<![CDATA[In the era of business growth supported by information technology, business competition and information needs are increasing. However, many businesses, including PT Sila Tirta Gemilang, have not utilized it optimally. The company faces challenges in managing drinking water stock and predicting sales efficiently, which leads to stock buildup and demand uncertainty. This research aims to apply data mining to PT Sila Tirta Gemilang's sales strategy in the past year. The research uses the Knowledge Discovery in Databases (KDD) method with six stages: data selection, data cleaning, data transformation, data mining, evaluation, and knowledge presentation. The apriori algorithm is used to determine the frequency of item sets and find customer purchase patterns. The types of drinking water studied include D. 200 ML, D. 600 ML, K. 200 ML, K. 600 ML, S. 200 ML, QUA. F, Q. 600 ML, COCO, Fresh Tea, and GMES. The results showed that the largest support value for one item was 62.61%, two items were 24.9%, and three items were 7.2%. Overall, the confidence value is 89.5%, and the lift ratio is 1.670. The resulting 55 association rules can be used by companies to improve sales efficiency and effectiveness.Dalam era pertumbuhan bisnis yang didukung oleh teknologi informasi, persaingan bisnis dan kebutuhan informasi semakin meningkat. Namun, banyak bisnis, termasuk PT Sila Tirta Gemilang yang bergerak dalam penjualan air minum, belum memanfaatkannya secara optimal. Perusahaan ini menghadapi tantangan dalam mengelola stok air minum dan memprediksi penjualan secara efisien yang menyebabkan penumpukan stok dan ketidakpastian permintaan. Penelitian ini bertujuan menerapkan data mining untuk strategi penjualan PT Sila Tirta Gemilang dalam satu tahun terakhir. Penelitian menggunakan metode Knowledge Discovery in Database (KDD) dengan enam tahapan, data selection, data cleaning, data transformation, data mining, evaluation, dan knowledge presentation. Algoritma apriori digunakan untuk mengetahui frekuensi itemset dan mencari pola pembelian pelanggan. Jenis air minum yang diteliti meliputi D. 200 ML, D. 600 ML, K. 200 ML, K. 600 ML, S. 200 ML, QUA. F, Q. 600 ML, COCO, Fresh Tea, dan GMES. Hasil penelitian menunjukan nilai support terbesar untuk satu itemset adalah 62.61%, dua itemset 24.9%, dan tiga itemset 7.2%. Secara keseluruhan, nilai confidence sebesar 89.5% dan lift ratio 1.670. Dihasilkan 55 aturan asosiasi yang dapat digunakan perusahaan untuk meningkatkan efisiensi dan efektivitas penjualan.]]>
