Aluminum is widely used in a variety of things, so that the emerging companies to process aluminum. More and more businesses in the same pose competition. To overcome this need a good marketing strategy. One study done analyzing Market Basket to determine product recommendations to customers. Many researchers associate a priori approach or fp-growth in analyzing market basket. But for a large dataset, the results of the application of the association becomes less accurate. Therefore, a large dataset to be segmented k-means that large datasets are divided into several smaller segments. Analyzing the results obtained in the market basket by applying k-means algorithm and fp-growth algorithm is shown to improve the accuracy of 70% to 90%, 80% and 90%. Appropriate product recommendations to assist in the marketing strategy,especially in the field of promotional products and to assist production planning. Keywords: k-means algorithm, fp-growth algorithm, market basket, product recommendation.