Customer value classification method and system based on RFAT model

Abstract

The invention discloses a customer value classification method and system based on an RFAT model, and relates to the technical field of customer classification, and the method comprises the steps: analyzing the transaction proximity, the customer frequency, the average transaction amount and the transaction trend of each customer in each gas station, and then classifying each customer in each gas station, the development trend of customers of each category is analyzed, then the category with the good development trend is selected for activity promotion, and the resource delivery proportion of the target category in each gas station in the future period is predicted, so that the defect of customer value classification based on an RFM model in the prior art is overcome, the diversity of customer classification results is improved, and the customer value classification efficiency is improved by using a "two-eight" criterion. According to the method, the flexibility of evaluation standards is improved, accurate data is provided for subsequent enterprise activity promotion and resource release, activity promotion is formulated for customer types with good development trends, the complexity and cost of enterprise activity promotion are greatly reduced, the satisfaction of customers is improved, and sustainable increase of services is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of customer classification, and in particular to a customer value classification method and system based on a RFAT model. Background Art

[0002] Gas stations are trading places where refined oil is sold. In the process of selling refined oil at gas stations, classifying the customer value at gas stations helps companies understand the development status of gas stations. At the same time, they can also analyze the development status of different types of customers based on the customer classification, and then formulate promotional activities for customer types with better development trends, so as to attract new customers and retain old customers, increase customer stickiness and satisfaction, and thus achieve sustainable business growth.

[0003] The existing customer value classification is mainly based on the RFM model, which quantifies customers based on three indicators: recency, frequency, and amount, and divides customers into lost customers, important recalled customers, new customers, general maintained customers, important retained customers, and important deep-cultivation customers. This classification results in fewer customer categories, and it is impossible to achieve in-depth analysis and classification of customer value, thereby failing to improve the diversity of customer classification results, and also failing to provide accurate data for subsequent corporate activity promotion and resource allocation, thus failing to improve customer retention and stickiness, and failing to achieve sustainable business growth;

[0004] The existing technical discrimination rules are based on the three average scores of RFM, with scores above the average being 1 and scores below the average being 0. For example, the scores of lost customers are (0,0,0) and those of important value customers are (1,1,1). This scoring standard has obvious disadvantages: it adopts a simple dichotomy and cannot accurately reflect the differences between customer groups. The evaluation standard is too mechanical and cannot meet the classification needs of different industries and different customer groups.

[0005] The existing technology specifies different strategies according to different categories when promoting activities, and does not screen out customers with better development trends according to the development status of different categories of customers, and then formulate activity promotions for customer types with better development trends. This greatly increases the complexity and cost of corporate activity promotion, and also increases the workload of gas station staff, thereby failing to improve the service effect of the staff, thereby failing to improve customer satisfaction, and failing to improve the operating efficiency of the gas station. Summary of the invention

[0006] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to provide a customer value classification method and system based on the RFAT model.

[0007] In order to solve the above technical problems, the present invention adopts the following technical solutions: In the first aspect, the present invention provides a customer value classification method based on the RFAT model, comprising the following steps: Step 1, gas station order acquisition: obtain the order information corresponding to each gas station within a specified period.

[0008] Step 2: RFAT data analysis: Based on the order information corresponding to each gas station, analyze the transaction recency, customer frequency, average transaction amount and transaction trend of each customer in each gas station.

[0009] Step 3: 80 / 20 classification: Based on the transaction recency, customer frequency, average transaction amount, and transaction trend corresponding to each customer in each gas station, obtain the standard transaction recency, standard customer frequency, and standard average transaction amount, and compare them with the transaction recency, customer frequency, and average transaction amount corresponding to each customer in each gas station. At the same time, compare the transaction trend of each customer in each gas station with 0 to obtain the category corresponding to each customer in each gas station within the specified period.

[0010] Step 4: Activity promotion category analysis: According to the categories corresponding to each customer in each gas station, obtain the customers corresponding to each category in each gas station, and then analyze the trend value of each category of customers in the specified period, and select the categories corresponding to the activity promotion in the future period, and then execute step 5.

[0011] Step 5. Activity promotion delivery forecast: obtain the activity promotion category corresponding to the company's future cycle, and record the activity promotion category corresponding to the company's future cycle as the target category, thereby obtaining the number of customers corresponding to the target category of each gas station in the specified period, and at the same time obtain the number of customers corresponding to the target category of each gas station in each historical period, and then predict the corresponding resource delivery ratio of each gas station in the future cycle.

[0012] Step 6. Display of resource allocation: Display the corresponding resource allocation ratio of each gas station in the future cycle.

