Across Europe, insurance leaders raise similar questions about connected insurance.
This series addresses them directly.
If you disagree or have questions of your own, email us in English at hello@cmt.ai, en Français bonjour@cmt.ai, in italiano ciao@cmt.ai, auf Deutsch hallo@cmt.ai.

Q2: Can I just use mileage for pricing risk?
I’ve heard it’s more predictive than behaviour.

For decades, motor insurers have priced risk through familiar lenses: mileage, driving experience, and claims history.

These variables have long accounted for much of the observed variation in losses, offering measures that are comparable, stable, and statistically robust. The challenge is that these variables are one dimensional; they lack context, only capturing the outlines of risk on the road.

Claims history records outcomes, not the decisions behind them. Driving experience signals age, but also luck. Mileage reflects how often a driver is exposed to risk, a practical proxy for actuarial exposure. These factors endure because they are easily accessible. But what they describe is what happened, not why it happened.

Risk is created, moment by moment, by what we do behind the wheel.

Speed offers a useful example. Higher speeds correlate with greater crash severity. Yet speed itself reveals little about how risk develops. The same velocity can reflect calm and control in one context and reactive pressure in another. Context matters.

Mileage tells a similar story. More distance means greater exposure, but not necessarily greater danger. 10,000 steady motorway kilometres bear little resemblance to 10,000 kilometers of congested urban driving. Again, context makes all the difference.

Driving behaviour has always been central to risk. It shows how often each driver exhibits patterns that lead to crashes. Until recently, however, these insights were out of reach; the technology to measure them didn’t exist.

Telematics data transforms this dynamic. By making driving behaviour visible and understandable, individually and at scale, it provides the missing context that traditional variables lack. Speeding risk gains meaning with hard brakes and distraction. When we add behaviours to mileage, we learn the true risk of that exposure. From distraction to acceleration to smoothness, we can now observe and quantify driving patterns to understand how risk develops. 

Actuaries have demonstrated that a telematics score including driver behaviour is at least three times more predictive than any rating variable previously employed. This is when comparing the difference in loss costs between the riskiest decile of insured vehicles and the safest decile. Source: NAIC

As we can see, telematics data complements traditional rating factors and exposure like mileage, aligning insurance more closely with how risk actually develops behind the wheel.

If you disagree or have questions of your own, email us in English at hello@cmt.ai, en français bonjour@cmt.ai, in italiano ciao@cmt.ai, auf Deutsch hallo@cmt.ai.