In vivo exposure therapy (IVET) is the treatment of choice for specific phobias, including animal phobia. Recently virtual reality exposure therapy (VRET) and Augmented reality exposure therapy (ARET) have been presented and validated as suitable tools, however, to the best of our knowledge, no other work has identified nor provided an explicit classification or taxonomy to classify, compare and even inform the design of these kind of systems. In this work we propose a taxonomy of feedback factors that affect the experience of use of mixed reality systems for small-animal phobias. The taxonomy comprises feedback factors in three categories, namely realism, interaction, and intensity; while considering: auditory, haptic and visual dimensions for each. To illustrate this taxonomy in use, we conducted a survey to compare mixed reality applications for small-animal phobias, analysing their features according to the factors proposed in the taxonomy. In addition, we developed the concept of smart objects of fear by means of a design fiction world. In this world, in addition to IVET, VRET and ARET, where the objects of fear are either real (in the former) or virtual (in the two later), we envision a new kind of mixed reality therapy, where the objects of fear are represented using context-aware smart objects, capable of inferring the emotional state of the patient, and of adapting their behavior to improve the system’s response during the therapy; which in turn, may enrich the patient’s therapy experience. We validated the perceived usability of this design fiction world with an expert in exposure therapy for small-animal phobias, whom perceived it as useful and correct. Finally, we show how the use of the taxonomy allowed us to discuss the features of the proposed future technologies in terms of how they furnish each of the factors of the taxonomy.