In this paper we present a quality model highlighting the completeness of sensor data with respect to its application. The model allows consistent handling of information loss as data propagates through a sensor network. The tradeoffs between various factors that influence completeness are quantified thereby allowing an integrated view of completeness at various levels in a system. The paper is presented in the context of the fast emerging field of smart spaces. All concepts in the paper have a foundation in real-life problems arising in this context. Preliminary implementation results are presented to illustrate the value of the completeness based approach versus one that does not use completeness.