Whether your organisation has a mature safety data practice or is just beginning to build one, the goal is the same: keeping people safe and continuously getting better at it.Â
The reporting systems, the processes, the conversations happening at every level of the business all represent commitment to safety, and that matters. The opportunity we want to explore is how to get even more value out of what you’re already collecting.
The truth is, safety data is one of the most underutilised assets in many organisations. Not through lack of effort, but because the tools and frameworks for making sense of it are still catching up with what we now know about how serious incidents actually occur. When it comes to managing critical risk, that gap between the data you’re collecting and the insight you’re extracting can have very real consequences.
How much safety data are organisations missing?
Research across more than 12,500 participants spanning nine industries shows that around 25% of safety incidents globally go unreported. In Australia, that figure sits closer to 31%, about 6% above the global benchmark. It’s tempting to read that as a failure. We’d encourage a different lens: it’s an opportunity to understand why, and to close that gap in a way that builds trust rather than erodes it.
To put that in practical terms, 50% of people experience a safety incident, 30% fail to report at least one, and the average number of unreported incidents per person is 6.3. For an organisation of 3,000 people, that’s approximately 450 incidents per year that never enter your system and never inform your decisions. Not because people don’t care, but because the conditions for reporting aren’t yet where they need to be.
The drivers of under-reporting tend to cluster around three areas: under-appreciation, where people feel that reporting doesn’t lead to anything changing (around 38% of cases); fear of social or professional consequences, which accounts for roughly 37%; and process friction, where the reporting system itself is simply too cumbersome to use, cited by about 25%. None of these are fixed. Each points toward a specific, actionable response, whether that be better feedback loops, psychologically safe environments, or simpler processes.
Organisations that have addressed these drivers haven’t just improved their data. They’ve seen near-miss reporting increase significantly, which means more learning, more early intervention, and ultimately fewer serious events.
Looking more closely at what the data is telling you
One of the most valuable shifts happening in safety right now is the move toward distinguishing between incidents based on their potential severity, not just their actual outcome. This is where the real insight lives.
Think about two incidents that result in the same injury, a fractured foot. In one case, an employee slips off the bottom step of a truck cab. In another, a worker is struck by a reversing forklift with a faulty backup alarm. The outcome is identical. The potential consequence of the second scenario is categorically different, and it deserves a different level of attention, investigation, and learning.
This isn’t about ranking people’s pain or dismissing lower-severity events. It’s about ensuring that the incidents with the greatest potential to cause serious harm get the focused scrutiny they deserve, while organisations aren’t stretched thin investigating everything with the same intensity.
Analysis across mature organisations in utilities and mining has found that up to 45% of Serious Injury and Fatality potential (SIFp) incidents are being classified at low or moderate risk levels meaning they bypass the scrutiny they warrant. At the same time, around 29% of incidents are attracting investigative effort disproportionate to their actual risk, pulling resources away from where they’re needed most. In one case study, 87% of SIFp-classified incidents had been internally categorised at a low or moderate level, while 47% of incidents drawing significant investigation effort turned out to carry no meaningful SIFp at all.
Sharpening this distinction is foundational to effective critical risk management. It’s also one of the highest-leverage improvements an organisation can make, not just for safety outcomes, but for the efficiency and focus of your entire safety function. You can’t manage critical risk well if your data isn’t telling you where it actually lives.
How one organisation used existing data to reduce serious injury exposure
One large utilities organisation worked with Sentis to do exactly this and the results show what’s possible when you look at existing data through a more refined lens.
By applying more nuanced SIFp classification to their incident data (i.e. near misses, hazards, and recorded events) they were able to identify exactly which risk categories were generating the most serious-injury exposure. Motor vehicle incidents were the most prevalent. Animal attacks, which had largely flown under the radar, turned out to be a more significant exposure than anyone had anticipated.
With that clarity, they were able to strengthen the critical controls that mattered most, redesign specific high-risk tasks to address identified human factors, and meaningfully improve workforce engagement at pre-start and shift handovers. Leaders at every level became more proactively focused on the presence of controls in the field, rather than simply responding to what had already gone wrong. Over two years of targeted interventions, they saw a material reduction in recordable incidents and a significant uplift in near-miss reporting. This is the kind of result that reflects a culture genuinely learning and improving, not just managing metrics.
None of this required starting from scratch. It required looking at what they already had with fresh eyes.
How does safety culture impact data quality and reporting?
Underpinning all of this is safety culture, and it’s worth naming because it’s so often the invisible variable. Research across 26,000 people spanning 86 organisations shows a clear pattern: organisations with lower cultural maturity tend to see safety as something done to them, with procedures perceived as compliance exercises rather than genuine tools. Reporting is inhibited, data quality suffers, and the willingness to really dig into what the data is saying is limited.
High-performing cultures look markedly different. Leaders actively seek out weak signals and respond constructively. Critical risk management isn’t a periodic audit or a board-level talking point; it’s embedded in how leaders inquire, how teams communicate and how work is actually done. Workers feel confident that raising a concern will lead to action, not scrutiny. There’s a shared sense that safety is something everyone contributes to proactively, and near-miss reporting and hazard identification bubble up from the workforce, not just down from management.
The feedback loop matters enormously here. When people see that their reports lead to visible improvements, reporting becomes meaningful. That meaning is what sustains the culture over time and it’s what ultimately makes the data reliable enough to act on.
Five ways to get more from your safety data
If this resonates, a few practical areas worth reflecting on for your organisation:
Create conditions for reporting: address the fear, friction and under-appreciation drivers specifically, not generically.
Discriminate between SIFp and non-SIFp incidents: different severity, different response, different learning. This is the foundation of reliable critical risk management.
Audit your classification integrity: don’t assume your categorisation system is catching what matters.
Move from point-in-time reviews to dynamic monitoring: analyse longitudinal data, then build continuous tracking into how you operate.
Close the feedback loop with your workforce: show people what their reporting led to. That’s what sustains a culture of reporting.
The organisations getting ahead of serious injury and fatality risk aren’t necessarily collecting more data. They’re making the data they already have actually work for them, and in doing so, building the kind of clarity that effective critical risk management depends on.
The question worth sitting with: if your safety data could tell a richer story, what decisions might you make differently? Who in your organisation could it better protect?
Understanding your critical risk exposure starts with knowing what your data is really telling you.
Sentis works with organisations to surface the insights already sitting in their safety data and build the culture that acts on them. If you’re ready to look more closely, we’d love to talk.
