Providing Synthetic Data Mining Model Using Association Rules and Clustering for Determining Discounting Strategy (Case Study: Pegah Distribution Co.)

Sales promotion is important issue in most of sales and distribution companies and finding the most appropriate strategy for this subject is marketers’ challenge. Discounting (offering) is one of sales promotion strategies. Using the fixed and constant discounting strategy for all customers and on all goods reduces chance for success. Discounting strategy needs a model for providing best prices to customers. This is applied descriptive survey in which data mining was applied as a tool for creating a discounting strategy. In this research, association rule and clustering were used for determining an appropriate discounting strategy.
RFM and K-Means were applied to cluster customers and achieve optimized cluster number. Then, rules available in each cluster for different goods were ex-tracted. Finally, appropriate discounting strategies were considered according to features of customers in each cluster. Findings indicated that the best number of clusters for this company’s customers is 8 clusters. The customers were catego-rized into two categories of premium and hesitant. Discounting strategies were identified based on results from association rule and interview with experts and specific discounting policies were assigned for different purchases of each cate-gory.