The Internet of Things (IoT) is a very dynamic and heterogeneous environment that generates plethora of sensor data, accessible to develop new smart pervasive applications. However, the substantial amount of effort required to collect and disseminate context data with sufficient quality prevents the context consumers to take advantage of the IoT to its full potential. Consequently, novel research efforts are required to design middleware solutions able to deliver relevant context data to consumer applications while hiding the complexity of data distribution in heterogeneous and large-scale environments. This paper presents the INCOME framework that enables context producers to express the level of Quality of Context (QoC) they are able to provide and context consumers to set thresholds on the QoC they expect in order to determine how to distribute context data. Our experiments show that context data can be annotated with QoC metadata and distributed from producers to consumers with a reasonable additional cost even on resource-constrained devices such as Raspberry Pi.