A customer relationship management system under the SaaS model

Abstract

The present invention relates to the technical field of customer data analysis, and specifically to a customer relationship management system under the SaaS model. The system includes: A customer behavior collection module, based on customer purchase records and service request data, statistically calculates the customer's purchase frequency, purchase amount, and the number of service requests according to time periods, extracts the behavioral characteristics of each time period, and generates an analysis result of the customer's time-series behavioral characteristics. In the present invention, by statistically analyzing the customer purchase records and service request data over time periods and extracting behavioral characteristics, it is possible to dynamically capture changes in customer behavior. At the same time, by cross-matching customer behavior data and channel activity performance, it is possible to accurately judge the changes in customer behavior before and after an event, thereby optimizing marketing strategies. Combining with the prediction of customer behavior at multiple time nodes, it is possible to gain insight into the customer's future purchase needs and behavioral trends in advance, and optimize the timing in customer relationship management.

Description

Technical Field

[0001] The present invention relates to the technical field of customer data analysis, and particularly to a customer relationship management system under the SaaS model. Background Art

[0002] In the technical field of customer data analysis, technical means such as data mining, machine learning, and statistical analysis are used to understand customer behavior, optimize the customer experience, and drive sales growth. At the same time, by analyzing data such as purchase history, customer feedback, and social media interactions, enterprises can identify consumer preferences and formulate more effective marketing strategies. In addition, analyzing customer data can also help enterprises improve customer satisfaction and loyalty, and optimize products and services.

[0003] Among them, the customer relationship management system is a technical solution for managing the interactions between an enterprise and its customers. The main functions include centrally storing customer information, tracking customer interactions, automating sales processes, and marketing activities. By providing a detailed view of customers, it helps enterprises serve customers more effectively, enhance customer relationships, thereby improving sales efficiency and the targeting of marketing. In addition, the system can also assist enterprises in providing the right products and services to the right customers at the right time, thereby optimizing customer lifecycle management and improving customer maintenance efficiency.

[0004] In the existing technology for the processing and analysis of customer data, more attention is paid to the static storage of customer information and the management of customer interaction records, making it difficult to effectively capture customer behavior information that changes over time. This data processing method makes it difficult for enterprises to respond promptly and accurately in customer relationship maintenance. In addition, for the evaluation of channel activities in the existing technology, it is difficult to understand the actual impact of activities on customer behavior based on the information of customer behavior changes before and after the activities. In addition, in terms of customer churn risk management and the exploration of potential customer demand data, enterprises can only react after obvious changes in customer behavior, which is not conducive to quickly capturing changes in customer behavior trends and potential demands. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose a customer relationship management system under the SaaS model.

[0006] To achieve the above purpose, the present invention adopts the following technical solution: A customer relationship management system under the SaaS model includes:

[0007] The customer behavior collection module statistically analyzes the purchase frequency, purchase amount, and service request times of customers according to time periods based on customer purchase records and service request data, extracts the behavioral characteristics of each time period, and generates the analysis results of customer time-series behavioral characteristics;

[0008] Based on the analysis results of the customer time-series behavior characteristics, the channel activity analysis module cross-matches the duration and frequency of channel activities with customer behavior data through the multi-tenant architecture of SaaS, determines the behavior changes of customers before and after channel activities, and generates activity performance analysis results;

[0009] Based on the activity performance analysis results, the customer behavior prediction module extracts the purchase behavior and service request data of customers at multiple time nodes, predicts the future purchase behavior and demand changes of customers, and at the same time analyzes the probability of customer churn according to the purchase potential of customers, and generates customer behavior change analysis results;

[0010] Based on the customer behavior change analysis results and activity performance analysis results, the customer relationship maintenance and optimization module matches the purchase behavior of customers with channel resource allocation according to the customer behavior changes and channel activity effect data, analyzes the critical periods for maintaining or following up services, and generates customer relationship maintenance and optimization results.

[0011] As a further solution of the present invention, the acquisition steps of extracting the behavior characteristics of each time period are specifically as follows:

[0012] Based on the customer purchase records and service request data, the daily purchase records are grouped into the integrated data set according to time periods, and the customer behavior data is cleaned according to the integrated time periods to obtain the time-series analysis input set;

[0013] Based on the time-series analysis input set, the purchase frequency, amount, and behavior response degree data of customers are grouped, the purchase behavior frequency and amount within multiple time periods are integrated, and combined with the data of customer behavior response degree, the formula is used:

[0014] ;

[0015] Compare the customer behavior frequency and consumption amount data, and calculate the average statistical value of the customer behavior frequency and response degree within the time period , to obtain the customer behavior characteristic data;

[0016] Among them, is the purchase behavior frequency data of the th customer in the rd time period, is the purchase amount of the th customer in the th time period, is the behavior response degree adjustment coefficient, is the behavior response degree of the th customer in the th time period, is the total number of days in the time period;

[0017] Based on the behavioral characteristic data of the customer, statistically analyze the behavioral changes of the customer, construct a customer behavioral change curve, use the curve to represent the change trend of the customer's behavior, and obtain the analysis result of the customer's time-series behavioral characteristics.

