Compressive Sensing (CS) has proved to be a promising approach for the Internet of things (IoT) due to the fact that CS can abate the magnitude of raw data which is to be transmitted to the sink. It further helps in acquiring the traffic load balancing in the whole network. Recently, a plethora of research is reported that combines the clustering with CS in three genres; a plain CS, hybrid CS and a multi-path hybrid CS. However, the number transmissions are too high by the nodes (plain CS) or by the Cluster Heads (CHs) (hybrid or multi-path hybrid). While adopting the aforementioned genres of CS-based clustering, the selection of CH has not been given significant attention. This results in inevitable reduction the network lifetime of IoT-based WSN. Therefore, to extenuate the aforementioned concerns, in this paper, two contributions are reported. Firstly, the CH selection is done by the energy, distance, node density and average energy of the network that helps in the befitting CH selection of a node. Consequently, data gathering is improved and compression is done at the CH level. Secondly, the data reconstruction is also made better as compared to the state-of-the-art protocols helping in enhancing the Signal to Noise Ratio. The proposed scheme is named as Energy efficient CS based clustering framework (ECSCF). It is evident from the simulation that the ECSCF outperforms the competitive CS-based algorithms on the platform of different metrics namely, network lifetime, stability period, energy consumption, network’s remaining energy, etc.