Currently, companies operating in the motorbike sector are experiencing a decline in sales of spare part products and are not appropriate in determining the promotional strategies given to customers. A lot of transaction data is used as a reference for selling products at capital prices which only results in small profits. If it is still not sold, the incoming goods are delayed because the capital has not been returned. Therefore, a system is needed to process information data more quickly and precisely in predicting motorbike sales patterns using a priori algorithm data mining applications. The results of calculations using the a priori algorithm show that if a consumer buys a Yamaha lower arm and tires then the support value = 23.33 and the confidence value = 77.78 and if the consumer buys a Yamaha lower arm and AC filter then the support value = 26.67 and the confidence value = 72.72. The research objective is to predict and analyze motorbike sales patterns implemented on a desktop-based application. This is to make it easier to analyze the competitiveness of the best-selling motorcycle products simultaneously. As a recommendation for decision makers to improve marketing and promotion of better motorbike products.