Clustering analysis is a crucial technique in data processing and pattern understanding. In this study, we compare the clustering results using the k-Means algorithm with two different approaches to centroid initialization: random centroids and manual centroids. The dataset consists of three observed variables. The analysis results indicate significant differences in centroid placement and cluster formation between the two approaches. The random centroid approach yields three clusters with centroids located at different coordinates: Cluster 1 [1.76, 2.5, 10.88], Cluster 2 [1.60, 1.87, 2.23], and Cluster 3 [1.64, 1.568, 15.88]. On the other hand, the manual centroid approach generates three clusters with centroids manually specified: Cluster 1 [1.64, 1.81, 14.84], Cluster 2 [1.61, 1.901, 2.04], and Cluster 3 [1.75, 1.7, 6.8]. The analysis and interpretation of these differences highlight the sensitivity of the k-Means algorithm to centroid initialization. The implications of these findings provide insights into the importance of selecting the appropriate initialization method in clustering analysis to ensure consistent and meaningful results. This research makes a significant contribution to understanding the factors influencing clustering results and can serve as a guide for researchers and practitioners in choosing clustering approaches that are suitable for their data and analytical goals.