The relevance of this study is driven by the need to understand the most effective clustering method for customer segmentation using machine learning. This is necessary to develop effective measures for implementing these methods in modern organizations while minimizing costs and maximizing efficiency. The purpose of the study is to evaluate four clustering methods, namely K-means, hierarchical, density-based, and affinity clustering, in comparison and determine the most effective one for companies in the process of forming target customer groups. The following methods were used in the study: analysis, comparison, abstract and logical. According to the results of the study, there is no single correct and most effective clustering method. Companies should choose the method according to the initial situation, because each clustering method has its own advantages and disadvantages. K-means with 10 clusters showed the average values of three characteristics. Hierarchical clustering, which does not require an initial number of clusters, identifies them using a dendogram, but has a higher time complexity. Density-Based Spatial Clustering of Applications with Noise is suitable for arbitrary cluster shapes, but formed only three clusters in the experiments due to limitations in data density differences. Affinity Propagation is efficient but time consuming, making it preferable for small to medium sized datasets. The practical significance of this study is that its results are important for companies to identify consumer segments and tailor their branding strategies to the unique needs of each group. Ultimately, this will allow companies to improve their performance and increase their financial results, which will affect their competitiveness in the marketplace.