<![CDATA[This study aims to apply the K-Means Clustering algorithm to segment consumers of tempe products at MSME Pabrik Tempe Mas Suri. The main problem addressed in this study is the limited use of consumer transaction data as a basis for developing targeted marketing strategies. Transaction records are often used only as sales documentation, not as analytical input to understand consumer purchasing behavior. This study used a quantitative approach with a data mining method. The dataset consisted of 1,000 consumer transaction records collected from December 1, 2024, to April 30, 2025. Three main variables were analyzed, namely purchase frequency, purchase quantity, and total spending. The analysis process was carried out using RapidMiner Studio through data cleaning, attribute selection, Min-Max normalization, K-Means clustering, and cluster quality evaluation using the Davies-Bouldin Index. The clustering process tested several values of K, from K=2 to K=5. The results showed that K=2 produced the lowest actual DBI value of 0.569. However, K=5 was selected because it provided more detailed and applicable consumer segments for marketing purposes. The five segments consisted of Main Priority Consumers, Loyal Consumers, Medium Consumers, Potential Consumers, and Passive Consumers. Classification validation produced an accuracy of 94.50%, indicating that the segments were well differentiated. The study concludes that K-Means Clustering can help MSMEs identify consumer characteristics and formulate more specific data-driven marketing strategies.]]>
Copyrights © 2026
