<![CDATA[Insurance is a mechanism of protection or protection from the risk of loss by transferring the risk to another party. Sometimes a product that has just emerged becomes a product that is superior in terms of sales, so that interest in a product is not absolutely measured from the year the product was released. The constraint factors include the marketing of the product when it was launched. Offering products with low premiums along with the benefits that customers want. However, insurance companies still have difficulty in classifying superior products that are in great demand by prospective customers. For this reason, a technique for grouping insurance products is needed to make it easier for companies to see superior products and choose products that suit the needs of their customers. Analyzing and processing data using the K-Means method in the clustering of insurance products is the aim of this study. The application of the K-Means algorithm is to help calculate the purity value from the results of the clustering carried out so that the clustering of insurance products is in accordance with the needs of its customers. The application of the K-Means method with clustering techniques for data mining produces information on insurance products that are more attractive to potential customers. This is very appropriate in grouping data types because it is easier to implement and its application can filter quickly and precisely. Calculations using the K-Means method with a data sample of 55 customers obtained 3 clusters, namely cluster 1 for fire insurance which has 30 customers, cluster 2 for accident insurance 24 people and cluster 3 for health insurance 1 person.]]>