[0013] Preferably, the analysis includes the transaction recency, customer frequency, average transaction amount and transaction trend of each customer in each gas station. The specific analysis process is as follows: the time and amount of each transaction corresponding to each customer are extracted from the order information corresponding to each gas station, and then the last transaction time of each customer in each gas station is extracted, and compared with the preset reference time to obtain the number of days between the last transaction time of each customer in each gas station and the reference time, which is used as the total transaction progress of each customer in each gas station.

[0014] The total number of transactions corresponding to each customer in each gas station is counted as the customer frequency of each customer in each gas station.

[0015] The amounts of each transaction corresponding to each customer at each gas station are accumulated to obtain the total transaction amount of each customer at each gas station, and then divided by the total number of transactions corresponding to each customer at each gas station to obtain the average transaction amount of each customer at each gas station.

[0016] Sort the transaction times of each customer at each gas station in ascending order, and then obtain the interval time between each adjacent transaction, recorded as Wherein j represents the number corresponding to each gas station, j=1,2......m, r represents the number corresponding to each customer, r=1,2......f, i represents the number corresponding to each transaction, i=1,2.......n, m, f, n are all arbitrary integers greater than 2.

[0017] According to the calculation formula Get the transaction trend T corresponding to the rth customer in the jth gas station jr .

[0018] Preferably, the steps for obtaining the category corresponding to each customer in each gas station within a specified period are as follows: S1. Integrate the transaction recency, customer frequency, average transaction amount, and transaction trend corresponding to each customer in each gas station to obtain a transaction recency set, a customer frequency set, an average transaction amount set, and a transaction trend set, and then select the values ​​of 80% of the corresponding positions in the transaction recency set, the customer frequency set, the average transaction amount set, and the transaction trend set as standard values, thereby obtaining standard recency, standard customer frequency, standard average transaction amount, and standard transaction trend.

[0019] S2. Compare the transaction recency, customer frequency, and average transaction amount corresponding to each customer in each gas station with the standard recency, standard customer frequency, and standard average transaction amount respectively. If the transaction recency corresponding to a customer in a gas station is greater than the standard recency, the transaction progress corresponding to the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0; similarly, if the customer frequency corresponding to a customer in a gas station is greater than the standard customer frequency, the customer frequency corresponding to the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0; if the customer frequency average transaction amount corresponding to a customer in a gas station is greater than the standard average transaction amount, the average transaction amount corresponding to the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0.

[0020] S3. If the transaction trend corresponding to a certain customer in a certain gas station is greater than or equal to 0, the transaction trend corresponding to the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0.

[0021] S4. According to the analysis process of steps S2-S3, the transaction recency, customer frequency, and average transaction amount of each customer in each gas station are obtained. If the transaction recency, customer frequency, average transaction amount, and transaction trend of a customer in a gas station are all 1, then the value of the customer in the gas station is 1111, and the category of the customer in the gas station is recorded as 1111. If the transaction recency of a customer in a gas station is 1, the customer frequency is 0, the average transaction amount is 1, and the transaction trend is 0, then the value of the customer in the gas station is 1010, and the category of the customer in the gas station is recorded as 1010. In this way, the category corresponding to each customer in each gas station is obtained.

[0022] Preferably, the analysis of the trend value of each category of customers within the specified period is specifically performed as follows: extracting the code of each customer corresponding to each category in each gas station within the specified period, extracting the code of each customer corresponding to each category in each gas station in each historical period from the database, and then comparing the code of each customer corresponding to each category in each gas station within the specified period with the code of each customer corresponding to each category in each gas station in each historical period; if the code of a customer corresponding to a category in a gas station within the specified period is different from the code of each customer corresponding to the category in the gas station in each historical period, the customer is recorded as a new customer, thereby counting the number of new customers of each category in each gas station in the specified period.

[0023] Count the number of customers of each category in each gas station during the specified period, and add up the total number of customers of each category in the specified period. Add the number of new customers of each category in each gas station during the specified period to get the total number of new customers of each category in the specified period. At the same time, extract the total number of customers and the total number of new customers of each category in each historical period from the database, and then substitute them into the calculation formula to calculate the trend value of each category of customers in the specified period.

[0024] Preferably, the calculation obtains the trend value of each category of customers in the specified period, and the specific calculation process is as follows: the number of customers corresponding to each category in the specified period and the total number of new customers are respectively recorded as Q1 q 、Q2 q , q represents the number corresponding to each category, q = 1, 2...p, p is any integer greater than 2, and the total number of customers and the total number of new customers corresponding to each category in each historical period are recorded as Q1 yq 、Q2 yq , where y represents the number corresponding to each historical period, y=1,2......z, z is any integer greater than 2.

