A recent analysis from Arity across multiple insurance carriers shows that the Zone Improvement Plan (ZIP) codes with the highest hard braking frequency showed 3.5 times higher bodily injury loss costs and 2.8 times higher personal injury protection loss costs, compared to ZIP codes with the lowest hard braking frequency. This highlights how behavioural signals translate directly into loss costs.
Henry Kowal, the Insurance Product Director at Arity, says the answer lies in better understanding real-world driving behaviour, as he tells Digital Journal: “Real-world driving behaviour data provides an actionable signal of territorial exposure. Our analysis shows that where hard-braking frequency is higher, injury-related loss costs rise sharply. That gives insurers a more current, behaviour-based view of territorial risk than loss history alone so that they can refine pricing with greater confidence”.
Digital Journal: Traditional actuarial models have long relied on historical claims and loss data to assess risk. What are the limitations of this backward-looking approach in today’s rapidly evolving driving environment?
Henry Kowal: Relying on historical claims data creates an unavoidable time lag, meaning changes in driving behaviour can take 12–36 months to appear in loss experience and pricing decisions. Claims are relatively infrequent events, and when analysed at the ZIP-code level, the resulting data is often sparse or lacks statistical credibility.
Loss experience typically requires multiple years to mature, which slows insurers’ ability to identify and respond to emerging territorial risk trends. Many commonly used third-party proxies, such as surveys, census data, or industry loss benchmarks, are updated infrequently and reflect conditions that may already be outdated. Rapid shifts in driving behaviour (such as post-COVID traffic patterns or return-to-office mandates) can materially change risk well before those shifts are visible in claims or loss ratios.
DJ: How can mobility and behavioural driving data help insurers detect emerging risk trends before those trends begin to show up in claims or loss ratios?
Kowal: Behavioural driving data reflects what is happening on the road today, rather than relying on outcomes from years in the past, making it a leading indicator of risk. Because driving behaviour is observed continuously and refreshed regularly, it captures risk shifts as they develop instead of after losses accumulate.
ZIP-level behavioural insights allow insurers to identify localized changes in risk before they aggregate into state-level loss trends. This enables actuarial and pricing teams to act proactively, rather than waiting for claims experience to justify rate or territorial changes.
DJ: What types of behavioural shifts are you seeing emerge in driving patterns today, and why should pricing and actuarial teams be paying close attention to them?
Kowal: Driving behaviour is becoming increasingly variable across ZIP codes, with risk diverging meaningfully even within the same state or rating territory. External forces, including return-to-office mandates, economic activity, and changes in roadway usage, are actively reshaping how, when, and where people drive.
These shifts reinforce the limitations of static or backward-looking inputs and highlight the need for current, behaviour-based data in pricing and actuarial analysis.
DJ: How can analysing driving behaviour at a ZIP or geographic level help insurers better understand localized risk and refine territorial pricing models?
Kowal: Behavioural insights help improve risk segmentation in areas where internal loss data is sparse or lacks credibility. Insurers can identify misalignment between historical territorial assumptions and current, real‑world driving patterns. ZIP codes serve as a practical and familiar linking key, making behavioural insights easier to integrate into existing workflows and datasets.
DJ: For pricing and actuarial teams looking to incorporate behavioural mobility insights into their work, what are 3–5 practical steps they can take to start identifying behavioural risk trends earlier?
Kowal: Pricing and actuarial teams can overlay ZIP‑level driving behaviour data on existing territorial models to test current assumptions. Evaluate whether similarly priced territories actually exhibit similar driving behaviour, and adjust segmentation where misalignment exists.
They can also use behavioural insights to supplement sparse loss experience, especially for regional carriers, and monitor year‑over‑year behaviour changes to detect emerging trends sooner. Incorporating observed behaviour changes into actuarial narratives can help support pricing decisions and regulatory filings.
DJ: As access to behavioural driving data expands, how do you see the role of actuarial and pricing teams evolving over the next five to ten years?
Kowal: Actuarial and pricing teams will increasingly shift from reacting to historical loss outcomes toward anticipating where territorial risk is likely to change next. Actuaries will be expected to identify emerging pockets of risk before those shifts are fully reflected in claims or loss ratios. Behavioural data will increasingly be used for this to strengthen actuarial judgment by adding timely, explainable context to traditional rating factors, rather than replacing them.
