Clustering E-Commerce Data: Application of K-Means to the SME Fashion Industry

In today's digital era, data is a crucial component. Data is recorded and stored in media or databases to be processed for useful information. This research discusses grouping username data for Lou's Fashion Brand buyers from e-commerce using the K-means algorithm. The data source for this research was collected from the sales results of Lou's products from September 1, 2024, to March 10, 2024. The variables used were Username (Buyer), Shipping Cost Paid by Buyer, and Total Income. The data will be processed using the RapidMiner application with the K-means operator, which will divide the data into three clusters to show which cluster has the highest total. From processing this data, we identified three clusters: cluster_0 includes data with total income other than Rp. 118,000 and has a value in the shipping cost attribute paid by the buyer, with 10 data points; cluster_1 includes data with a total income of Rp. 118,000, with 18 data points; and cluster_2 includes data with total income other than Rp. 118,000 and no shipping cost value (Rp. 0), with 16 data points. These results can be used as input for Lou's fashion products to create products based on cluster_1, which has the highest total income. The purpose of this research is to perform customer segmentation to better understand and target different buyer groups.