Customer-segmentation-using-RFM-

Data Description:

The data consists of 541909 rows x 8 columns, totaling in 33.1+ MB approximately. The Goal based on our features we have to categorize the customers.

• InvoiceNo
• StockCode
• Description
• Quantity
• InvoiceDate
• UnitPrice
• CustomerID
• Country

Data preprocessing:

• Transforn the InvoiceDate column from objest to date.

• Excluded Negative Values in Quantity and UnitPrice.

• Dropped all the Duplicated Values.

• Droped all the missing values in CustomerID column.

• Creating Total Price Column, whihc is the result of multipling the quanitity and the item price.

  RFM Ananlsis:

We are going to perform following opertaions:

• For Recency, Calculate the number of days between present date and date of last purchase each customer, InvoiceDate.
• For Frequency, Calculate the number of orders for each customer, InvoiceNo.
• For Monetary, Calculate sum of purchase price for each customer, Total_Price.

Then ordered the values from 1 to 5. Then combing them all, in the RFM score.

Also I created a segments for the customers:

• Promising.
• About to Sleep.
• Hibernating.
• Need Attention.
• Potential Loyalists.
• At Risk.
• Can't Loose.
• Loyal Customers.
• New Customers.
• Champions.