The widespread popularity of Internet of things (IoT) enables a huge amount of fine-grained thermal comfort and, implicitly, energy efficiency data to be assembled. Meanwhile, the movement toward a human-centric approach in the domain of smart homes has constantly been advancing worldwide. How to collect, prepare, store and employ the huge amount of data collected by the IoT devices to promote and enhance the thermal comfort inside buildings while preserving energy constraints is an important issue. To date, different IoT applications have been introduced in the thermal comfort domain. To provide a thorough overview of the prevailing research and to identify challenges for the upcoming research, this paper conducts a review on the IoT thermal comfort applications and thermal comfort analytics in buildings. Following the three stages of analytics: descriptive, predictive and prescriptive, we propose the basic application areas as thermal comfort analysis, thermal comfort prediction and thermal environment control. We also review the novel techniques endorsed by each application. In addition, we discuss some research trends associated with the topic like machine learning techniques, data privacy and security and federated learning.