Category: Telematics · Reviewed by Taylor Watts, Broker · New Business · Last reviewed 2026-06-10
Driver scoring is the algorithmic process by which a telematics dataset of a person’s driving — comprising journey-level GPS, accelerometer, speed, time-of-day and, increasingly, phone-handling signals — is reduced to a single numeric or letter index used as a rating, pricing or eligibility factor in motor insurance.
Category: Telematics Aliases: driver score, telematics score, behavioural risk score, PHYD score Established: Embedded in UK PHYD products from c. 2010; methodologies updated continuously Related: Pay-how-you-drive, Telematics insurance / UBI, FCA Consumer Duty telematics, GDPR telematics
A driver score is a derived attribute. The inputs are observed driving events; the outputs are a scalar (typically 0–100, sometimes letter-graded A–E) that summarises risk relative to a reference population. The underlying methodology is normally proprietary; published documentation generally explains what contributes to the score (braking, acceleration, cornering, speed in relation to limit, time of day, journey duration, phone handling) without disclosing weights or interaction terms.
Driver scoring is used inside pay-how-you-drive products, in hybrid PHYD-PAYD policies, in fleet driver-behaviour management, and in some claims investigations.
The FCA’s Consumer Duty in PS22/9 (July 2022) requires firms to act to deliver good outcomes for retail customers. The consumer understanding cross-cutting rule applies in full to driver scoring: the policyholder should be able to grasp how their driving translates into price. The products and services and price and value outcomes constrain how a score can be used to surcharge, cancel or exclude. PROD 4 requires consideration of foreseeable harm at the design stage.
The General Insurance Pricing Practices rules in PS21/5 (May 2021) prevent renewal price walking — relevant because a renewing customer with a long score history must not be priced less favourably than an equivalent new customer.
UK GDPR Article 22 places restrictions on decisions based solely on automated processing, including profiling, which produce legal or similarly significant effects. Renewal refusal, mid-term cancellation and material price changes generally fall within scope. Available bases include contract necessity (Article 22(2)(a)) or explicit consent (Article 22(2)(c)); in both cases the data subject’s safeguards in Article 22(3) — meaningful human intervention, the ability to express a point of view, the right to contest — must be respected. The Information Commissioner’s Office’s published guidance on automated decision-making and profiling, and the 2024 connected vehicles guide, set the supervisory expectations.
The Equality Act 2010 prohibits direct and indirect discrimination on protected characteristics (age, sex, race, disability, religion and others). Driver score inputs that correlate with protected characteristics — for example, night-time-driving weights that may disproportionately affect shift workers from specific demographic groups, or phone-handling sensitivity that may disadvantage disabled drivers using accessibility tools — can give rise to indirect discrimination concerns under section 19 unless a proportionate means of achieving a legitimate aim can be demonstrated. The ICO’s pre-existing concerns about algorithmic opacity reinforce the practical importance of internal testing.
Telematics suppliers ingest raw event data, normalise for vehicle type (e.g. acceleration thresholds adjust for kerb weight), aggregate over a defined exposure window (commonly the trailing 90 days), and apply a weighting model. Suppliers periodically recalibrate against claims experience. Insurers map the score to a rating factor in their pricing model; in the most transparent products the customer-facing portal explains how to improve the score.
Independent academic and industry studies (Geneva Association Promoting Loss Prevention through Digital Risk Engineering and similar) report that the predictive lift of behavioural telematics is real but modest, with much of the predictive power historically attributable to mileage and time-of-day exposure rather than discrete event severity.
Variations include claim-frequency-trained scores, claim-cost-trained scores, supplier-pooled “universal” scores, and OEM-native scores generated within the vehicle. As phone-handling detection matures, products are integrating distraction signals — and so increasing the salience of Article 22 controls. Predictive maintenance and crash-detection layers (e.g. Cambridge Mobile Telematics’ crash detection) sit alongside scoring.
A customer disputes a renewal price increase driven by a fall in score from 84 to 71. The insurer’s complaints team obtains a meaningful-human-intervention review under Article 22(3): an analyst reviews journey data, identifies that a new commute through a 20 mph zone has triggered repeated marginal speed flags, and applies a manual adjustment, with the customer being offered the option of an updated score after a further 90-day window.
This entry is part of the Apex Insurance Wiki. Last reviewed by Matt Bartlett on 2026-06-10. Next review: 2026-12-10.
Apex Insurance Brokers Limited. Authorised and regulated by the Financial Conduct Authority, FRN 724952. Registered in England and Wales, Companies House 07014570. This entry provides general information about UK insurance concepts and is not regulated advice. Consult your insurance broker on your specific position.
Apex Insurance Brokers serves UK professional services firms and commercial businesses. Call 0117 325 0027, email hello@apexinsurancebrokers.co.uk, or request a quotation.
Get a quote