Analysis of RFM model

The meaning of RFM is as follows:

1. R (Recency): the interval between the customer’s last transaction time. The larger the R value, the longer the date when the customer's transaction occurred, and vice versa, the closer the date when the customer's transaction occurred.

2. F (Frequency): The number of transactions the customer has made in the recent period. The larger the F value is, the more frequent the customer's transactions are, and the opposite indicates that the customer's transactions are not active enough.

3. M (Monetary): The amount of transactions made by the customer in the recent period. The larger the M value is, the higher the customer value is, and vice versa.

4. RFM analysis is a method of segmenting customer value based on the contribution of customer activity and transaction amount.

The rfm analysis method is as follows:

We usually use the format of transaction data for analysis. Because transaction data can be organized into customer data, but customer data cannot be restored into transaction data. That is, you can use the fields of transaction data to get the fields of customer data, but not vice versa.

Whether it is "transaction data" or "customer data" depends on the format of the data source file.

Variables: Select each variable

Binning: What is the total score of the score

Save: Which new variables are generated, you can customize the name.

Output: All can be checked in order to fully interpret the RFM analysis results.

After confirmation, four new variables were generated

New-score: the time interval score of the last transaction;

Frequency-score: the total number of transactions Score;

Consumption amount-score: total transaction amount score;

RFM score: RFM score

Interpretation of analysis results:

The The graph is mainly used to see whether the number of customers with each RFM summary score is evenly distributed.

We expect a uniform distribution. If it is not evenly distributed, you should reconsider the applicability of RFM or try another binning method (reduce the number of bins or randomly assign binding values)

"RFM heat map" is a graphical representation of the average transaction amount on the matrix chart drawn by RS and FS. The color depth is used to indicate the size of the average transaction amount. The darker the color, the average customer transaction amount in the corresponding matrix block. The higher.

As in this example, as the scores of RS and FS increase, the color becomes darker and darker, which means that the more recent the customer's last transaction is and the more times the customer has made transactions, the higher the average transaction amount will be.

This graph is a scatter graph between the last transaction time, the total number of transactions, and the total transaction amount.

Through the scatter plot, you can clearly and intuitively see the relationship between the three analysis indicators, which facilitates the evaluation of indicator correlation.

In this example, there is an obvious linear relationship between the total number of transactions and the total transaction amount, while the correlation between the last transaction time and the other two analysis indicators is weak.