Customer behavior analysis entails the analysis of the buying behavior, churn behavior, satisfaction, and loyalty of your customers. These aspects make up a customer’s behavioral profile, and you can then place customers exhibiting a similar behavioral profile in the same customer segment. The more you know about the behavioral profile of your customers, the more effective your measures become, such as measures to counteract customer churn. You can target individual customers or customer segments according to their needs, provide the customer care as appropriate, and thereby achieve greater customer satisfaction.

The most crucial part of customer behavior analysis relates to identifying hidden patterns in your customer behavior, which you do on the basis of historic data. By identifying patterns that could otherwise go unnoticed, you obtain powerful insights that allow you to implement measures with accuracy. Such hidden patterns are determined using data mining. SAP BW delivers various data mining methods (such as Clustering for customer segmentation). You can supplement these methods by creating your own models according to your requirements and then by using these to draw information relevant for decision-making from the data in your SAP BW. You can also analyze historic data to make predictions about the future. Alongside the actual data mining methods, SAP also delivers analytical applications that use these data mining methods.

You find the data mining methods in the SAP Business Information Warehouse (SAP BW) with Release 2.1C (Support Package 8) or higher.

The integration of the data mining methods with SAP BW means that you can use as source for the analysis any system that extracts data into SAP BW.

You need access in SAP BW to historic customer data for use as the basis for your analyses.

The following data mining methods are delivered by SAP and can be supplemented by models that you can create yourself:

• Decision trees
• Clustering
• Association analysis
• Weighted Score Tables (available with SAP BW with Release 2.1C (Support Package 10) or higher).
• Linear and nonlinear regression (available with SAP BW with Release 2.1C (Support Package 10) or higher).

In addition, there is an interface to the IBM Intelligent Miner.

For more information about the data mining methods delivered by SAP, see Business Content ® Analytical Applications ® Customer Relationship Management ® Data Mining in the SAP Library for SAP BW.

You can use the data mining methods for the purposes described below. In some cases, SAP also delivers analytical applications for the same purposes.

Decision trees in particular enable you to use historic data to identify which behavioral profile has frequently preceded the loss of a customer. When you then witness the same behavior in your active customers, you can take timely measures to retain customers that are liable to churn. With SAP BW 3.0B, SAP also delivers the analytical application Churn Management, which uses other data mining methods alongside the decision trees and thereby produces a comprehensive view of customer behavior. For more information, see the documentation in the SAP Library for SAP BW under Business Content ® Analytical Applications ® Customer Relationship Management ® Churn Management.

Association analysis in particular allows you to determine which products are generally bought in combination and hence determine product association rules. This information can then be converted into cross-selling opportunities. If you establish, for example, that a large number of customers who bought product A also bought product B, you can then offer product B to those specific customers who only bought product A. For this, you can use the analytical application Cross Selling, delivered with SAP BW 3.0B or higher. For more information, see the documentation in the SAP Library for SAP BW under Business Content ® Analytical Applications ® Customer Relationship Management ® Cross Selling.

You determine the satisfaction and loyalty of your customers from the answers customers provide to a survey. You can analyze the results and then use the customer knowledge base to draw parallels with customers‘ buying behavior, for example.

From SAP BW 3.0B, you can analyze customer satisfaction and loyalty using an analytical application that allows you to draw up and analyze surveys. For more information, see the documentation in the SAP Library for SAP BW under Business Content ® Analytical Applications ® Customer Relationship Management ® Customer Satisfaction and Loyalty.

The most important data mining method for customer segmentation is Clustering. You can use it to establish previously hidden groups of associated data from randomly ordered datasets. This method therefore allows you to achieve customer segments that each have typical behavioral profiles.