[0018] As a further solution of the present invention, the obtaining step of judging the behavioral changes of the customer before and after the channel activity is specifically as follows:

[0019] Based on the analysis result of the customer's time-series behavioral characteristics, integrate the channel activity data through the multi-tenant architecture of SaaS, cross-match the participation frequency of the customer with the duration and frequency of the channel activity, and obtain the channel activity data set;

[0020] Based on the channel activity data set, use the formula:

[0021] ;

[0022] Calculate the average purchase change amplitude of the customer during the channel activity , and generate the analysis result of the customer consumption change during the activity;

[0023] Wherein, is the purchase amount of the th customer in the th time period, is the purchase amount of the th customer in the previous time period, is the number of time periods during the channel activity;

[0024] Based on the analysis result of the customer consumption change during the activity, evaluate the behavioral changes of the customer before and after the channel activity, analyze the impact of the activity on the customer's behavior, and obtain the change trend information of the customer's consumption behavior.

[0025] As a further solution of the present invention, the obtaining step of comparing and analyzing multiple channel activities is specifically as follows:

[0026] According to the change trend information of the customer's consumption behavior, extract the ratio of the number of customers participating to the total number of customers, quantify the participation degree of the customer in the channel activity, and obtain the customer participation data of the activity;

[0027] According to the customer participation data of the activity, based on the historical data of other channel activities, compare the customer participation rates between multiple activities, and combine the purchase change situation to identify the key activity types, and obtain the activity performance analysis result.

[0028] As a further solution of the present invention, the obtaining step of predicting the future purchase behavior and demand changes of the customer is specifically as follows:

[0029] Based on the analysis results of the activity performance, extract the purchase behavior and service request data of the customer at multiple time nodes, classify and organize them according to time periods, analyze the purchase frequency and the number of service requests of the customer, and generate a customer behavior data set;

[0030] Based on the customer behavior data set, use the formula:

[0031] ;

[0032] Calculate the conversion probability from stage to stage of the customer, and generate a customer behavior prediction result;

[0033] wherein, represents the possibility that the customer will continue to maintain the behavior in the future stage, is the total number of customer behaviors in stage , is the total number of customer behaviors in stage , is the weight coefficient, is the customer satisfaction in stage , is the number of customer service support times in stage , is the customer participation rate.

[0034] As a further solution of the present invention, the obtaining step of analyzing the customer churn probability is specifically:

[0035] Based on the customer behavior prediction result, use the formula:

[0036] ;

[0037] Calculate the future consumption behavior potential of the customer, and generate customer purchase potential data;

[0038] wherein, is the purchase frequency of the customer, is the cumulative purchase behavior of the customer, is the total consumption ability of the customer, is the customer loyalty weight;

[0039] Extract relevant information from the feedback data and transaction records, analyze the behavior changes of the customer in multiple time periods, and combine the customer purchase potential data to evaluate the future consumption trend of the customer, and generate a customer behavior change analysis result.

[0040] As a further solution of the present invention, the obtaining step of matching the customer's purchase behavior with the channel resource allocation is specifically:

[0041] Based on the analysis results of the changes in customer behavior and the effects of channel activities, analyze the customer's consumption growth and participation behavior, and perform data processing to generate a customer behavior and channel activity dataset;

[0042] Based on the customer behavior and channel activity dataset, compare the changes in customer purchases with the allocation of channel resources, evaluate whether the existing resources meet customer needs, and analyze the rationality of resource allocation in combination with the customer's participation behavior to generate an adjusted resource configuration.

[0043] As a further solution of the present invention, the steps for obtaining the optimized result of customer relationship maintenance are specifically as follows:

[0044] According to the adjusted resource configuration, use the formula:

[0045] ;

[0046] Calculate the critical period for maintenance or follow-up services , and generate critical period data;

[0047] Wherein, is the incremental customer demand, is the allocation of channel resources, represents the interaction frequency of the customer in the activity, is the critical resource weight;

[0048] Based on the critical period data, adjust the allocation priority of channel resources, analyze the rationality of resources, and perform optimization to generate an optimized result for customer relationship maintenance.

