Accurate prediction and analysis of power consumption trends across multiple zones is essential for effective energy management and sustainability. However, present approaches frequently lack precision and fail to capture the complex dynamics of power consumption. This paper addresses this issue by presenting a comparative evaluation of machine learning (ML) techniques for power consumption analysis across multiple zones. The issue statement focuses on the requirement to discover the best algorithms for predicting and analysing power usage patterns. Several algorithms, including decision tree (DT), random forest (RF), K-Nearest Neighbour (KNN), AdaBoost, logistic regression (LR), naive Bayes (NB), and artificial neural network (ANN), are tested for their effectiveness in this task. The findings are meant to provide insights into selecting the best algorithm for optimising energy management techniques and supporting sustainability activities.