Over the last years, the Internet of Things has fostered a growing interest in context-aware mobile applications; this fact is mainly due to highly favoring information provision from multiple Internetconnected devices. To identify user context, these applications collect information from the user and his/her environment and typically filter app information, so that the user receives only the interesting and relevant information. However, such a task usually implies further resource consumption on user mobile devices, not only regarding battery usage but also in terms of network traffic. Accordingly, although context-aware applications can improve user experiences in their daily lives, they must ensure the maintenance of lowlevel resource consumption; otherwise, the applications are promptly replaced by less consuming ones, and therefore, removed from the mobile market. In this paper, we evaluate and discuss several architectural styles for context-aware mobile applications, as well as, providing a set of guidelines to decide on the right architecture for a particular app depending on its characteristics. The use of such guidelines when choosing the right architectural style can strongly influence the resource consumption of context-aware mobile applications. Following these guidelines, user satisfaction of a context-aware mobile application may be improved, thus guaranteeing the app success.