Rethinking Risk Through Events That Never Result in Claims
by Adriana Zambon
For a long time—and to a large extent still today—the insurance industry has built its understanding of risk by observing events that result in a loss: accidents, claims, damages, and compensation requests.
But what if we tried to look at risk from a different perspective?
A claim represents the moment when risk materializes and becomes measurable: accident frequency, loss severity, geographical distribution, and the characteristics of the drivers and vehicles involved.
Today, however, the growing availability of telematics data is opening new perspectives on risk.
If a claim represents the final outcome of a sequence of events, is it enough to observe only the result? Or is there an even broader source of information hidden within all those situations where an accident could have happened but did not?
The question may seem paradoxical, but it raises an important consideration: the future of risk analysis may depend less on understanding what happened and more on understanding what was about to happen.
A Claim Tells Only Part of the Story
A road accident is a relatively rare event.
Every day, millions of people travel billions of kilometres without generating a single claim. Yet during these journeys, situations involving potential exposure to risk occur continuously: a sudden slowdown in traffic, emergency braking, an unexpected pedestrian crossing, an evasive manoeuvre, or the successful navigation of a particularly complex road segment.
These events do not necessarily result in an insurable loss. They do not appear in claims statistics. They do not generate compensation requests. Yet they reveal a great deal about how risk develops and is managed in everyday mobility.
Focusing exclusively on accidents means, in a sense, concentrating only on the final moments of a much more complex sequence of events.
The evolution of telematics has made it possible to observe mobility behaviour with a level of detail that would have been unimaginable just a few years ago. Today, millions of journeys can be analysed, providing insights not only into where and when accidents occur, but also into how risk emerges, evolves, and—in most cases—is managed before it turns into a claim.
The concept of a near miss is not new and has long been used in various fields of safety management and insurance research. What is changing today is the ability to observe these events at scale through telematics data and transform them into a new source of risk intelligence.
In this context, information emerges that has traditionally not been part of insurers’ analytical assets:
- high-risk driving events that do not result in a claim;
- exposure to complex traffic environments;
- interactions with particularly challenging infrastructure;
- behaviors that help avoid potentially dangerous situations;
- recurring patterns that precede accidents.
These signals make it possible to observe risk at a much earlier stage than when it ultimately materializes as a claim.
Perhaps the most interesting aspect is that many risk situations follow recurring patterns. Before an accident occurs, there are almost always several conditions that increase its likelihood: inappropriate speed for the context, high traffic density, distractions, aggressive driving behaviours, or simply a combination of unfavourable factors. The difference is that, in most cases, the chain of events is interrupted before damage occurs. A driver reacts correctly. Another road user changes course. Traffic conditions shift. The risk is neutralized.
Analysing these situations means studying not only system failures, represented by accidents, but also successful outcomes. In other words, it means understanding the factors that enable millions of people every day to avoid becoming involved in a claim.
From Avoided Accidents to Prevention
This perspective opens new opportunities for the insurance industry.
While traditional models are primarily based on the historical observation of claims, the availability of telematics data now makes it possible to integrate a preventive dimension. Understanding where latent risks are concentrated, which behaviours most effectively reduce accident probability, and which contexts generate greater exposure can help insurers develop more effective prevention and risk management strategies.
The objective is not to replace claims analysis, which remains a fundamental component of the insurance business, but to enrich it with a more comprehensive understanding of the dynamics that precede a loss.
In this scenario, the most valuable data may not be only the data describing accidents that have already occurred. It may be the data that tells the story of the thousands of avoided accidents each day.
In this context, OCTO observes millions of journeys and billions of kilometres travelled every day. The growing availability of telematics data creates opportunities to explore new dimensions of risk, including those associated with critical events that do not result in an accident.
Understanding avoided accidents means beginning to interpret risk differently: not merely as an event to be recorded, but as a phenomenon to be understood and, potentially, anticipated. It is precisely within this still largely unexplored dimension that one of the next frontiers of insurance intelligence may lie.