[0049] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0050] In the present invention, by performing time period statistics on customer purchase records and service request data and extracting behavioral characteristics, the changes in customer behavior can be dynamically captured. At the same time, by cross-matching customer behavior data and channel activity performance, the behavioral changes of customers before and after the activity can be accurately judged, thereby optimizing the marketing strategy. Combining the behavioral predictions of customers at multiple time nodes can help to anticipate the future purchase needs and behavioral trends of customers in advance, and optimize the timing in customer relationship management. In addition, through the intelligent matching of customer purchase behavior and channel resources, the efficiency of customer maintenance is improved, ensuring that the enterprise provides the most suitable service or product at the best time. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is the system flow chart of the present invention;

[0052] Figure 2Flowchart of the acquisition steps for extracting the behavior characteristics of each time period in the present invention;

[0053] Figure 3 Flowchart of the acquisition steps for determining the behavior changes of customers before and after channel activities in the present invention;

[0054] Figure 4 Flowchart of the acquisition steps for comparing and analyzing multiple channel activities in the present invention;

[0055] Figure 5 Flowchart of the acquisition steps for predicting the future purchase behavior and demand changes of customers in the present invention;

[0056] Figure 6 Flowchart of the acquisition steps for analyzing the customer churn probability in the present invention;

[0057] Figure 7 Flowchart of the acquisition steps for matching the purchase behavior of customers with channel resource allocation in the present invention;

[0058] Figure 8 Flowchart of the acquisition steps for the optimization result of customer relationship maintenance in the present invention. Detailed implementation manners

[0059] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0060] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as limiting the present invention. In addition, in the description of the present invention, "a plurality of" means two or more, unless otherwise specifically defined.

[0061] Please refer to Figure 1 , a customer relationship management system under the SaaS model includes:

[0062] The customer behavior collection module integrates the customer behavior information within each time period from the cloud data source based on the customer purchase records and service request data, counts the customer purchase frequency, purchase amount, and service request times according to the time period, obtains the behavior change curve of the customer in multiple periods based on the data, extracts the behavior characteristics of each time period, and generates the analysis result of the customer time-series behavior characteristics;

[0063] The channel activity analysis module is based on the analysis results of customer temporal behavior characteristics. Through the multi-tenant architecture of SaaS, it integrates the channel activity time period and the changing trend of customer behavior, cross-matches the duration and frequency of channel activities with customer behavior data, calculates the change in customer purchases during the channel activity period, determines the changes in customer behavior before and after the channel activity, and then calculates the customer participation rate index of the activity. By comparing and analyzing multiple channel activity information, it generates activity performance analysis results;

[0064] The customer behavior prediction module extracts the customer's purchase behavior and service request data at multiple time points based on the activity performance analysis results, calculates the probability of the customer's conversion from the current stage to the next stage, predicts the customer's future purchase behavior and demand changes, and analyzes the customer churn probability based on the customer's purchase potential, generating customer behavior change analysis results;

[0065] The customer relationship maintenance optimization module is based on the analysis results of customer behavior changes and activity performance. According to customer behavior changes and channel activity effect data, it matches customer purchasing behavior with channel resource allocation, analyzes the key periods for maintenance or follow-up services, and generates customer relationship maintenance optimization results.

[0066] The results of the customer's temporal behavior characteristics analysis include customer behavior frequency distribution, purchase amount changes and service request number statistics; the activity performance analysis results include activity response rate, purchase behavior change range and activity frequency impact; the customer behavior change analysis results include purchase intention prediction value, expected change in service requests and life cycle stage transition probability; the customer relationship maintenance optimization results include maintenance trigger time point, resource investment allocation standard and customer communication frequency recommendation.

[0067] See also Figure 2 ,The specific steps for extracting the behavioral features of each time period are:

[0068] Based on customer purchase records and service request data, daily purchase records are grouped into integrated data sets by time period. Customer behavior data is cleaned based on the integrated time period to obtain a time series analysis input set.

[0069] Through the cloud data source, the enterprise's database or external cloud platform is connected, and historical data related to customer behavior is obtained through API, SQL query or data interface. The data includes customer purchase records and service request information. Then, for the timestamp in each record, the daily purchase records and service request data are classified according to the timestamp according to the predetermined time period (such as every hour, every working day, etc.). This process requires the system to accurately match the timestamp according to the setting of each time period to ensure that each record is accurately classified into the corresponding time period. In the integrated data set, abnormal high or low values ​​are screened out by removing outliers in the data. During the cleaning process, data screening conditions will be set, such as setting a reasonable range of purchase amounts and service request times. Values ​​outside these ranges will be marked or eliminated, and the cleaned data will be processed according to unified standards within the same time period. To ensure data standardization, all data types will be converted to a consistent format, such as the purchase amount will be unified into the same currency unit, the time will be unified into a standard format, and data integrity and consistency will be ensured.

