In most countries, to have permission to drive vehicles on public roads one must have insurance against civil
liability for vehicles. In many cases, the insurance fees depend on the age of the driver, the number of years
one holds a driving license, and the driving history. The usual assumption taken by insurance companies that
younger drivers are always more risky than others are not always correct, penalizing young good drivers. In
this paper, we follow a pay-as-you-drive approach based on trip behavior data of different drivers. First, we
build a dataset from real trip data. Then, we apply a two-stage clustering approach to the dataset to identify
trip profiles. The experimental results show that we can cluster and identify distinct trip profiles in which
many trips have a non-aggressive style, some have an aggressive style and only a few are risky style trips. Our
solution finds application in fair insurance fee calculation or fleet management tasks, for instance.