This paper proposes a novel method named exploiting spatio-temporal attributes of videos, which focuses on reducing bandwidth consumption in high-definition video streaming services. The method sequentially embeds every ninth frame of a video sequence in the spatial domain of its preceding eight consecutive frames. In this way, every ninth video frame does not require transmission, hence saving substantial amount of bandwidth. To avoid any visible degradation in videos, the embedding is done in an intelligent fashion by considering the variations in temporal information between consecutive video frames. The results show that up to 19% of bandwidth can be saved, while maintaining high video quality.