[0070] Based on the time series analysis input set, the customer's purchase frequency, amount and behavior responsiveness data are aggregated, the purchase behavior frequency and amount in multiple time periods are integrated, and combined with the customer behavior responsiveness data, the formula is used:

[0071] ;

[0072] Compare the customer's behavior frequency and consumption amount data, and calculate the average statistical value of the customer's behavior frequency and responsiveness within the time period , obtain the customer's behavioral characteristic data;

[0073] in, Used to represent the comprehensive behavioral characteristics of customers in a specific period of time. It is Customers in The purchase behavior frequency data for a time period indicates the number of times a customer makes a purchase during that time period. It is obtained by counting the purchase records on a daily or hourly basis. It is Customers in The purchase amount in a time period is the consumption amount of the customer in the corresponding time period in the integrated data set, which comes from the customer's consumption record database. It is the behavioral responsiveness adjustment coefficient, which is set according to the weight of the impact of customer response behavior on the business. It is usually determined by the business characteristics of the enterprise. For example, if the response speed directly affects customer satisfaction, the coefficient can be estimated through customer service response records. It is Customers in The behavioral response degree of a time period, which represents the enthusiasm of customers for channel activities or service requests, can be obtained by quantifying the feedback speed or participation degree of customers. is the total number of days in the time period, used to calculate the average value, which can be 7 days, 30 days or other total time periods set by the enterprise.

[0074] Set the time period to 7 days, and count the behavior from the 1st day to the 3rd day, the purchase frequency , amount and response degree , and then in turn , , , and continue in the same way until the 3rd day, that is:

[0075] ;

[0076] Substitute the values to complete the calculation:

[0077] If on each day within the time period, on the 1st day, there are 5 purchases with an amount of 100 and a response degree of 0.8, on the 2nd day, there are 3 purchases with an amount of 75 and a response degree of 0.6, on the 3rd day, there are 4 purchases with an amount of 120 and a response degree of 0.7, and the response degree coefficient , then the calculation is as follows:

[0078] ;

[0079] ;

[0080] ;

[0081] The result shows that the average behavioral characteristic value of the customer within this time period is 436.67. If the historical behavioral characteristic mean value of this customer is 400, compared with the current behavioral characteristic value, it indicates that the customer's recent activities are higher than their historical behavior. Through this analysis, it can help the enterprise understand the changes in customer behavior and identify that the customer's behavior is more active within this time period.

[0082] Based on the behavioral characteristic data of customers, statistically analyze the changes in customer behavior, construct a customer behavior change curve, use the curve to represent the change trend of customer behavior, and obtain the analysis result of customer time-series behavioral characteristics;

[0083] Based on the characteristic data within each time period, sequentially integrate the behavioral data of different time periods into a curve graph to display the change trend of customer behavior over time. Further, perform curve smoothing processing by combining the characteristic values of each time period, segmentally summarize the curves of different customer behavior patterns, compare the change amplitudes of behavioral characteristics in each time period, and finally generate the analysis result of customer time-series behavioral characteristics to obtain the overall trend curve of customer behavior changes over time.

[0084] Please refer to Figure 3 , and the specific steps for obtaining the behavioral changes of customers before and after the channel activity are as follows:

[0085] Based on the analysis results of customer time-series behavior characteristics, integrate the channel activity data through the multi-tenant architecture of SaaS, cross-match the participation frequency of customers with the duration and frequency of channel activities to obtain a channel activity dataset;

[0086] According to the comprehensive behavior characteristics of customers within a specific time period , through the multi-tenant architecture of SaaS, in this architecture, different tenants share system resources by allocating isolated data tables and logical partitions, and separate and uniformly manage the channel activity data of each tenant in a distributed data management manner, integrate the channel activity time period and the change trend of customer behavior, and use the comprehensive behavior characteristics of customers within a specific time period as the analysis input; during the activity, obtain the duration and frequency data of each channel. First, summarize the channel activity duration and frequency by time period, and then the comprehensive behavior characteristics of customers within a specific time period of the dataset for cross-matching operations, that is, one-to-one correspondence between the participation frequency of customers within a specific time period and the duration and frequency of channel activities, and summarize according to the time period distribution to calculate the participation degree of customers in each time period; after the matching is completed, associate the customer participation degree result with , compare the participation frequency and behavior characteristic values in different time periods, sort the interaction frequency of customers during the activity and establish the distribution of the activity time period. According to behavior characteristics and participation frequency, analyze the differences in customer behavior in the same channel activity.

[0087] Based on the channel activity dataset, use the formula:

[0088] ;

[0089] Calculate the average purchase change amplitude of customers during the channel activity , and generate the analysis result of customer consumption change during the activity;

[0090] Among them, is used to represent the consumption growth brought by customers' participation in the activity, is the purchase amount of the th customer in the th time period, which is obtained through customer purchase record data, and the recorded amount value directly reflects the consumption behavior of customers, is the consumption amount of the th customer in the previous time period (the th segment), and the acquisition method is the same as Similarly, the difference in amount can be obtained by recording the purchase behavior of customers in adjacent time periods. It is the number of time periods during channel activity that are averaged to quantify the average level of consumption growth.

