Strategy for recalling old customers——RFM model and application!

Suppose for some reason, you need to recall your old customers. Old customers with different consumption attribute levels require different recall trigger points, so you may need to treat your old customers in a stratified manner. At this time, a customer relationship management model was introduced: the RFM model . This article focuses on sharing ideas for recalling old customers based on the RFM model: how to divide old customers of different consumption levels into quadrants and prescribe the right medicine for customers in different quadrants.

1. Overview of RFM Model

Among the many customer relationship management (CRM) analysis models, the RFM model is widely mentioned.

The RFM model is an important tool and means to measure customer value and customer profitability. The mechanical model describes a customer's value status through three indicators: the customer's recent purchase behavior, the overall frequency of purchases, and how much money was spent.

According to Arthur Hughes of the American Database Marketing Institute, there are three elements in the customer database: R (Recency), F (Frequency), and M (Monetary).

1. Recent consumption R (Recency)

When was the customer's last purchase? Customers whose last purchase was made recently are the most likely to respond to the goods or services offered.

If the number of customers whose last purchase was very recent (consumption within 1 month) increases, it means that the company is growing steadily; otherwise, it is a sign of an unhealthy path.

It is much easier to get a customer who came in a few months ago to buy than to get a customer who came in more than a year ago.

2. Consumption frequency

The number of purchases made by a customer during a specified period. The customers who buy most frequently are also the most satisfied customers.

If you believe in brand and store loyalty, then the most frequent shoppers will be the most loyal. Increasing the number of times a customer buys means stealing market share from your competitors and earning sales at someone else's expense.

3. Monetary: The customer's purchase amount (can be divided into cumulative purchases and average purchases per time)

Spending amounts are the backbone of all database reports and can also verify Pareto's Law - 80% of a company's revenue comes from 20% of its customers.

"↑" indicates that the value is greater than the mean, and "↓" indicates that the value is less than the mean.

Because there are three variables, a three-dimensional coordinate system is used for display. The X-axis represents Recency, the Y-axis represents Frequency, and the Z-axis represents Monetary. The eight quadrants of the coordinate system represent eight types of users. According to the classification in the above table, the following graphic can be used to describe it:

The above is a rough framework introduction of the RFM model. Next, let’s talk about how to use this model to classify old customers in actual work.

2. RFM Standard Analysis

In similar CRM systems such as Shuyun, customers are divided into five categories. This five-category analysis is equivalent to a "loyalty ladder". The trick is to let consumers keep climbing up the ladder. Think of sales as pushing a two-time customer to a three-time customer, and turning a one-time buyer into a two-time buyer.

To facilitate the following explanation, the corresponding quadrants are represented by letters 1-25 (as shown in the figure below).

For example: a customer's F=1, 30＜R≤90, then he is in quadrant 22.

Before using this model to recall old customers, you need to first clarify the meaning of each quadrant:

1. Customers closer to the upper right quadrant are of higher quality, have a higher probability of repurchasing, and are more loyal to the brand.
2. Customers in quadrants 21-25 will directly become customers in quadrant 16 if they make one more purchase; customers in quadrants 6-10 will directly become customers in quadrant 1 if they make one more purchase.
3. Quadrant 25 belongs to lost customers, Quadrant 1 belongs to absolutely loyal old customers (this type of customer is most directly communicated by phone), and focus on customers in Quadrants 5 and 10 (why did your loyal old customers churn?)...

Take a look at the table below to get a more intuitive feeling.

The RFM standard analysis also derives another parameter: number of customers/percentage. Therefore, it can be divided into: quadrants based on the number of customers/percentage, quadrants based on the average purchase amount per time; quadrants based on the cumulative purchase amount.

Table 2.1 RFM standard analysis of quadrants divided by cumulative amount

Table 2.1 shows that the more times a customer purchases, the smaller the proportion. Pay attention to quadrant 24 (brown). This type of data means that you have too many lost customers! It’s time to pay more attention to the marketing work for new customers and move the customers in Quadrants 21-25 to Quadrant 16.

Table 2.2 RFM standard analysis of quadrants divided by average purchase amount

Analyzing Table 2.2 from the perspective of M (amount of consumption), we can focus on Quadrant 2 and Quadrant 3 (in yellow). These customers have high contributions and can be visited or contacted as a priority to regain more business opportunities in the most effective way.

Table 2.3 RFM standard analysis of quadrants divided by cumulative purchase amount

Table 2.3: Still analyzing from the perspective of M (consumption amount), we can find that the main contribution value of RMB lies in lost customers. In other words, you are squeezing too little money from old customers! Your CRM maintenance work is not good enough. The secondary recall of new customers is the key issue that needs to be focused on in the next stage.

3. Logic of recalling old customers based on RFM model

Let’s take another assumption:

You have 10,000 customers and need to recall them to the greatest extent possible (in terms of number of customers or spending amount) by sending text messages or emails. However, your budget is limited and you can only select 2,000-3,000 customers at most. So how would you find the most optimized customer sample?

Once you understand the logic of RFM, it is possible for ROI to jump from 1:6 to 1:30, and the cost savings in marketing will be considerable.

Based on the periodic changes in different quadrants, we can infer the changes in customer consumption and list the customers according to the possibility of customer churn.

If you don’t know how to take samples, just carpet bomb all customers in all quadrants and calculate the ROI of different quadrants. I have a plan for the next event~

Once you have a concept, you still have to keep trying → summarizing → adjusting to reach the most ideal state.

Next time, you can ask your boss with your chest held high:

“What ROI do you want to achieve by recalling old customers this time?”

References

1. "RFM Analysis of Customer Relationships on E-commerce Websites" Webmaster Home

2. Baidu Encyclopedia : “FRM Model”

Mobile application product promotion service: APP promotion service Qinggua Media advertising

This article was compiled and published by @廖小虫爱吃肉 by (Qinggua Media). Please indicate the author information and source when reprinting! Site Map