[0025] According to calculation Get the trend value α of the qth category customer in the specified period q , where Q1 zq 、Q2zq They represent the number of customers in the qth category and the total number of new customers in the zth historical period, respectively, and κ1 q , κ2 q They respectively represent the historical customer quantity change rate and the historical total number of new customers change rate of the qth category, Δκ1 and Δκ2 are the preset difference in change rate of the number of licensed customers and the difference in change rate of the total number of new customers, ε1 and ε2 are the weight factors of the set customer quantity change rate and the weight factors of the total number of new customers, respectively.

[0026] Preferably, the specific selection process for selecting the category corresponding to the activity promotion in the future period is as follows: comparing the development trend value of each category of customers in the specified period with the preset development trend value threshold; if the development trend value of a certain category of customers in the specified period is greater than the preset development trend value threshold, then the category is recorded as a to-be-selected category, thereby obtaining each to-be-selected category; then extracting the number of new customers of each to-be-selected category in the specified period, and sorting each to-be-selected category in the specified period from large to small according to the total number of new customers, and selecting the first-ranked to-be-selected category as the category corresponding to the activity promotion in the future period.

[0027] Preferably, the prediction process of the resource allocation ratio of each gas station in the future period is as follows: the number of customers of the target category of each gas station in the specified period and the number of customers of the target category of each gas station in each historical period are calculated by the rate of change to obtain the rate of change of the number of customers of the target category of each gas station between adjacent periods, and then the maximum and minimum values ​​of the rate of change of the number of customers are removed, and then the reference rate of change of the number of customers of the target category of each gas station is obtained by the mean calculation.

[0028] Based on the number of customers of the target category corresponding to each gas station and the rate of change of the number of reference customers in the specified period, the number of reference customers of the target category corresponding to each gas station in the future period is calculated, and then the total number of reference customers of the target category in the future period is obtained by accumulation. The number of reference customers of the target category corresponding to each gas station in the future period is then divided by the total number of reference customers of the target category in the future period to obtain the corresponding resource allocation ratio of each gas station in the future period.

[0029] In a second aspect, the present invention provides a customer value classification system based on the RFAT model, comprising the following modules: a gas station order acquisition module, used to obtain order information corresponding to each gas station within a specified period.

[0030] The RFAT data analysis module is used to analyze the transaction recency, customer frequency, average transaction amount and transaction trend of each customer in each gas station based on the order information corresponding to each gas station.

[0031] The 80 / 20 criterion classification module is used to obtain the standard transaction recency, standard customer frequency, and standard average transaction amount based on the transaction recency, customer frequency, average transaction amount, and transaction trend corresponding to each customer in each gas station, and compare them with the transaction recency, customer frequency, and average transaction amount corresponding to each customer in each gas station. At the same time, the transaction trend of each customer in each gas station is compared with 0 to obtain the category corresponding to each customer in each gas station within a specified period.

[0032] The activity promotion category analysis module is used to obtain the customers corresponding to each category in each gas station according to the categories corresponding to each customer in each gas station, and then analyze the trend value of each category of customers in a specified period, and select the categories corresponding to the activity promotion in the future period.

[0033] The activity promotion delivery prediction module is used to obtain the category of activity promotion corresponding to the company's future period, and record the category of activity promotion corresponding to the company's future period as the target category, thereby obtaining the number of customers corresponding to the target category of each gas station in the specified period, and at the same time obtaining the number of customers corresponding to the target category of each gas station in each historical period, and then predicting the corresponding resource delivery ratio of each gas station in the future period.

[0034] The display terminal is used to display the corresponding resource allocation ratio of each gas station in the future cycle.

[0035] The beneficial effects of the present invention are as follows: the present invention provides a customer value classification method and system based on the RFAT model, by analyzing the transaction recency, customer frequency, average transaction amount and transaction trend of each customer in each gas station, and then classifying each customer in each gas station, analyzing the customer development trend of each category, and then selecting the target category for activity promotion, and predicting the resource allocation ratio of the target category in each gas station in the future cycle, solving the shortcomings of the prior art based on the RFM model for customer value classification, improving the diversity of customer classification results, using the "28" criterion when classifying customers, accurately reflecting the differences between customer groups, improving the flexibility of the evaluation criteria, being able to adapt to the classification needs of different industries and different customer groups, providing accurate data for subsequent corporate activity promotion and resource allocation, and formulating activity promotion for customer types with better development trends, greatly reducing the complexity and cost of corporate activity promotion, and also reducing the workload of gas station staff, thereby improving the service effect of staff, improving customer satisfaction, and also improving customer retention and stickiness, and achieving sustainable business growth. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 The present invention is a schematic flow chart of the steps for implementing the method.