[0091] If a channel activity lasts for 3 days, the daily consumption amounts of customers are , ,and ,but:

[0092] ;

[0093] ;

[0094] The results show that the average daily consumption change of customers during the channel activity is 26.67. If the historical consumption value is 20 yuan, by comparing with the customer's historical consumption value of 20 yuan, it shows that the customer's consumption during the channel activity has increased significantly. This result reflects the customer's positive response to the activity and illustrates the role of the activity in increasing the frequency and amount of customer consumption, providing a useful quantitative basis for the analysis of the effect of channel activities.

[0095] Based on the analysis results of customer consumption changes during the event, evaluate the changes in customer behavior before and after the channel event, analyze the impact of the event on customer behavior, and obtain information on the changing trend of customer consumption behavior;

[0096] According to the calculation results Determine the changes in customer behavior before and after channel activities, collect and divide customer behavior data according to the time periods before, during, and after the activity, and collect statistics on customer purchase behavior and participation in each stage. Compare the average purchase frequency and consumption amount of customers before the activity with those after the activity. Compare and identify the incremental changes in customer purchases during the event, further summarize the regular consumption level before the event, and compare the magnitude of changes during the event Together with regular data, analyze the impact of activities on customer behavior, identify the degree of fluctuation in customer purchasing behavior, and derive changing trends in customer consumption behavior during channel activities.

[0097] See also Figure 4 , the specific steps for comparing and analyzing various channel activities are as follows:

[0098] Based on the changing trend information of customer consumption behavior, extract the ratio of the number of participating customers to the total number of customers, quantify the degree of customer participation in channel activities, and obtain the customer participation data of the activities;

[0099] To extract the ratio of the number of customer engagements to the total number of customers, the formula is used:

[0100] ;

[0101] Among them, is the customer participation rate, which represents the participation ratio of customers during a certain channel activity. The customer participation rate is usually used to measure the ability of an activity to attract customers. is the number of customers who actually participated in the activity during the channel activity. This data is obtained through customer participation records, such as customer click, purchase, or interaction behavior data, and obtained through activity monitoring. is the total number of customers who were invited or targeted during the channel activity. Usually, the number of target customer groups is extracted from the customer database, such as the number of customers to whom activity text messages or emails were sent.

[0102] If during the activity, the number of target customers for the activity is 5000 people, and the record shows that 1200 customers participated in the activity, clicked, or completed a purchase, calculate the customer participation rate. :

[0103] ;

[0104] The result shows that the customer participation rate of the channel activity is 24%, that is, 24% of the target customers participated in this activity. If the average customer participation rate of historical activities is 20%, then the participation rate of this activity is 4 percentage points higher than the historical value, indicating that this channel activity performs well in attracting customer participation.

[0105] Based on the customer participation data of the activity, according to the historical data of other channel activities, compare the customer participation rates between various activities, and combine the purchase change situation to identify key activity types and obtain the activity performance analysis results;

[0106] First, extract the historical customer participation data of each channel activity from the system, including the participation rate and purchase change amount of each activity, and classify and organize these data according to dimensions such as activity type, duration, and frequency; then, by comparing the calculated value item by item with the participation rates of other activities, analyze which activity types and settings can more effectively attract customer participation, so as to discover activities with higher participation rates and accompanied by greater purchase changes. Finally, by subdividing and analyzing the correlation between the participation rate and purchase change, identify which type of channel activity is most helpful for promoting customer consumption.

[0107] Please refer to Figure 5 for the specific steps to obtain the prediction of customers' future purchase behavior and demand changes:

[0108] Based on the analysis results of activity performance, purchase behavior and service request data of customers at multiple time nodes are extracted, classified and sorted by time period, and the purchase frequency and service request times of customers are analyzed to generate a customer behavior dataset;

[0109] First, obtain the purchase behavior and service request data of customers at different time periods through the Customer Relationship Management System (CRM), including the number of purchases, service requests and other interaction behavior data at each time node. These data are classified through the timestamp records in the CRM system. Then, integrate and analyze these data with the data of channel activity effects, and analyze specific behavior performances such as the frequency, consumption amount, and service request response time of customer behavior. By analyzing the behavior characteristics of customers in different time periods, the changing trends of customer behavior can be identified. After the data are summarized by time period, the customer behaviors at each node are correlated to track the purchase tendencies and demand changes of customers.

