This work presents a new approach designed to reduce the costs (e.g., signalling, latency, power consumption, etc.) of providing location services in cellular networks that use a combination of location techniques in a hybrid fashion: MiLCO (middleware for location-services cost optimization). This proposal is based on managing location requests according to their quality of service (QoS) requirements of accuracy and latency, and selecting the most convenient technique or combination, i.e., the one that optimizes the use of the system's resources for positioning, whilst fulfilling QoS requirements. A possible implementation of this location middleware is evaluated and the results in comparison with other wide-deployed solutions are presented.