Wireless Sensor Networks (WSNs) are crucial for data collection in Internet of Things (IoT) devices, consisting of numerous small, spatially distributed, battery-operated sensors. These sensors, often placed in hard-to-reach areas, are constrained by limited energy resources, making energy-efficient data transmission vital. In order to address this issue, various clustering algorithms are being used to reduce energy consumption and extend the lifespan of networks. This paper presents a novel energy-efficient clustering protocol for both homogeneous and heterogeneous WSNs. The proposed EECP2-IOT protocol uses a balanced probability threshold to identify optimal cluster head candidates, selecting the final cluster head based on remaining energy and proximity to the base station. The proposed EECP2-IOT was tested on both homogeneous and heterogeneous networks. Simulation results and comparative analysis show that the EECP2-IOT (HETRO) performed better as compared to EECP2-IOT (HOMO) approach and significantly improved energy efficiency and extended network lifespan.