[0110] Based on the customer behavior dataset, use the formula:

[0111] ;

[0112] Calculate the conversion probability of customers from stage to stage to generate the customer behavior prediction result; where,

[0113] represents the possibility that customers will continue to maintain their behavior in the future stage, is the total number of customer behaviors in stage including the number of purchases and service requests. These data are obtained through the CRM system, is the total number of customer behaviors in stage reflecting the number of behaviors of customers in the next stage, which is obtained from the CRM system, is the weight coefficient, reflecting the importance of customer behavior. The weight is obtained through historical data analysis, and higher weights can be assigned to different behavior types (such as high-frequency purchase behaviors), is the customer satisfaction in stage collected through customer feedback surveys or scoring information, used to reflect the experience quality of customers in this stage, is the number of customer service support times in stage and the number of services enjoyed by customers in this stage is counted through the service records of the CRM system.

[0114] If the total number of purchase behaviors and service requests of customers in stage (within 1 month) is 800 times, and the total number of behaviors in stage (the next month) is 600 times, and the weight coefficient (not translated as it seems to be incomplete in the original) , the customer satisfaction score for stage is 80 points, and the number of service support times for stage is 50 times, and the customer participation rate is 0.24. The calculation is as follows:

[0115] ;

[0116] The result indicates that the conversion probability of customers from stage to stage is 2.74. A higher conversion probability indicates that the possibility of customers continuing their behavior in the next stage is very high. If the customer participation rate in the previous stage is 24% ( ), it indicates that one-fourth of the target customers actively participated in the marketing activities, which has a positive effect on the continuity of customer behavior, thus predicting a higher conversion probability in the subsequent stage. Therefore, a higher combined with a higher value indicates that customers not only have positive behavior performance in the current stage but also have a very high possibility of participating in the future stage. This combination method provides a more comprehensive perspective for customer behavior prediction and can help enterprises formulate more accurate future marketing and customer maintenance strategies.

[0117] Please refer to Figure 6 for the specific steps to obtain the customer churn probability:

[0118] Based on the customer behavior prediction results, use the formula:

[0119] ;

[0120] Calculate the potential of customers' future consumption behavior to generate customer purchase potential data;

[0121] Among them, is the purchase frequency of customers, indicating the average number of purchases by customers within a specific time period. This data is obtained through transaction records in the customer relationship management system (CRM). By analyzing the number of purchases by customers in the CRM system within a specific past period, the average purchase frequency is statistically calculated. is the cumulative purchase behavior of customers, indicating the total purchase volume by customers within a specific time period. The data is obtained through sales records in the CRM system. By accumulating the purchase amounts of customers, the total consumption behavior within this period is obtained. is the total consumption ability of customers, indicating the disposable consumption ability of customers, usually estimated from the financial data or income situation of customers. This data can be estimated through the financial reports, income surveys, or credit information systems of customers, reflecting the overall financial situation of customers. It is the customer loyalty weight, which represents the degree of customer loyalty to the brand. It is usually determined by analyzing factors such as customers' repeat purchase behavior, satisfaction scores, and brand interaction years. It can be extracted from the customer's long-term purchase history and satisfaction feedback system and given an appropriate weight value to reflect the customer's loyalty.

[0122] If the customer's purchase frequency times, the cumulative purchase volume , the total customer consumption ability , the customer loyalty weight , then the customer purchase potential is calculated as follows:

[0123] ;

[0124] The result shows that the customer's purchase potential is 5.2. If the historical average purchase potential value is 4.5, then the result of 5.2 for this customer indicates that their consumption potential is significantly higher than the average level, which makes this customer worthy of special attention by the enterprise. It means that the customer's future consumption behavior is very likely to continue and may even increase. Combining with the high value can not only identify which customers theoretically have high purchase potential but also evaluate the possibility of realizing these potentials in actual situations. This helps enterprises make more data-supported decisions when formulating promotional activities, pricing strategies, and customer loyalty programs.

[0125] Extract relevant information from the feedback data and transaction records, analyze the behavioral changes of customers over multiple time periods, combine with the customer purchase potential data, evaluate the future consumption trends of customers, and generate the analysis results of customer behavioral changes;

[0126] First, extract the customer's historical purchase data from the customer relationship management system, analyze the customer's purchase frequency and cumulative consumption amount in the past year, classify and organize these data by month, then combine with the customer's satisfaction feedback and service request records, analyze the trend of customer satisfaction changes, determine the customer's consumption behavior in different time periods, further use these data to evaluate the customer's future consumption trends, and combine with the calculated customer purchase potential to judge whether the customer's future consumption behavior will continue and the trend of change in consumption amount. By analyzing that the customer's purchase potential is higher than the historical value and combining with the customer's satisfaction data, it can be identified that this customer has a lower risk of churn and it can be predicted that they will continue to maintain a high consumption frequency in the future.