[0038] Figure 2 It is a schematic diagram of the system structure connection of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] See also Figure 1 As shown, in the first aspect, the present invention provides a customer value classification method based on the RFAT model, comprising the following steps: Step 1, gas station order acquisition: obtain the order information corresponding to each gas station within a specified period.

[0041] It should be noted that the order information corresponding to each gas station within a specified period is extracted from the database; the order information includes the corresponding code of each customer, the time and amount of each transaction.

[0042] Step 2: RFAT data analysis: Based on the order information corresponding to each gas station, analyze the transaction recency, customer frequency, average transaction amount and transaction trend of each customer in each gas station.

[0043] In a specific embodiment, the transaction recency, customer frequency, average transaction amount and transaction trend corresponding to each customer in each gas station are analyzed, and the specific analysis process is as follows: the time and amount of each transaction corresponding to each customer are extracted from the order information corresponding to each gas station, and then the last transaction time corresponding to each customer in each gas station is extracted, and compared with the preset reference time to obtain the number of days between the last transaction time of each customer in each gas station and the reference time, which is used as the total transaction progress corresponding to each customer at each gas station.

[0044] The total number of transactions corresponding to each customer in each gas station is counted as the customer frequency of each customer in each gas station.

[0045] The amounts of each transaction corresponding to each customer at each gas station are accumulated to obtain the total transaction amount of each customer at each gas station, and then divided by the total number of transactions corresponding to each customer at each gas station to obtain the average transaction amount of each customer at each gas station.

[0046] Sort the transaction times of each customer at each gas station in ascending order, and then obtain the interval time between each adjacent transaction, recorded as Wherein j represents the number corresponding to each gas station, j=1,2......m, r represents the number corresponding to each customer, r=1,2......f, i represents the number corresponding to each transaction, i=1,2.......n, m, f, n are all arbitrary integers greater than 2.

[0047] According to the calculation formula Get the transaction trend T corresponding to the rth customer in the jth gas station jr .

[0048] It should be noted that customers are classified based on the RFM model, that is, based on four indicators: transaction recency, customer frequency, average transaction amount and transaction trend.

[0049] Step 3: 80 / 20 classification: Based on the transaction recency, customer frequency, average transaction amount, and transaction trend corresponding to each customer in each gas station, obtain the standard transaction recency, standard customer frequency, and standard average transaction amount, and compare them with the transaction recency, customer frequency, and average transaction amount corresponding to each customer in each gas station. At the same time, compare the transaction trend of each customer in each gas station with 0 to obtain the category corresponding to each customer in each gas station within the specified period.

[0050] In a specific embodiment, the steps for obtaining the category corresponding to each customer in each gas station within a specified period are as follows: S1. Integrate the transaction recency, customer frequency, average transaction amount, and transaction trend corresponding to each customer in each gas station to obtain a transaction recency set, a customer frequency set, an average transaction amount set, and a transaction trend set, and then select the values ​​of 80% of the corresponding positions in the transaction recency set, the customer frequency set, the average transaction amount set, and the transaction trend set as standard values, thereby obtaining standard recency, standard customer frequency, standard average transaction amount, and standard transaction trend.

[0051] It should be noted that, if there are 10 values ​​in the transaction proximity set, the value at the 80% corresponding position is the 8th one. For example, the transaction proximity set is {1, 2, 3, 4, 5, 6, 7, 8, 9, 10}, then the value at the 80% corresponding position is the 8th one in the set, that is, 8, and 8 is the standard value.

[0052] It should also be noted that users can customize the selection criteria for standard values, such as the "four-six" criterion, selecting the values ​​at 60% of the corresponding positions in the set as the standard values.

[0053] S2. Compare the transaction recency, customer frequency, and average transaction amount corresponding to each customer in each gas station with the standard recency, standard customer frequency, and standard average transaction amount respectively. If the transaction recency corresponding to a customer in a gas station is greater than the standard recency, the transaction progress corresponding to the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0; similarly, if the customer frequency corresponding to a customer in a gas station is greater than the standard customer frequency, the customer frequency corresponding to the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0; if the customer frequency average transaction amount corresponding to a customer in a gas station is greater than the standard average transaction amount, the average transaction amount corresponding to the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0.

[0054] S3. If the transaction trend corresponding to a certain customer in a certain gas station is greater than or equal to 0, the transaction trend corresponding to the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0.

[0055] It should be noted that a transaction trend equal to 0 indicates that the customer's purchasing behavior is fixed; a value less than 0 indicates that the customer's consumption interval is shortened and they are increasingly inclined to consume at the company; a value greater than 0 indicates that the customer's consumption interval is prolonged and they are gradually leaving the company.

