Using RFM Methodology For Rule-Based (Dynamic) Segmentation
RFM Analysis is a substantial marketing model that analyzes customer’s purchase behavior, browsing behavior, cart abandoned behavior and email behavior to formulate rule-based dynamic Customer Segmentation. The objective of RFM Analysis is to segment customers according to their different behaviors and turn them into loyal customers by recommending products of their choice.
In RFM methodology, three parameters are analyzed, each denoted by the letters R, F, and M. To satisfy the need of knowing true customer value, analysis of just one parameter will give an inaccurate report of the customer base, so the customer segmentation won’t be that effective and even customer’s lifetime value can’t be reliable. This is why at least three parameters of customer’s behavior are analyzed; with the freedom to add other analytical parameters too.
RFM stands for Recency, Frequency and Monetary Value. For purchase behavior, Recency means how recently a customer bought from your store, Frequency means how frequently a customer is buying from you, and Monetary Value means what is the amount of money a customer spent in your store. These three parameters altogether determine the importance of customer for the retail shop and just like Purchase behavior, they are also obtained for Browsing, Cart abandonment, and Email behavior.
RFM Analysis is useful due to its simplicity, intuitiveness, and utilization of objective on numerical scales, that yields a comprehensive and informative depiction of customers. The output of this segmentation method is easy to interpret and adapt.
Before knowing the method by which RFM is calculated and analyzed, first, let’s check out the questions that RFM answers with reliable accountability.
Using RFM Analysis, retailers can find answers to the following questions, which have always placed a question mark between them and customers:
- Who are the loyal customers of a store?
- Which customers a store must retain?
- Which customers have the potential to convert into regular profitable customers?
- Who are those inconsistent customers who don’t engage regularly with the store?
- Which group of customers is most probable to respond to your campaign?
- Who are the churned out customers and who are about to churn?
RFM Analysis hands out the answers to these questions and brings affront the true customer value to the business.
As we have discussed the Importance of RFM Analysis, it’s time to know How RFM is calculated, and how it is analyzed to segment customers for campaigning.
RFM scores are calculated to know the purchase behavior, browsing behavior, cart abandoned behavior and email behavior of any customer. It provides a simple intuitive way of calculating each of the three aspects in a simple rating of 1-5, where 1 is the least important and 5 is the most important one. For example, a customer with R=5, F=5, and M=5 is the most profitable and loyal customer, while a customer with R=1, F=1, and M=1 is the least contributing one.
Although, RFM methodology can be applied to any of the customer behaviors but in the example below is the RFM calculation on the basis of Purchase Behavior. Following are the methods to evaluate the R, F, and M Values:
Recency (R) Value
Recency (R) Value represents how recently a customer has made a purchase. The number of days since the last purchase is acknowledged and a score of 1-5 is assigned to customers. The customers who made a purchase most recently are given a score of 5, while the customer who bought far back is given a score of 1.
For example:
Days since last purchase= R
0-20 = 5
20-100 = 4
100-300 = 3
300-1000 = 2
> 1000 = 1
Frequency (F) Value
Frequency (F) Value represents the number of purchases made by a customer. If a customer purchased 10 times over a period of time, the second customer purchased 7 times and the third customer purchased 5 times, then the first one will be assigned the F score of 5, second with 4 and third with 3 and so on.
Monetary Value (M)
Monetary Value (M) ranging from 1 to 5 denotes the monetary contribution of an individual customer. Every customer who contributes to the revenue of the store can be assigned a value of M.
For example, if there are 10 customers and they contributed a revenue of $10,000, then this total amount can be divided into 5 segments of $2000 each. If a customer contributed something between $500-$1500, then it will get the M score of 1. Similarly, if a customer contributed $2500 in the total revenue, then it will get a score of 2 and so on.
After figuring out the RFM Values of customers, proper customer segmentation is targeted to achieve. Let’s consider an example dataset of customer transactions to know how Customer Segmentation can be implemented.