[0127] Please refer to Figure 7 for the specific steps of matching the customer's purchase behavior with the channel resource allocation:

[0128] Based on the analysis results of customer behavior changes and channel activity effects, analyze the customer's consumption growth and participation behavior, and perform data processing to generate a customer behavior and channel activity dataset;

[0129] Combine the known purchase changes and participation rates, quantify the consumption growth caused by the activity by comparing the purchase data before and after the activity, and at the same time evaluate the response degree of different customer groups to the activity. Further data integration will include the combination of purchase change data and customer participation data with other relevant behavior data (such as the number of service requests, customer feedback, etc.), use statistical and data mining techniques to identify the patterns and trends of consumption and participation behavior, and finally integrate the customer's behavior changes with the channel effect results.

[0130] Based on the customer behavior and channel activity dataset, compare the customer's purchase changes with the allocation of channel resources, evaluate whether the existing resources meet the customer's needs, combine the customer's participation behavior, analyze the rationality of resource allocation, and generate an adjusted resource configuration;

[0131] By analyzing the amplitude of the customer's purchase change and match it with the existing channel resource allocation, evaluate whether the resources of different channels reasonably meet the actual needs of customers, and then combine the customer's participation rate to analyze the usage of customers in each channel, compare and data to determine whether the resources need to be reallocated. For example, some channels with high participation rates but insufficient resources may lead to poor customer experience and unexploited purchase potential, indicating uneven resource allocation. Ensure that customers can obtain sufficient channel support during critical periods, complete the precise matching of customer purchase behavior and channel resources, so as to improve resource utilization and customer satisfaction.

[0132] Please refer to Figure 8 The specific steps for obtaining the optimization results of customer relationship maintenance are as follows:

[0133] According to the adjusted resource configuration, use the formula:

[0134] ;

[0135] Calculate the critical period of maintenance or follow-up services to generate critical period data;

[0136] Among them, represents the best time point for the enterprise to carry out customer relationship maintenance or follow-up, is the incremental customer demand. By analyzing the transaction records or service request logs of the customer relationship management system, extract the demand change situation of customers during a specific period, It is the channel resource allocation, indicating the resource situation allocated to different channels, which can be obtained through the channel management system. It represents the interaction frequency of customers in the activity. Data such as the activity participation frequency is extracted through the CRM system. It is the key resource weight, indicating the importance of specific resources for maintaining customer relationships, which is obtained based on the analysis of historical usage effects.

[0137] If the customer demand increment , the channel resource allocation , the customer participation , the key resource weight , then the calculation is as follows:

[0138] ;

[0139] The calculation results show that this period is the best time for maintaining customer relationships or follow-up services. By comparing with historical data (if the benchmark value is 1.5), it can be seen that the customer demand increment in the current period is relatively large and the resource allocation is reasonable. Therefore, it is necessary to prioritize the arrangement of customer relationship maintenance services.

[0140] Based on the data of key periods, adjust the allocation priority of channel resources, analyze the rationality of resources, and optimize to generate the optimization results of customer relationship maintenance;

[0141] First, extract the behavior changes of customers and channel resource data, including the previous customer demand increment , the participation rate and the resource allocation situation . Then, integrate these data into the customer relationship management system. By analyzing the customer demand and channel resource allocation in different time periods, compare whether the resources can meet the actual needs of customers, and adjust the priority of resources according to the key resource weight to optimize the resource allocation strategy. The final optimization information of customer relationship maintenance can ensure that the enterprise can reasonably allocate resources during key periods to improve customer satisfaction and the effect of long-term relationship maintenance.

[0142] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A customer relationship management system under the SaaS model, characterized in that: The system comprises: The customer behavior collection module collects statistics on the customer's purchase frequency, purchase amount, and service request times according to time periods based on the customer's purchase records and service request data, extracts the behavior characteristics of each time period, and generates customer time series behavior characteristics analysis results; Based on the analysis results of the customer's temporal behavior characteristics, the channel activity analysis module cross-matches the duration and frequency of channel activities with customer behavior data through the multi-tenant architecture of SaaS, determines the changes in customer behavior before and after the channel activities, and generates activity performance analysis results; The steps for determining the changes in customer behavior before and after the channel activity are specifically as follows: Based on the analysis results of the customer's temporal behavior characteristics, the channel activity data is integrated through the multi-tenant architecture of SaaS, and the customer's participation frequency is cross-matched with the duration and frequency of channel activities to obtain a channel activity data set; Based on the channel activity data set, the formula is used: ； Calculate the average change in purchases made by a customer during their channel activity , generate analysis results of customer consumption changes during the activity; in, It is Customers in The purchase amount in a certain period of time, It is The purchase amount of each customer in the previous period, is the number of time periods during which the channel is active; Based on the analysis results of customer consumption changes during the event, evaluate the changes in customer behavior before and after the channel event, analyze the impact of the event on customer behavior, and obtain information on the changing trend of customer consumption behavior; The specific steps for comparing and analyzing multiple channel activities are as follows: According to the change trend information of the customer consumption behavior, extract the ratio of the number of customer participation to the total number of customers, quantify the customer participation in the channel activity, and obtain the customer participation data of the activity; Based on the customer participation data of the activities, and the historical data of other channel activities, compare the customer participation rates between various activities, and identify the key activity types in combination with the purchase changes to obtain the activity performance analysis results; The customer behavior prediction module extracts the customer's purchase behavior and service request data at multiple time points based on the activity performance analysis results, predicts the customer's future purchase behavior and demand changes, and analyzes the customer churn probability based on the customer's purchase potential, generating customer behavior change analysis results; The customer relationship maintenance optimization module is based on the customer behavior change analysis results and activity performance analysis results. According to the customer behavior change and channel activity effect data, it matches the customer's purchasing behavior with the channel resource allocation, analyzes the key time periods for maintenance or follow-up services, and generates customer relationship maintenance optimization results.