[0056] S4. According to the analysis process of steps S2-S3, the transaction recency, customer frequency, and average transaction amount of each customer in each gas station are obtained. If the transaction recency, customer frequency, average transaction amount, and transaction trend of a customer in a gas station are all 1, then the value of the customer in the gas station is 1111, and the category of the customer in the gas station is recorded as 1111. If the transaction recency of a customer in a gas station is 1, the customer frequency is 0, the average transaction amount is 1, and the transaction trend is 0, then the value of the customer in the gas station is 1010, and the category of the customer in the gas station is recorded as 1010. In this way, the category corresponding to each customer in each gas station is obtained.

[0057] Step 4: Activity promotion category analysis: According to the categories corresponding to each customer in each gas station, obtain the customers corresponding to each category in each gas station, and then analyze the trend value of each category of customers in the specified period, and select the categories corresponding to the activity promotion in the future period, and then execute step 5.

[0058] In a specific embodiment, the trend value of each category of customers in the specified period is analyzed, and the specific analysis process is as follows: extract the code of each customer corresponding to each category in each gas station in the specified period, extract the code of each customer corresponding to each category in each gas station in each historical period from the database, and then compare the code of each customer corresponding to each category in each gas station in the specified period with the code of each customer corresponding to each category in each gas station in each historical period. If the code of a customer corresponding to a category in a gas station in the specified period is different from the code of the customer corresponding to the category in the gas station in each historical period, the customer is recorded as a new customer, and the number of new customers of each category in each gas station in the specified period is counted.

[0059] Count the number of customers of each category in each gas station during the specified period, and add up the total number of customers of each category in the specified period. Add the number of new customers of each category in each gas station during the specified period to get the total number of new customers of each category in the specified period. At the same time, extract the total number of customers and the total number of new customers of each category in each historical period from the database, and then substitute them into the calculation formula to calculate the trend value of each category of customers in the specified period.

[0060] In the above, the calculation obtains the trend value of each category of customers in the specified period. The specific calculation process is as follows: the number of customers corresponding to each category in the specified period and the total number of new customers are recorded as Q1 and Q2, respectively. q 、Q2 q , q represents the number corresponding to each category, q = 1, 2...p, p is any integer greater than 2, and the total number of customers and the total number of new customers corresponding to each category in each historical period are recorded as Q1 yq 、Q2 yq , where y represents the number corresponding to each historical period, y=1,2......z, z is any integer greater than 2.

[0061] According to calculation Get the trend value α of the qth category customer in the specified period q , where Q1 zq 、Q2 zq They represent the number of customers in the qth category and the total number of new customers in the zth historical period, respectively, and κ1 q , κ2 q They respectively represent the historical customer quantity change rate and the historical total number of new customers change rate of the qth category, Δκ1 and Δκ2 are the preset difference in change rate of the number of licensed customers and the difference in change rate of the total number of new customers, ε1 and ε2 are the weight factors of the set customer quantity change rate and the weight factors of the total number of new customers, respectively.

[0062] It should be noted that ε1 and ε2 are both greater than 0 and less than 1.

[0063] It should also be noted that the historical customer number change rate of the qth category κ1 q The calculation formula is: Where Q1( y-1 ) q represents the number of customers in the qth category in the y-1th historical period; the change rate of the total number of new customers in the qth category in history κ2 q The calculation formula is: Where Q2( y-1 ) q Represents the total number of new customers in the qth category in the y-1th historical period.

[0064] The specific selection process for selecting the categories for promotional activities corresponding to the future period is as follows: comparing the trend value of each category of customers in the specified period with the preset trend value threshold; if the trend value of a category of customers in the specified period is greater than the preset trend value threshold, then the category is recorded as a to-be-selected category, thereby obtaining the to-be-selected categories; then extracting the number of new customers of each to-be-selected category in the specified period, and sorting each to-be-selected category in the specified period from large to small according to the total number of new customers, and selecting the first-ranked to-be-selected category as the category for promotional activities corresponding to the future period.

[0065] The embodiment of the present invention screens out categories with better development trends, and then selects the category with the most new customers as the category for activity promotion, thereby maintaining the retention rate and viscosity of new users to the greatest extent and improving the effect of activity promotion.

[0066] Step 5. Activity promotion delivery forecast: obtain the activity promotion category corresponding to the company's future cycle, and record the activity promotion category corresponding to the company's future cycle as the target category, thereby obtaining the number of customers corresponding to the target category of each gas station in the specified period, and at the same time obtain the number of customers corresponding to the target category of each gas station in each historical period, and then predict the corresponding resource delivery ratio of each gas station in the future cycle.

[0067] It should be noted that the number of customers corresponding to the target category of each gas station in each historical period is extracted from the database.

