Turning behavioural data into measurable gains in motor insurance
Sarah Vaughan, Director, Angelica Solutions
For over a decade, the insurance industry has acknowledged the potential of telematics to deliver fairer pricing, better risk selection, and safer roads. Yet it remains a largely untapped resource. While some insurers have made moves in the right direction, many are still operating at surface level with basic propositions that use telematics data solely to refine risk selection at renewal. The broader potential remains underused, perhaps due to the lack of meaningful evidence linking driver behaviour to claims outcomes. Until now, that evidence has remained largely theoretical.
A persistent industry challenge is how to move behavioural insights from potential to performance. The insurance industry is no stranger to data accumulation and telematics programmes may collect vast quantities of behavioural data but without the infrastructure or commitment to act on it, the potential gains are lost. In our experience, even well constructed scoring models can fall short of delivering value if they aren’t aligned with business processes and decision making workflows.
In a first of its kind independent study of telematics data, we analysed Howden Driving Data’s (HDD) behavioural scoring system. Analysing over 1.2 billion miles of telematics data from Ingenie, one of the UK’s most established young driver portfolios, which showed that HDD’s Driver Score didn’t just correlate with risk, it could predict it with consistency and materiality.
The difference in claims frequency between the highest and lowest scoring drivers exceeded a factor of three. Driving at night was particularly high risk, with frequent night drivers 17 times more likely to incur a large loss than those who drive exclusively during daylight hours.
This insight enabled our teams to model the commercial impact of early intervention using HDD Driver Scores. Acting on the worst 10% of scores within 30 days of policy inception could cut year one loss ratios by 6%. Applying the same approach at renewal could improve renewal portfolio performance by up to 13%. These results were not theoretical but grounded in live portfolio performance and real world conditions. This insight hit at the heart of the industry challenge, demonstrating how behavioural data could be practically converted into measurable performance gains.
The issue hasn’t been a lack of data, but a lack of integration. Many early telematics propositions focused on gathering information rather than using it meaningfully. Too often, the outputs were siloed. Whilst they were technically insightful, they were operationally disconnected. Real value emerges only when behavioural data influences real time decisions across underwriting, pricing, and claims. That requires more than data science, it takes cross-functional coordination, operational infrastructure, and the willingness to engage with customers in a relevant, timely way. Some providers simply weren’t set up for that, and as a result, the propositions didn’t demonstrate enough value to persist long term.
But when done properly, the potential of telematics is clear. The HDD study, for example, modelled two powerful interventions and demonstrated that just 30 days of driver data provides a powerful predictor of claims frequency for the overall policy term. Policy cancellation for the worst drivers after 30 days led to a 6% loss ratio reduction, while behavioural improvement strategies such as warnings and feedback, if well executed, could deliver another 5% or more in improved year one loss ratios. This proves that insurers can reduce risk, not just price it when they act on the data.
HDD’s platform represents an important evolution in this space. By converting behavioural inputs into predictive, usable intelligence, it addresses a long standing shortfall in telematics propositions. And in doing so, it equips insurers with the means to act and not just analyse, presenting a practical route to improved performance and greater fairness in underwriting.
Turning insight into operational impact requires more than just technology, it demands deep industry knowledge, practical experience, and a clear understanding of what works in the real world. Our work on the HDD study demonstrates how even complex behavioural data can be transformed into measurable commercial outcomes when the right expertise and frameworks are applied. For insurers ready to move from theory to ROI, the opportunity is not just within reach, it’s already here. The challenge now is how swiftly and confidently the industry is prepared to act.