2. The customer relationship management system under the SaaS model according to claim 1, characterized in that: The steps of extracting the behavior features of each time period are specifically as follows: Based on customer purchase records and service request data, daily purchase records are grouped into integrated data sets by time period. Customer behavior data is cleaned based on the integrated time period to obtain a time series analysis input set. Based on the time series analysis input set, the customer's purchase frequency, amount and behavior responsiveness data are aggregated, the purchase behavior frequency and amount in multiple time periods are integrated, and combined with the customer behavior responsiveness data, the formula is used: ； Compare the customer's behavior frequency and consumption amount data, and calculate the average statistical value of the customer's behavior frequency and responsiveness within the time period , obtain the customer's behavioral characteristic data; in, It is Customers in Purchase behavior frequency data for a period of time, It is Customers in The purchase amount in a certain period of time, is the behavioral responsiveness adjustment coefficient, It is Customers in Behavioral responsiveness in a time period, is the total number of days in the time period; Based on the customer's behavior characteristic data, the customer's behavior changes are counted, a customer behavior change curve is constructed, the curve is used to represent the changing trend of the customer's behavior, and the customer's time series behavior characteristic analysis results are obtained.

3. The customer relationship management system under the SaaS model according to claim 1 is characterized in that: The steps for obtaining the predicted customer's future purchasing behavior and demand changes are specifically as follows: Based on the activity performance analysis results, extract the customer's purchase behavior and service request data at multiple time points, classify and organize them according to time periods, analyze the customer's purchase frequency and service request times, and generate a customer behavior data set; Based on the customer behavior data set, the formula is adopted: ； Calculate the customer from the stage To stage The conversion probability , generate customer behavior prediction results; in, Indicates the possibility that the customer will continue to behave in the future. It is a stage Total number of customer actions, It is a stage Total number of customer actions, is the weight coefficient, It is a stage Customer satisfaction, It is a stage Customer service calls is the customer engagement rate.

4. The customer relationship management system under the SaaS model according to claim 3 is characterized in that: The steps for obtaining the probability of analyzing customer churn are specifically as follows: Based on the customer behavior prediction results, the formula is adopted: ； Calculate the customer's future consumption behavior potential , generate customer purchase potential data; in, is the customer's purchase frequency, It is the customer’s cumulative purchase behavior. is the total spending power of the customer, is the customer loyalty weight; Extract relevant information from feedback data and transaction records, analyze customer behavior changes over multiple time periods, combine the customer purchase potential data, evaluate the customer's future consumption trends, and generate customer behavior change analysis results.

5. The customer relationship management system under the SaaS model according to claim 4 is characterized in that: The steps for obtaining the matching of the customer's purchasing behavior and the channel resource allocation are specifically as follows: Based on the customer behavior change analysis results and channel activity effect analysis results, analyze the customer's consumption growth and participation behavior, and perform data processing to generate customer behavior and channel activity data sets; Based on the customer behavior and channel activity data set, compare the customer's purchase changes with the allocation of channel resources, evaluate whether the existing resources meet customer needs, analyze the rationality of resource allocation in combination with the customer's participation behavior, and generate adjusted resource configuration.

6. The customer relationship management system under the SaaS model according to claim 5, characterized in that: The steps for obtaining the customer relationship maintenance optimization result are specifically as follows: According to the adjusted resource allocation, the formula is adopted: ； Calculate critical periods for maintenance or follow-up services , generate key period data; in, is the increase in customer demand, It is the allocation of channel resources. Indicates the frequency of customer interaction in the activity, is the key resource weight; Based on the key period data, the allocation priority of channel resources is adjusted, the rationality of resources is analyzed, and optimization is performed to generate customer relationship maintenance optimization results.

Images