[0068] In a specific embodiment, the resource allocation ratio of each gas station in the predicted future period is predicted, and the specific prediction process is as follows: the number of customers of the target category of each gas station in the specified period and the number of customers of the target category of each gas station in each historical period are calculated by the change rate to obtain the change rate of the number of customers of the target category of each gas station between adjacent periods, and then the maximum and minimum values ​​of the customer number change rate are removed, and then the reference customer number change rate of the target category of each gas station is obtained by mean calculation.

[0069] It should be noted that, referring to the calculation of the historical customer quantity change rate of the qth category, the customer quantity change rate of each gas station corresponding to the target category between each adjacent period is calculated.

[0070] Based on the number of customers of the target category corresponding to each gas station and the rate of change of the number of reference customers in the specified period, the number of reference customers of the target category corresponding to each gas station in the future period is calculated, and then the total number of reference customers of the target category in the future period is obtained by accumulation. The number of reference customers of the target category corresponding to each gas station in the future period is then divided by the total number of reference customers of the target category in the future period to obtain the corresponding resource allocation ratio of each gas station in the future period.

[0071] Step 6. Display of resource allocation: Display the corresponding resource allocation ratio of each gas station in the future cycle.

[0072] See also Figure 2 As shown, in the second aspect, the present invention provides a customer value classification system based on the RFAT model, including the following modules: a gas station order acquisition module, a RFAT data analysis module, a 28 rule classification module, an activity promotion category analysis module, an activity promotion delivery prediction module, a display terminal and a database.

[0073] The gas station order acquisition module is used to obtain the order information corresponding to each gas station within a specified period.

[0074] The RFAT data analysis module is used to analyze the transaction recency, customer frequency, average transaction amount and transaction trend of each customer in each gas station based on the order information corresponding to each gas station.

[0075] The 80 / 20 criterion classification module is used to obtain the standard transaction recency, standard customer frequency, and standard average transaction amount based on the transaction recency, customer frequency, average transaction amount, and transaction trend corresponding to each customer in each gas station, and compare them with the transaction recency, customer frequency, and average transaction amount corresponding to each customer in each gas station. At the same time, the transaction trend of each customer in each gas station is compared with 0 to obtain the category corresponding to each customer in each gas station within a specified period.

[0076] The activity promotion category analysis module is used to obtain the customers corresponding to each category in each gas station according to the categories corresponding to each customer in each gas station, and then analyze the trend value of each category of customers in a specified period, and select the categories corresponding to the activity promotion in the future period.

[0077] The activity promotion delivery prediction module is used to obtain the category of activity promotion corresponding to the company's future period, and record the category of activity promotion corresponding to the company's future period as the target category, thereby obtaining the number of customers corresponding to the target category of each gas station in the specified period, and at the same time obtaining the number of customers corresponding to the target category of each gas station in each historical period, and then predicting the corresponding resource delivery ratio of each gas station in the future period.

[0078] The display terminal is used to display the corresponding resource allocation ratio of each gas station in the future cycle.

[0079] The database is used to store the order information corresponding to each gas station in a specified period, the total number of customers and the total number of new customers corresponding to each category in each historical period, the code of each customer corresponding to each category in each gas station in each historical period, and the number of customers of the target category corresponding to each gas station in each historical period.

[0080] The embodiment of the present invention analyzes the transaction recency, customer frequency, average transaction amount and transaction trend of each customer in each gas station, and then classifies each customer in each gas station, analyzes the customer development trend of each category, and then selects the target category for activity promotion, and predicts the resource allocation ratio of the target category in each gas station in the future cycle, thereby solving the shortcomings of the prior art based on the RFM model in customer value classification, improving the diversity of customer classification results, using the "28" criterion when classifying customers, accurately reflecting the differences between customer groups, improving the flexibility of evaluation standards, and being able to adapt to the classification needs of different industries and different customer groups, providing accurate data for subsequent corporate activity promotion and resource allocation, and formulating activity promotion for customer types with better development trends, greatly reducing the complexity and cost of corporate activity promotion, and also reducing the workload of gas station staff, thereby improving the service effect of staff, improving customer satisfaction, and also improving customer retention and stickiness, and achieving sustainable business growth.

[0081] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they shall all fall within the protection scope of the present invention.

