Customer prioritization using ABC Classification Model

[Private_Member_9643]

"80% of revenues are usually generated by 20% of your customer base."       -  Richard Koch

 but what if the stakes change… Read on to find out how to prioritize your customer profits.

Time and again, businesses have faced the dilemma of numbers against value. Is the business more conducive towards a vertically- deep niche market or is the business more conducive towards a horizontal number driven approach. The ABC classification model helps companies in analyzing the profitability in terms of the revenues generated by the three classes of customer base prioritized on the basis of some characteristics.

The ABC Classification Model of Data Mining is used to categorize objects on the basis of classification rules. Object could be any Characteristic which you like to categorize. Input to the Model will be generated by SAP BW query. Classification rules apply on any of Key Figure. We must have at least one characteristic and one key figure available for analyses whenever we are creating ABC Classification Model.

We can classify customers on the basis of revenue generated from them. We can divide them into different classes on the basis of different threshold values.

Example

Classification Rule:

Category as per Classification Rule (class)	Threshold Value (Revenue- thousand dollars)
A	0-50
B	51-300
C	301-1000

The output generated on the basis of above classification rule is:

Customer ID	Revenue(thousand dollars)	Class
CUST001	14	A
CUST002	100	B
CUST003	18	A
CUST004	800	C
CUST005	120	B
CUST006	200	B
…	…	…
…	…	…

Following are the types of fields which are required for creating this Model.

1. Classified Object – In our example its Customer ID
2. Classification Criterion – Revenues generated by customers.
3. Grouping Attribute – where to store result.

Following are the types of Classification Rules supported by ABC Classification method:

1. Absolute Values of Classification Criterion – Data is classified on the basis of values of Key Figure.
2. Cumulated Percentage of Classification Criterion- Data is classified on the basis of highest value to the lowest value of Key Figure.
3. Cumulated Percentage of Classified Object – Data is classified on the basis of combination of highest value of key figure and maximum number of records available for characteristics which is selected as a classified object.
4. Absolute Values of Classified Object – Data is classified on the basis of maximum to minimum number of records available for characteristics which is selected as a classified object.

This is the simple example of ABC Classification Model of Data Mining using which we can create any possible scenario for categorizing the data and making it available for any further analysis.

For any organization, all its customers are equally important but that does not necessitate into equal resource spending on all of them. Techniques like these help organizations to understand the more profitable customers from the ordinary ones and make use for further analytics to enhance their value offerings to this set of customers.