Take Advantage of Association Analysis and See Your Sales Grow
- by Anton Karnaukhov, Senior IT Manager, Pacific Coast Companies, Inc.
- February 1, 2009
Most people are familiar with the “customers who bought this item also bought...” concept popularized by Web giants such as amazon.com. Association analysis, a data mining algorithm available in the Data Mining Workbench, helps you identify these related product sets. Find out how to set up association analysis in three steps.
The purpose of association analysis is to identify patterns and formulate rules that are applicable to a set of data. For example, association analysis could help a supermarket identify that when customers buy hamburger patties they are likely to buy hamburger buns 75% of the time. You can use the rules that association analysis produces to optimize marketing campaigns, product placement, and cross-selling opportunities.
Data mining has gained in popularity recently, with many companies using some form of data mining algorithms to extract value out of terabytes of data stored in their SAP ERP systems. For example, you can use data mining algorithms to identify the source of a specific hardware problem in a laptop by analyzing millions of support tickets and searching for commonalities.
Association analysis is a data mining algorithm that enables you to identify sets of products that share strong sales relationships. However, many people are unsure about how to carry out association analysis. The following step-by-step guide shows you how you can use the Data Mining Workbench and Analysis Process Designer (APD) to perform association analysis. Specifically, I will take a look at a store that sells three products: a DVD movie, a Blu-ray movie, and a movie poster. By using association analysis, I can analyze how these three products relate to each other and which of them are potentially good candidates for cross-selling opportunities.
Would you like to see this full item?