Claims

1. A customer value classification method based on the RFAT model, characterized in that: The steps include: Step 1: Obtaining gas station orders: Obtaining order information corresponding to each gas station within a specified period; Step 2: RFAT data analysis: Based on the order information of each gas station, analyze the transaction recency, customer frequency, average transaction amount and transaction trend of each customer in each gas station; Step 3, 80 / 20 classification: Based on the transaction recency, customer frequency, average transaction amount, and transaction trend of each customer at each gas station, obtain the standard transaction recency, standard customer frequency, and standard average transaction amount, and compare them with the transaction recency, customer frequency, and average transaction amount of each customer at each gas station. At the same time, compare the transaction trend of each customer at each gas station with 0 to obtain the category corresponding to each customer at each gas station within the specified period; Step 4: Activity promotion category analysis: According to the categories corresponding to each customer in each gas station, obtain the customers corresponding to each category in each gas station, and then analyze the trend value of each category of customers in the specified period, and select the category corresponding to the activity promotion in the future period, and then execute step 5; Step 5: Prediction of activity promotion: Obtain the categories of activity promotion corresponding to the company's future cycle, and record the categories of activity promotion corresponding to the company's future cycle as target categories, thereby obtaining the number of customers corresponding to the target category of each gas station in the specified cycle, and at the same time obtain the number of customers corresponding to the target category of each gas station in each historical cycle, and then predict the corresponding resource allocation ratio of each gas station in the future cycle; Step 6. Display of resource allocation: Display the corresponding resource allocation ratio of each gas station in the future cycle.

2. A customer value classification method based on RFAT model according to claim 1, characterized in that: The specific analysis process of analyzing the transaction recency, customer frequency, average transaction amount and transaction trend of each customer at each gas station is as follows: Extract the time and amount of each transaction of each customer from the order information corresponding to each gas station, and then extract the last transaction time of each customer in each gas station, and compare it with the preset reference time to obtain the number of days between the last transaction time of each customer in each gas station and the reference time as the total transaction progress of each customer in each gas station; Count the total number of transactions corresponding to each customer at each gas station as the customer frequency of each customer at each gas station; The amount of each transaction corresponding to each customer at each gas station is accumulated to obtain the total transaction amount of each customer at each gas station, and then divided by the total number of transactions corresponding to each customer at each gas station to obtain the average transaction amount of each customer at each gas station; Sort the transaction times of each customer at each gas station in ascending order, and then obtain the interval between each adjacent transaction, recorded as Where j represents the number corresponding to each gas station, j = 1, 2...m, r represents the number corresponding to each customer, r = 1, 2...f, i represents the number corresponding to each transaction, i = 1, 2.......n, m, f, n are all any integers greater than 2; According to the calculation formula Get the transaction trend T corresponding to the rth customer in the jth gas station jr .

3. A customer value classification method based on RFAT model according to claim 1, characterized in that: The steps of obtaining the category corresponding to each customer at each gas station within a specified period are as follows: S1. Integrate the transaction recency, customer frequency, average transaction amount, and transaction trend corresponding to each customer in each gas station to obtain a transaction recency set, a customer frequency set, an average transaction amount set, and a transaction trend set, and then select the values ​​of 80% of the corresponding positions in the transaction recency set, the customer frequency set, the average transaction amount set, and the transaction trend set as standard values, thereby obtaining a standard recency, a standard customer frequency, a standard average transaction amount, and a standard transaction trend; S2. Compare the transaction recency, customer frequency, and average transaction amount of each customer in each gas station with the standard recency, standard customer frequency, and standard average transaction amount respectively. If the transaction recency of a customer in a gas station is greater than the standard recency, the transaction progress of the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0; similarly, if the customer frequency of a customer in a gas station is greater than the standard customer frequency, the customer frequency of the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0; if the customer frequency average transaction amount of a customer in a gas station is greater than the standard average transaction amount, the average transaction amount of the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0; S3. If the transaction trend corresponding to a certain customer in a certain gas station is greater than or equal to 0, the transaction trend corresponding to the customer in the gas station is assigned a value of 1, otherwise it is assigned a value of 0; S4. According to the analysis process of steps S2-S3, the transaction recency, customer frequency, and average transaction amount of each customer in each gas station are obtained. If the transaction recency, customer frequency, average transaction amount, and transaction trend of a customer in a gas station are all 1, then the value of the customer in the gas station is 1111, and the category of the customer in the gas station is recorded as 1111. If the transaction recency of a customer in a gas station is 1, the customer frequency is 0, the average transaction amount is 1, and the transaction trend is 0, then the value of the customer in the gas station is 1010, and the category of the customer in the gas station is recorded as 1010. In this way, the category corresponding to each customer in each gas station is obtained.

