are we measuring whether that activity works?

Training completed is an activity.

Competence demonstrated is an outcome.

Inspections completed are activities.

Critical controls verified as effective provide stronger information.

Actions closed are activities.

Recurrence prevented is an outcome.

Procedures reviewed are activities.

Procedures understood and usable in the field provide stronger information.

This does not mean activity measures should disappear.

They can be useful for understanding implementation.

But leadership should not confuse completion with effectiveness.

That distinction sounds obvious.

In practice, dashboards blur it constantly.

Measure What Prevents the Bad Day

A better approach begins with the organisation's significant risks.

What are the events we absolutely do not want?

• A fatal fall.

• Loss of containment.

• Vehicle-pedestrian collision.

• Dropped object.

• Electrical contact.

• Confined-space fatality.

• Major environmental release.

Now work backwards.

What prevents those events?

Those controls deserve measurement.

For work at height, it might include the condition of fall-prevention systems, scaffold integrity, edge protection and competence.

For lifting, it may include equipment integrity, lift planning, exclusion zones and competent personnel.

For process safety, it may include containment integrity, alarm availability, shutdown systems, inspection status and operating-envelope compliance.

This creates a fundamentally different dashboard.

Instead of asking:

“How many inspections did we complete?”

leadership begins asking:

“Are the controls preventing our most serious events healthy?”

That is a far more useful question.

Critical Control Verification Changes the Conversation

Critical control management is particularly valuable because it connects risk assessment with operational assurance.

A bow-tie or similar risk assessment may identify dozens of controls.

Not all controls are equally important.

Some are critical because their absence or failure would materially increase the likelihood or consequence of a major event.

Those controls deserve particular attention.

• Are they present?

• Are they functional?

• Are they being used?

• Are they capable of performing as intended?

This moves verification beyond paperwork.

A procedure may specify that an exclusion zone is required.

Verification asks whether an effective exclusion zone actually exists around the work.

A standard may require gas testing.

Verification asks whether appropriate testing is occurring with functioning equipment and competent personnel.

This provides leadership with information about the condition of risk controls rather than merely the existence of requirements.

Trends Matter More Than Snapshots

Dashboards often reduce performance to current status.

• Green.

• Amber.

• Red.

  
  

This is useful for rapid communication.

But it can conceal deterioration.

Imagine maintenance backlog remains within the organisation's approved threshold.

Green.

Over the previous twelve months, however, the backlog has increased every month.

Still green.

Which information matters more?

Both.

But the trend may tell leadership something the status does not.

The same applies to workforce turnover, overtime, overdue actions, equipment defects, audit findings and contractor performance.

A metric can remain within tolerance while moving consistently in the wrong direction.

Good HSE reporting therefore needs to show direction of travel.

Not simply:

Where are we?

But:

Where are we going?

Repeated Findings Are a Powerful Indicator

Suppose an audit identifies inadequate control of lifting accessories.

An action is created.

The action is closed.

Three months later, another inspection identifies a similar problem elsewhere.

Another action is created.

Another closure.

The dashboard may show excellent action closure performance.

The organisation may actually be demonstrating failure to learn.

Repeated findings are valuable because they tell us that the problem is broader than the individual action.

Perhaps the standard is unclear.

Perhaps procurement is inconsistent.

Perhaps supervision is weak.

Perhaps the corrective action addressed the symptom rather than the system.

A dashboard focused only on closure can miss this entirely.

Leaders should therefore ask not only how many findings are open.

They should ask:

What keeps coming back?

Recurrence is one of the strongest signals that corrective action is not producing organisational learning.

Overdue Actions Need Context

Overdue actions are another popular metric.

Red is bad.

Green is good.

Again, reality is more complicated.

Ten overdue low-risk administrative actions may be less important than one overdue action addressing a degraded critical control.

Treating them equally can distort management attention.

Action management should therefore be risk-based.

