How to carry out refined operations of financial products?

Now that the cost of acquiring customers is getting higher and higher, I believe every operator is aware of the importance of conversion, retention and activation. RFM can help us better understand users and implement accurate operational methods. Below I will share with you the entire process of my data analysis using the RFM model.

1. First, determine three important indicators based on the product

I am in charge of a financial product, and the KPI is the recharge amount and the number of new recharges. So I determined RFM as the recharge amount, recharge times, and login times respectively.

(Specific indicators depend on your own products. For example, social products can be: number of logins, number of followers, and usage time)

The picture shows the data model I created. There are 998 user data in total. It is an array randomly generated by the RANDBETWEEN function. It is not real user data and is only used for analysis.

2. Now let’s calculate the R, F, and M values.

The first step is to group the data into 5 levels. The grouping can be based on the company's business. My grouping here is to first draw a trend chart of the user's consumption amount, and then group them according to the data trend.

As can be seen from the figure, there are obvious distribution nodes, so I divided these 5 areas into:

• 1=10~100
• 2=101~1200
• 3=1201~5000
• 4=50001~12000
• 5=12001~20000

Next, the value corresponding to each recharge amount is calculated based on the partition data, and the function is as follows:

 =IF（B2>12000,5,IF（B2>5000,4,IF（B2>1200,3,IF（B2>100,2,1））））

The calculation method of F and M is the same.

After grading the consumption frequency, these five areas are divided into:

• 1=1~5
• 2=6~7
• 3=8~10
• 4=10~15
• 5=15~20

The function is as follows:

 =IF（C2>15,5,IF（C2>10,4,IF（C2>7,3,IF（C2>5,2,1））））

Next is the number of logins:

• 1=1~10
• 2=11~20
• 3=21~30
• 4=31~40
• 5=41~50

The function is as follows:

 =IF（D2>40,5,IF（D2>30,4,IF（D2>20,3,IF（D2>10,2,1））））

3. Find the average RFM

After finding the corresponding values, use the AVERAGE function to calculate the average value of each column:

• Average value of R = 1.19
• Mean value of F = 1.28
• Mean value of M = 3.33

(Some friends asked me why we need to calculate the value and then find the average, instead of finding the average based on the original consumption amount? Think about it, when the judges give scores in the competition, they will remove the highest score and the lowest score. The principles of these two are actually a bit similar. This will make the array more integrated.)

4. Determine the RFM level

Next, determine the RFM level. By comparing with the average, we can find out the value.

The function is as follows:

• R=IF(E2>1.19,"High","Low")
• F=IF(F2>1.28,"High","Low")
• M=IF(G>3.33,"High","Low")

5. Determine user value

Now comes the most important step, which is to calculate the user's value. Use IF and AND functions to determine what the corresponding values ​​are for users with the same high value, the same low value, or different high and low values.

The function is as follows:

 =IF（AND（H2="高",I2="高",J2="高"）,"重要價值用戶", IF（AND（H2="高",I2="低",J2="高"）,"重要發展用戶", IF（AND（H2="低",I2="高",J2="高"）,"重要保持用戶", IF（AND（H1="低",I2="低",J2="高"）,"重要挽留用戶", IF（AND（H2="高",I2="高",J2="低"）,"一般價值用戶", IF（AND（H2="高",I2="低",J1="低"）,"一般發展用戶", IF（AND（H1="低",I2="高",J2="低"）,"一般保持用戶","一般挽留用戶" ）））））））

Finally, different operational measures can be taken based on the judgment results. Here are 2 specific implementation workflows for your reference:

Important value users are the top customers in our pyramid, so we need to discover their greater value. Through data comparison, we found that this type of users are concentrated in the male group aged 30-40, who have money and time. We provide these users with large coupons, as well as privileges and lucky draws after successful recharge to stimulate them to complete large recharges.

Important to retain users, these users are mostly young people between 18-25 years old, with low recharge amount and frequent login. Young people have a strong desire to share. Through testing, we selected more interesting content and copywriting to guide them to recommend it to their friends in order to expand the user base and increase the sharing rate by 10%.

…

Through testing of users at different levels, you can summarize the characteristics of this group of users, and then mobilize colleagues to put forward ideas and opinions. If necessary, you should also invite users to ask questions. After summarizing the opinions, select feasible solutions through AB testing. At this time, you will know what to do next...

The above is the refined stratification and operation process of financial products. I’m new here, so please give me your advice! Welcome to give your comments and suggestions~Continue learning and progress!

Author: Wang Youyou

Source: Wang Youyou