A method and apparatus (400) for evaluating the performance of a user service executed in a communication network by means of one or more user-independent system services. A measuring block (400a) measures a set of system service key performance indicators S-KPIs for each system service, the S-KPIs being carefully selected to reflect the performance of the system services from a user perspective. Relevant data for measuring the S-KPIs is collected from sensors and counters, including infrastructure data (402c), traffic data (402b) and end user data (402a). A service evaluating block (400b) then evaluates or estimates the performance of the user service based on the measured S-KPIs. By making a distinction between system services and user services in this way and using measurable S-KPIs, more consistent and reliable service evaluations can be obtained.