What exposure exists while the action remains open?

Are temporary controls adequate?

How long has the organisation accepted the condition?

Who has formally accepted the residual risk?

This turns action management from an administrative process into risk management.

The goal is not simply to make the red number smaller.

It is to understand what unresolved exposure remains.

Listen to Weak Signals

Some of the most valuable safety information does not arrive through formal HSE reporting.

A supervisor says the team is struggling to keep up.

Maintenance mentions that a particular component is becoming difficult to source.

Employees complain that a new system is cumbersome.

A contractor says mobilisation has been rushed.

An experienced employee resigns.

None of these events is necessarily an HSE incident.

But each can alter the conditions under which risk is controlled.

These are weak signals.

Individually, they may mean little.

Together, they can indicate deterioration.

Strong organisations develop ways of listening for them.

This requires qualitative information as well as quantitative data.

Not everything important can be reduced to a percentage.

Sometimes leadership needs narrative.

What are people concerned about?

Where is work becoming harder?

What are we compensating for?

Where are experienced people uncomfortable?

These questions can reveal information no dashboard algorithm will find automatically.

Worker Perception Is Data

Employee surveys are sometimes treated as soft information.

They should not be dismissed so easily.

If large numbers of employees believe schedule pressure is affecting safe work, that is data.

If contractors do not believe they can stop work without consequence, that is data.

If supervisors report that administrative requirements are reducing their field presence, that is data.

Perception does not automatically prove that a problem exists exactly as described.

But perception influences behaviour.

That makes it relevant.

The strongest organisations combine workforce perception with other evidence.

If employees report workload pressure and overtime is increasing, investigate.

If workers say maintenance is deteriorating and equipment defects are rising, investigate.

Different information sources begin validating one another.

That is much stronger than relying on a single KPI.

Beware of KPI Gaming

People respond to measurement.

This is not necessarily dishonest.

It is human.

Tell a manager that their performance depends on closing 95% of HSE actions within 30 days and they will become very interested in closing actions within 30 days.

That may improve performance.

It may also create unintended behaviour.

Actions may be downgraded.

Completion evidence may become superficial.

Complex actions may be broken into smaller items.

Deadlines may be extended.

None of these behaviours necessarily involves deliberate manipulation.

The measurement system simply shapes priorities.

This is often summarised through Goodhart's Law: when a measure becomes a target, it can stop functioning well as a measure.

HSE should be particularly cautious because many safety indicators are proxies.

We are often measuring something because we believe it represents something else.

Inspections represent attention to workplace conditions.

Training represents competence development.

Observations represent workforce engagement.

Once the proxy becomes the target, the original purpose can disappear.

Compare Indicators That Should Tell the Same Story

One way to strengthen HSE analysis is to compare related indicators.

Training completion is high, but competency failures are increasing.

Why?

Audit findings are declining, but incident reports are increasing.

Why?

Safety observations are rising, but almost none relate to significant risk.

Why?

Action closure is excellent, but repeat findings remain common.

Why?

Injury performance is strong, but critical control failures are increasing.

Why?

Contradictions are valuable.

They tell us where the simple explanation may be wrong.

A sophisticated dashboard should therefore not merely display indicators independently.

It should help leadership identify where the indicators disagree.

That is often where the interesting questions begin.

Different Audiences Need Different Dashboards

A frontline supervisor and a board member should not receive the same HSE dashboard.

They make different decisions.

The supervisor needs operational information.

Which controls require attention today?

Which actions are outstanding?

Where are permits conflicting?

What equipment is unavailable?

The executive needs strategic information.

Where is material risk changing?

Which critical controls are deteriorating?

Where are systemic weaknesses emerging?

What decisions require leadership intervention?

Trying to serve every audience with one dashboard usually produces too much information for everyone.

Good reporting starts with the decision-maker.

What decisions are they expected to make?

Then provide the information needed to make them.

