Many innovations in the retail sector are closely related to digitalization and technological advances. The value of sales data is increasing as a resource for businesses in today's digital era. Sales data consists of details about the items sold, the clients they serve, sales patterns, and other elements that influence how well a business operates. However, effectively handling and understanding vast and complex sales data can be difficult and confusing. So with this problem the author will determine the annual sales level of the products that are most widely marketed by applying Exploratory Data Analysis (EDA) and the K-Means Clustering Algorithm, the method used is to determine the sales level based on the attributes used, the sales level will later be divided into 3, namely clusters 1 is high, cluster 2 is medium and cluster 3 is low. Based on the results of the EDA and K-Means methods, the results of a sales comparison for 4 years have a average value, indicating that sales in 2019 have large average values between 2020, 2021 and 2022. From the visualization results it can be concluded that there are 27 products in the cluster-1 category, 10 products in the cluster-2 category, and 5 products in the cluster-3 category.