Example:
| Customer ID | Recency | Frequency | Monetary |
| 1 | 3 | 5 | 520 |
| 2 | 5 | 19 | 920 |
| 3 | 45 | 1 | 35 |
| 4 | 22 | 2 | 65 |
| 5 | 14 | 3 | 159 |
| 6 | 31 | 2 | 56 |
| 7 | 6 | 3 | 120 |
| 8 | 49 | 1 | 930 |
| 9 | 33 | 14 | 2610 |
| 10 | 9 | 5 | 171 |
The above table contains the Recency, Frequency and Monetary Values for 10 customers based on their transactions with a store. Now let’s get ahead and find out how RFM analysis is made for Customer Segmentation.
RFM Analysis Of Customer Segmentation (RFM)
After R, F, and M Values are taken from the transactional history, each of them is categorized in increasing order for each customer. First, we’ll arrange all the (R) Values in increasing order for all customers and respectively score them with values of 1-5 in accordance with their related Receny. Let’s see how:
| Customer ID | Recency Value | Recency Score |
| 1 | 3 | 5 |
| 2 | 5 | 5 |
| 3 | 6 | 4 |
| 4 | 9 | 4 |
| 5 | 14 | 3 |
| 6 | 22 | 3 |
| 7 | 31 | 2 |
| 8 | 33 | 2 |
| 9 | 45 | 1 |
| 10 | 49 | 1 |
We’ve divided the Score in 5 quintiles of 20% each (which gives scoring of 1-5 to every two customers as per their Recency). Similarly, we will score the customer’s Frequency and Monetary Value in order of Most Frequent and Big Spenders to Least Frequent and Low Spenders.
| Customer ID | Frequency Value | Frequency Score |
| 1 | 14 | 5 |
| 2 | 10 | 5 |
| 3 | 5 | 4 |
| 4 | 5 | 4 |
| 5 | 3 | 3 |
| 6 | 3 | 3 |
| 7 | 2 | 2 |
| 8 | 2 | 2 |
| 9 | 1 | 1 |
| 10 | 1 | 1 |
| Customer ID | Monetary Value | Monetary Score |
| 1 | 2610 | 5 |
| 2 | 930 | 5 |
| 3 | 920 | 4 |
| 4 | 520 | 4 |
| 5 | 171 | 3 |
| 6 | 159 | 3 |
| 7 | 120 | 2 |
| 8 | 65 | 2 |
| 9 | 56 | 1 |
| 10 | 35 | 1 |
When we are done with assigning the RFM scores to each customer, we’ll now rank Customers by combining their individual R, F and M Scores. The RFM scores utilized for each customer will be the average of all the three individual values of RFM.
| Customer ID | RFM | RFM Score |
| 1 | (5,4,4) | 4.3 |
| 2 | (5,5,4) | 4.6 |
| 3 | (1,1,1) | 1.0 |
| 4 | (3,2,2) | 2.3 |
| 5 | (3,3,3) | 3.0 |
| 6 | (2,2,1) | 1.6 |
| 7 | (4,3,2) | 3.0 |
| 8 | (1,1,5) | 2.3 |
| 9 | (2,5,5) | 4.0 |
| 10 | (4,4,3) | 3.6 |
By figuring out the RFM Score, the customers having relevant scores can be grouped in a Customer Segment and then the related campaigns and promotions are made for each customer segment.
The implementation of RFM Analysis helps to figure out the most loyal customers, potential customers, new customers as well as those customers who perform differently throughout the sales channel.
However, Enalito uses the RFM methodology to implement rule-based customer segments, which are dynamic in nature. This gives a provision to retailers as the system tracks and updates the customer segments with the ever-changing and differing behaviors of an online store. In the absence of such a system, it would be hectic, and likely impossible for a retailer, to govern and handle everything by ownself. Enalito with Machine Learning and Robust methodology provides the solutions which are capable to uplift any retail business to unimagined success.
As it is clear now what is customer segmentation and how it is achieved by analyzing Browsing, Purchase, Cart abandoned and Email behavior with the vital role of marketing automation and RFM methodology, now it is time to know: How Enalito uses all of these things to deliver customer segments in an effective grid-based format to ensure that retailers get every actionable insight they deserve to get.