A Green Dashboard Should Still Generate Questions

Perhaps the greatest danger of performance dashboards is psychological.

Green feels safe.

Once the page is green, curiosity declines.

That is precisely when leaders should remain curious.

What are we not measuring?

Where could the data be misleading?

Which risks are poorly represented?

What has changed since last month?

Where are people compensating for weak systems?

Which indicator concerns our technical experts even though it remains within target?

The purpose of a dashboard should not be to reassure leadership.

It should help leadership understand.

Sometimes the correct message will be reassuring.

Sometimes it should be uncomfortable.

A good reporting system must be capable of both.

What Should an Executive HSE Dashboard Contain?

There is no universal answer.

The dashboard should reflect the organisation's risk profile.

But a strong executive view will usually combine several perspectives.

It should show serious events and meaningful lagging outcomes.

It should show the condition of controls associated with material risks.

It should identify significant trends.

It should highlight recurring issues.

It should show important organisational factors such as competence, contractor performance, maintenance or workload where relevant.

It should identify unresolved high-risk actions.

It should provide enough narrative to explain what the numbers cannot.

Most importantly, it should identify where executive attention or decision-making is required.

If the dashboard merely describes what happened last month, it is incomplete.

It should help leadership decide what to do next month.

Measure Less. Understand More.

Organisations sometimes respond to uncertainty by adding more indicators.

The dashboard becomes larger.

More charts.

More percentages.

More traffic lights.

Eventually, the volume of information becomes part of the problem.

Leaders cannot give equal attention to fifty KPIs.

Nor should they.

Measurement should create focus.

A smaller number of well-designed indicators linked directly to significant risks can provide more value than hundreds of activity measures.

This requires discipline.

Every metric should earn its place.

What question does this indicator answer?

What decision could it influence?

What risk does it help us understand?

What would we do differently if it changed?

If nobody can answer those questions, consider removing it.

The Dashboard Is a Window, Not the Reality

No HSE dashboard can completely describe an organisation.

It is a simplified representation of a complex operating system.

That is not a weakness.

It is what dashboards are supposed to do.

The danger comes when leaders begin confusing the representation with reality.

Green does not mean safe.

Red does not automatically mean unsafe.

A KPI does not become meaningful simply because it can be calculated to two decimal places.

Good HSE performance measurement combines data with judgement, field intelligence, assurance and an understanding of risk.

So keep the dashboards.

Use the technology.

Track the trends.

Measure performance.

But occasionally look beyond the screen and ask a harder question:

What would we need to know today to prevent our worst day tomorrow?

Then ask whether the dashboard can answer it.

If it cannot, perhaps the problem is not your safety performance.

Perhaps you are simply measuring the wrong things.

References and Further Reading

International Association of Oil & Gas Producers (IOGP). Process Safety – Recommended Practice on Key Performance Indicators. IOGP guidance supports the use of leading and lagging indicators to provide insight into process-safety performance and the health of systems intended to prevent major events.

International Association of Oil & Gas Producers (IOGP). Operating Management System Framework. Industry framework addressing risk management, performance monitoring, assurance and continual improvement within oil and gas operations.

UK Health and Safety Executive (HSE). Developing Process Safety Indicators – HSG254. Guidance explaining how organisations can develop leading and lagging indicators linked to risk-control systems rather than relying solely on outcome statistics.

UK Health and Safety Executive (HSE). Managing for Health and Safety – HSG65. Guidance covering monitoring, measuring and reviewing health and safety performance within the Plan, Do, Check, Act management framework.

International Organization for Standardization (ISO). ISO 45001:2018 – Occupational Health and Safety Management Systems. Requirements addressing monitoring, measurement, analysis, performance evaluation, audit, management review and continual improvement.

Center for Chemical Process Safety (CCPS). Process Safety Leading and Lagging Metrics. Guidance supporting the development and use of indicators intended to measure process-safety performance and identify weaknesses before major events occur.