4. A customer value classification method based on RFAT model according to claim 1, characterized in that: The analysis specifies the trend value of each category of customers within a period, and the specific analysis process is as follows: Extract the code of each customer in each category of each gas station in the specified period, extract the code of each customer in each category of each gas station in each historical period from the database, and then compare the code of each customer in each category of each gas station in the specified period with the code of each customer in each category of each gas station in each historical period. If the code of a customer in a certain category of a gas station in the specified period is different from the code of the customer in the same category of the gas station in each historical period, then the customer is recorded as a new customer, and the number of new customers in each category of each gas station in the specified period is counted; Count the number of customers of each category in each gas station during the specified period, and add up the total number of customers of each category in the specified period. Add the number of new customers of each category in each gas station during the specified period to get the total number of new customers of each category in the specified period. At the same time, extract the total number of customers and the total number of new customers of each category in each historical period from the database, and then substitute them into the calculation formula to calculate the trend value of each category of customers in the specified period.

5. A customer value classification method based on RFAT model according to claim 4, characterized in that: The calculation obtains the trend value of each category of customers within a specified period. The specific calculation process is as follows: The number of customers in each category and the total number of new customers in the specified period are recorded as Q1 q 、Q2 q , q represents the number corresponding to each category, q = 1, 2...p, p is any integer greater than 2, and the total number of customers and the total number of new customers corresponding to each category in each historical period are recorded as Q1 yq 、Q2 yq , where y represents the number corresponding to each historical period, y=1,2......z, z is any integer greater than 2; According to calculation Get the trend value α of the qth category customer in the specified period q , where Q1 zq 、Q2 zq They represent the number of customers in the qth category and the total number of new customers in the zth historical period, respectively, and κ1 q , κ2 q They respectively represent the historical customer quantity change rate and the historical total number of new customers change rate of the qth category, Δκ1 and Δκ2 are the preset difference in change rate of the number of licensed customers and the difference in change rate of the total number of new customers, ε1 and ε2 are the weight factors of the set customer quantity change rate and the weight factors of the total number of new customers, respectively.

6. A customer value classification method based on RFAT model according to claim 5, characterized in that: The specific selection process of selecting the category of activity promotion corresponding to the future period is as follows: The trend value of each category of customers in the specified period is compared with the preset trend value threshold. If the trend value of a category of customers in the specified period is greater than the preset trend value threshold, then the category is recorded as the to-be-selected category, thereby obtaining the to-be-selected categories; then the number of new customers of each to-be-selected category in the specified period is extracted, and each to-be-selected category in the specified period is sorted from large to small according to the total number of new customers, and the to-be-selected category ranked first is selected as the category for corresponding activity promotion in the future period.

7. The customer value classification method based on the RFAT model according to claim 1 is characterized in that: The specific prediction process of predicting the resource allocation ratio of each gas station in the future cycle is as follows: The number of customers of the target category of each gas station in the specified period and the number of customers of the target category of each gas station in each historical period are calculated by the change rate to obtain the change rate of the number of customers of the target category of each gas station between adjacent periods, and then the maximum and minimum values ​​of the customer number change rate are removed, and then the reference customer number change rate of the target category of each gas station is obtained by the mean calculation; Based on the number of customers of the target category corresponding to each gas station and the rate of change of the number of reference customers in the specified period, the number of reference customers of the target category corresponding to each gas station in the future period is calculated, and then the total number of reference customers of the target category in the future period is obtained by accumulation. The number of reference customers of the target category corresponding to each gas station in the future period is then divided by the total number of reference customers of the target category in the future period to obtain the corresponding resource allocation ratio of each gas station in the future period.

8. A customer value classification system using the customer value classification method based on the RFAT model according to any one of claims 1 to 7, characterized in that: Includes the following modules: The gas station order acquisition module is used to obtain the order information corresponding to each gas station within a specified period; RFAT data analysis module, used to analyze the transaction recency, customer frequency, average transaction amount and transaction trend of each customer in each gas station according to the order information of each gas station; The 80 / 20 criterion classification module is used to obtain the standard transaction recency, standard customer frequency, and standard average transaction amount based on the transaction recency, customer frequency, average transaction amount, and transaction trend of each customer in each gas station, and compare them with the transaction recency, customer frequency, and average transaction amount of each customer in each gas station. At the same time, the transaction trend of each customer in each gas station is compared with 0 to obtain the category corresponding to each customer in each gas station within the specified period; The activity promotion category analysis module is used to obtain the customers corresponding to each category in each gas station according to the categories corresponding to each customer in each gas station, and then analyze the trend value of each category of customers in a specified period, and select the categories corresponding to the activity promotion in the future period; The activity promotion delivery prediction module is used to obtain the category of activity promotion corresponding to the enterprise's future period, and record the category of activity promotion corresponding to the enterprise's future period as the target category, thereby obtaining the number of customers corresponding to the target category of each gas station in the specified period, and at the same time obtaining the number of customers corresponding to the target category of each gas station in each historical period, and then predicting the corresponding resource delivery ratio of each gas station in the future period; the display terminal is used to display the corresponding resource delivery ratio of each gas station in the future period.

Images