Customer Migration Analysis 

Purpose

Customer Migration Analysis enables you to measure the customer migration and deviation. Using this analysis, you can compare:

·        the results obtained from the same data mining method (for example, two results of ABC Classification or two results of Clustering)

·        the results obtained from the two different data mining method (for example, results of ABC Classification and Clustering or Clustering and Decision Tree)

The Business Content for Customer Migration Analysis is provided for two scenarios:

·        CRM scenario where you can compare the analyses built on Business Partners (BPARTNER)

·        R/3 scenario where you can compare the analyses built on Customers (0CUSTOMER)

 

 For each Business Partner, the accumulated sales is retrieved from the SAP CRM Sales for a specific Marketing Campaign. For example, Query ABC Scoring (0CSALMC02_Q004) on CRM Sales Funnel / Pipeline Cube (0CSALMC02). After this, you perform ABC Classification (for example, ABC Analysis (Sales Volume) (0CRM_ABC_SAL). 

When you repeat the campaign  sometime later, Customer Migration Analysis enables you to monitor how many customers who were classified as A in the first campaign have now turned into B customers in the second campaign. This means that the customers who were first classified as A in the first analysis have now migrated to B in the second analysis. This group of  customers who have migrated could be the target group for a follow-up campaign, for example, a campaign that has the goal to prevent churn.

The queries for Customer Migration Analysis follow a pattern. Following types of queries are available:

• Queries for Migration Analysis
• Queries for Deviation Analysis
• Pre-queries for Deviation Analysis

Let us suppose that the first analysis produces for each Business Partner a result field <Field_1>. This field carries the values <Field_1_Value_1>, <Field_1_Value_2>, <Field_1_Value_3>, … , <Field_1_Value_M>.

The second analysis  produces for each Business Partner a result field <Field_2>. This field carries the values <Field_2_Value_1> , <Field_2_Value_2>, <Field_2_Value_3>, … , <Field_2_Value_N>.

The queries for customer migration computes how many Business Partners, which have the value <Field_1_Value_K> in the first analysis, have value <Field_2_Value_L> in the second analysis for K=1, 2, … , M and L=1, 2, … , N. This applies to all Business Partners, which appear in both the first and the second analysis.

Prerequisites

• Queries to extract Customer data for the analysis process has been defined (based on either Business Partner or Customer Number)
• A Model has been created in the Data Mining Environment (which uses the query to extract the data for the analysis process)

·         At least two models are Trained in the Data Mining Environment

• For the trained Models, DataSources have been maintained
• The Business Content for Customer Migration is Activated
• SAP BW is attached to the activated InfoSources as SourceSystem and Transfer Rules have been maintained
• At least two results from data mining methods have been loaded into the InfoProviders for the comparison.

Process Flow

The process flow for Customer Migration Analysis is illustrated in the following diagram: