CN117726357A - Electronic commerce marketing method based on SCRM - Google Patents

Abstract

An electronic commerce marketing method based on SCRM realizes the deep portrait analysis of clients and staff visitors by establishing a comprehensive platform database and recording visitor activity data. And a carefully designed matching algorithm is adopted to divide the visitors into different groups, so that a foundation is provided for personalized service and accurate marketing. By defining the interaction weight, calculating the interaction score and intelligent resource allocation, intelligent management and personalized recommendation of user interaction are realized, and the purposes of improving customer satisfaction, promoting user loyalty and promoting continuous growth of e-commerce sales performance are achieved. In the whole, the invention skillfully combines SCRM and E-business marketing, and provides a comprehensive and efficient marketing strategy system for enterprises.

Description

Electronic commerce marketing method based on SCRM

Technical Field

The invention relates to a marketing scheme, in particular to an electronic commerce marketing method based on SCRM.

Background

The CRM system of the traditional electronic commerce business mainly focuses on order data, and relates to the aspects of transaction amount, transaction evaluation, repurchase rate, return rate and the like, however, the CRM system does not fully cover other activity data of customers on a platform. The current CRM system has limited data sources and cannot deeply record and analyze various activities of clients on the platform, such as commodity browsing, searching records, interaction behaviors and the like. This limits the enterprise's full understanding of customer behavior, resulting in limitations in personalized services and accurate marketing. Therefore, there is a need to increase customer portrait depth and accuracy to more effectively meet customer needs, increase customer satisfaction, and promote sales growth.

Disclosure of Invention

In order to overcome the above drawbacks of the prior art, the object of the present invention is:

an electronic commerce marketing method based on SCRM is established. Unlike conventional CRM systems, which focus only on order data, the present invention creates a comprehensive platform database and records various activity data of visitors on the platform, including browsing merchandise, searching records, purchase history, interaction records, etc. Through the cleaning of data, user portrait analysis and the design of a matching algorithm, the invention aims to realize more comprehensive and deep user understanding, divide visitors into different groups and provide a more accurate basis for subsequent personalized service and accurate marketing. By defining interaction weights, calculating interaction scores and intelligent resource allocation, the invention aims to optimize user interaction management, improve customer satisfaction, promote user loyalty and finally promote continuous increase of e-commerce sales performance. By integrating SCRM and E-commerce marketing, the invention aims to overcome the defects of the prior art and provide a set of more comprehensive and efficient marketing strategy system.

The technical scheme adopted for solving the technical problems is as follows:

an electronic commerce marketing method based on SCRM specifically comprises the following steps:

1. establishing a platform database and recording the activity data of the visitor

The platform establishes a comprehensive database for storing various activity data of the visitor on the platform, including commodity browsing, searching record, purchasing history, interaction record, etc. This provides a rich data base for subsequent personalized services and accurate analysis.

2. Cleaning the data to finish user duplication removal

And the accuracy and consistency of the data are ensured by cleaning and processing the collected data. The deduplication operation eliminates multiple records for the same user, enabling each user in the database to be uniquely identified, providing a clean dataset for subsequent analysis.

3. User profile analysis of guests:

the guests are divided into client guests and employee guests, and user portrait analysis with different levels is performed:

(1) customer portrait analysis:

through deep analysis of the activity data of the clients and the visitors, client portraits are established, including browsing tracks, purchasing behavior, hobbies and interests, participation topics, interaction records and the like. The establishment of the customer portrait aims at deeply understanding the requirements, preferences and behaviors of customers and provides a basis for establishing personalized services and accurate marketing;

(2) staff portrait analysis:

user portrait analysis is carried out on staff and visitors, staff portraits are established, and the staff portraits comprise subdivision fields, content creation, interaction records, attention numbers and the like. By analyzing the characteristics of staff and visitors, the platform can better know the influence, the professional field and the interaction level of the staff on the platform, and provides a basis for the subsequent establishment of a secondary marketing scheme;

4. matching according to a matching algorithm

By designing a matching algorithm, the platform divides guests into different groups to better understand their characteristics and behavior. The step helps the platform to locate the user type more accurately, and lays a foundation for personalized service and accurate marketing.

5. Formulating personalized services

In the key step, a platform adopts a layering strategy to formulate a first-level marketing scheme according to the user portrait of the client, and then a second-level marketing scheme is formulated according to the user portrait of the staff visitor matched by the client:

(1) first-level marketing scheme formulated according to user portrayal of customer

The platform establishes a primary marketing scheme by comprehensively analyzing the information such as interests, purchasing behavior, geographic positions and the like of clients. This may include providing personalized product recommendations, periodic proprietary offers, custom services, etc. to the customer to improve customer satisfaction and purchase conversion.

(2) Making a secondary marketing scheme according to the user portrait of staff visitor matched by the customer

Based on the user portraits of employee visitors to which the customers are matched, the platform formulates a finer secondary marketing scheme. This step may include establishing a more intimate relationship with employee guests, providing personalized services, introducing more customers through recommendations of employee guests, etc., to further increase customer loyalty and promote more cross-sales.

As a further improvement of the invention: the method for establishing the database is as follows: the method of creating a database includes performing demand analysis, selecting an appropriate database management system (DBMS), designing a database structure, creating a physical database, selecting an appropriate data type, implementing data security measures, importing initial data, formulating periodic backup and restore policies, optimizing database performance, documenting database structure and operating procedures, and continuously monitoring and maintaining database performance and stability. The process ensures that the database can meet the actual requirements, ensures the data security, improves the performance, and is maintained and monitored in time.

As a further improvement of the invention: the method for recording the activity data of the visitor is as follows: the method for recording the activity data of the visitor comprises the steps of implementing page tracking, and recording page browsing behaviors and clicking events of the visitor through browsing data; recording search behaviors including search keywords and search result clicking conditions; recording purchase history, including order information and transaction details; recording interaction behaviors such as comments, scores, participation topics and the like; performing user identification to correlate activity data to a particular visitor; introducing a real-time data updating technology to ensure timeliness of the data; and emphasizes privacy protection, follows relevant regulations, and protects the privacy rights of users. The methods provide a comprehensive visitor activity data recording system and provide a rich data basis for subsequent personalized services and accurate analysis.

As a further improvement of the invention: the method for cleaning and washing the data is as follows: data cleaning and deduplication are key links of data processing, and aim to ensure data quality and consistency. The process comprises the steps of processing missing values and abnormal values, unifying data formats, adopting a deduplication operation to wash by using a database tool and a script based on key fields and time windows, verifying logical relations and uniqueness, and finally regularly maintaining a database to keep information clean and accurate, thereby providing a reliable basis for subsequent data analysis and application.

As a further improvement of the invention: the method for carrying out user portrait on the visitor comprises the following steps: the method for user portraying the visitor includes creating comprehensive platform data base and recording various kinds of activity data, such as browsing, searching, purchasing and interactive record, of the visitor. And the data is cleaned and de-duplicated, so that the accuracy of the data is ensured. The visitors are divided into customer visitors and employee visitors according to activities, user portrait analysis of different levels is carried out, browsing tracks, purchasing behaviors, interests and hobbies of the customers are known in depth, and analysis is carried out aiming at characteristics of subdivision fields, content creation and the like of the employee visitors. And carrying out group division through a matching algorithm, formulating a primary marketing scheme to provide personalized service for clients, and further formulating a secondary marketing scheme based on staff visitor so as to promote the loyalty of clients and cross sales. This process helps the platform to more accurately understand the user's needs, providing personalized services and accurate marketing strategies.

As a further improvement of the invention: the design approach to matching algorithms covers a number of key steps. First, the matching objective is explicitly defined, the application purpose of the algorithm is explicitly defined. The applicable matching features are then selected, taking into account data features including user behavior, interest tags, etc. Data preprocessing is performed to ensure the quality and consistency of the data. The matching algorithm is selected according to specific requirements, and can be a rule-based method, a similarity-based method or a machine learning algorithm. And (3) formulating a matching rule or establishing a model, and improving the accuracy of the algorithm through training and tuning. In the application stage, the real-time data is matched, so that the real-time performance of the algorithm is ensured. Finally, the performance of the algorithm is assessed and optimized, privacy and security factors are considered, and a real-time update mechanism is implemented to maintain the effectiveness of the algorithm. The integrated design method is favorable for developing a matching algorithm with higher accuracy and practicability, and the personalized service and recommendation effect of the system are improved.

As a further improvement of the invention: the matching algorithm also comprises an SCRM interaction weight resource allocation scheme, and various interaction types (such as praise and comment) are definitely defined and are given different weights in an interaction weight definition stage. The weights are used for evaluating the influence of interaction in real time and are adjusted at any time so as to more accurately reflect the actual influence of the interaction of the user. In the interaction score calculation, a linear regression model is adopted to calculate the interaction score of the user, and the frequency and the quality of interaction are considered, so that the liveness and the influence of the user on a social platform are comprehensively evaluated. The weight threshold setting stage adopts setting based on a statistical method, meanwhile, the diversity of user groups is considered, and the threshold is adjusted to better adapt to the characteristics of different user groups. Finally, in the resource allocation logic stage, intelligent resource allocation decisions are carried out by means of a decision tree algorithm according to the user interaction scores and the set threshold value, and personalized recommendation and service are provided for the users.

As a further improvement of the invention: the method for making personalized marketing service scheme is as follows: the users are subdivided into different groups, and customized product or service suggestions are provided for the users by utilizing a personalized recommendation engine. And formulating personalized promotion, providing special offers according to the purchase history and the attention field of the user, and stimulating the purchase desire. The personalized service is ensured to keep consistent in each marketing channel, and is continuously communicated with the user through a real-time interaction and feedback mechanism. The A/B test is used for effect comparison and optimization, and meanwhile, the privacy of users is respected in scheme design, and related regulations and compliance requirements are met. Finally, by monitoring and measuring the effect of the scheme, the user satisfaction is improved, the user loyalty is promoted, and the sales growth is realized.

Compared with the prior art, the invention has the beneficial effects that:

by establishing a comprehensive platform database and recording visitor activity data, deep portrait analysis of clients and staff visitors is realized. And a carefully designed matching algorithm is adopted to divide the visitors into different groups, so that a foundation is provided for personalized service and accurate marketing. By defining the interaction weight, calculating the interaction score and intelligent resource allocation, intelligent management and personalized recommendation of user interaction are realized, and the purposes of improving customer satisfaction, promoting user loyalty and promoting continuous growth of e-commerce sales performance are achieved. In the whole, the invention skillfully combines SCRM and E-business marketing, and provides a comprehensive and efficient marketing strategy system for enterprises.

Drawings

FIG. 1 is a schematic flow chart of the present invention.

Detailed Description

The invention will now be further described with reference to the accompanying drawings and examples:

detailed description of the invention:

1. establishing a database:

before the database is built, first, a requirement analysis is performed to define the target, scope and required functions of the database. And determining key information such as data types, data amounts, access frequencies and the like which need to be stored so as to ensure that the database design meets the actual requirements. Selecting an appropriate database management system, such as MySQL, postgreSQL, microsoft, according to the needs and system requirements; SQL; server, etc. The selected DBMS should be matched to the technical architecture and performance requirements of the platform. The table structure of the database is designed to define the relationships and fields between the tables. Visualization of database structures using database design tools (e.g., ER graphs) ensures that the storage and management of data is organized and efficient. Creating a physical database in the selected DBMS, and building corresponding tables, fields, indexes and the like according to the designed database structure. The standardization and the integrity of the database are guaranteed, the best practice is followed, and the performance and the maintainability are improved. And selecting proper data types to store different kinds of information, so as to ensure the accuracy and the effectiveness of the data. For example, standard data types such as integer, floating point number, date and time, etc. are used. Data security measures, including but not limited to access control, authentication, encryption, etc., are introduced to ensure that information in the database is protected, conforming to privacy and security standards. The existing initial data is imported into the database, ensuring that the database contains the necessary initial information. The data import operation may be performed using a database management tool or a scripting language. Regular backup and restore policies are formulated to prevent data loss. And ensuring that the backed-up data can be restored in time so as to maintain the integrity and availability of the database. Database performance optimization, including index optimization, query optimization, normalization, etc., is performed periodically to ensure efficient operation of the database. To facilitate team members' understanding and manipulating databases, the database structures, field descriptions, and operating procedures are documented, providing clear documentation and guidance. The performance and stability of the database are continuously monitored, and the problems are timely identified and solved. And necessary maintenance work is carried out, so that the database can meet the changing requirements.

The above method provides a general step framework for establishing a database, and the specific operation can be adjusted according to the specific requirements and technical requirements of the project.

2. Recording activity data:

when a visitor browses a page on a platform, recording browsing behaviors of the visitor, wherein the browsing behaviors comprise information such as the accessed page, residence time, browsing sequence and the like; click events of the visitor on the page, including clicked buttons, links, commodities, etc. are recorded to learn interests and preferences of the visitor. And recording the search behavior of the visitor on the platform, wherein the search behavior comprises search keywords, search result clicking conditions, and search time and frequency. Recording the purchase history of the visitor, including purchased goods, purchase time, order amount and the like; transaction details including payment means, delivery information, etc. are recorded. The method comprises the steps of carrying out a first treatment on the surface of the Recording comments and scores of visitors on goods or services so as to know satisfaction degree and feedback of the visitors on products; if the platform has social functions or topic discussion areas, the participation of the visitor is recorded, including published contents, replies, praise and the like. In order to associate the activity data to a specific guest, a user identification method needs to be implemented, for example using user login information, cookies, device identification, etc., to ensure personalization of the data. By adopting a real-time data updating technology, the activity data of the visitor can be recorded and stored in real time so as to respond to the demands and changes of the user in time. When recording the activity data, the user is ensured to meet the related privacy regulations, the collection and use purpose of the data is explicitly informed, the privacy rights of the user are respected, and privacy setting options are provided.

The combination of the above methods can build a comprehensive visitor activity data recording system, and provide a rich data basis for subsequent personalized services and accurate analysis.

Further, the data acquisition process may include: the social media platform and the types of user interaction data to be collected, such as posting, praise, comment and sharing; selecting a proper Web crawler tool to realize effective data crawling; deep knowledge of social media page structure, identification of data storage locations and element tags for targeted crawling; user interaction behaviors needing to be recorded are definitely defined, including posting, praise, comment, sharing and the like; extracting relevant fields of user interaction data, such as user ID, time stamp, content, praise number, comment content and the like, through a crawler technology; the collected data is stored in a database or a file, so that the structuring and the inquiry are ensured.

Such a data collection flow can help build a rich user interaction data set, providing an important information basis for subsequent analysis, user portrayal, and personalized services. During data acquisition, relevant regulations and platform policies are required to be followed, and validity and privacy protection are ensured.

3. User portrayal:

establishing a comprehensive database, and recording various activity data of the visitor on a platform, including commodity browsing, searching records, purchase history, interaction records and the like; through cleaning and deduplication operations, accuracy and consistency of data are ensured, and a plurality of records of the same visitor are eliminated, so that each visitor in the database can be uniquely identified. The guests are divided into client guests and employee guests, and user portrayal analysis of different levels is performed, including client portrayal analysis and employee portrayal analysis. Through further analysis of the activity data of the client visitor, a client portrait is established, including browsing tracks, purchasing behaviors, hobbies and interests, participation topics, interaction records and the like, and a foundation is provided for establishing personalized services and accurate marketing. User portrait analysis is carried out on staff and visitors, staff portraits are established, including subdivision fields, content creation, interaction records, attention numbers and the like, and basis is provided for making a secondary marketing scheme. Matching algorithms are designed to divide guests into different groups to better understand their characteristics and behavior. The platform is helpful for positioning the user types more accurately, and lays a foundation for personalized service and accurate marketing. And a hierarchical strategy is adopted to formulate a primary marketing scheme according to the user portraits of the clients, wherein the primary marketing scheme comprises the steps of providing personalized product recommendation, regular exclusive preference, customized service and the like for the clients, so that the satisfaction degree and the purchase conversion rate of the clients are improved. Based on the user portraits of employee and visitor matched by the customers, a finer secondary marketing scheme is formulated, including establishing more intimate relationship with employee and visitor, providing personalized services, introducing more customers through the recommendation of employee and visitor, etc., to further improve customer loyalty and promote more cross-sales. By the method, the platform can deeply understand the characteristics and behaviors of the visitor, so that more personalized and accurate service and marketing schemes are provided.

Further, when user portrayal is performed, the process of performing behavior pattern analysis is as follows: user behavior to be analyzed, such as posting, praise, comment, etc., is explicitly defined; collecting time sequence data of user behaviors, including time stamps, behavior types, related contents and the like; observing the distribution condition of user behaviors in time by drawing a time sequence chart, and finding out the periodicity and trend possibly existing; calculating key statistical indexes such as mean and variance to better understand the overall characteristics of the user behavior; fourier transforming the time series data to analyze whether significant periodic components are present; identifying possible periodic features, determining whether the behavior exhibits a repeating pattern over a particular period of time; smoothing the time sequence data by using a moving average method to help identify trends; determining whether a user behavior has a significant rising or falling trend by using a trend line fitting technology; a clustering algorithm is applied to divide users into different groups according to behavior patterns; classifying the user behaviors by using a classification algorithm, and identifying different user behavior types; establishing a time sequence prediction model, and attempting to predict the trend and periodicity of future user behaviors; optimizing personalized recommendation and service strategies according to the analysis result of the behavior mode so as to better meet the requirements of users; collecting user feedback, and verifying the accuracy and the effectiveness of behavior pattern analysis; and according to the feedback result, the user portrait and the personalized service strategy are adjusted, so that more accurate user experience is realized.

Through the steps, the behavior mode of the user can be comprehensively known, including periodicity, trend and specific mode identification, and powerful support is provided for formulating more targeted personalized service and marketing strategies.

4. Matching algorithm

The features selected for matching may be visitor behavior data, interest tags, geographic locations, etc. The selection of features should be consistent with matching objectives and business requirements. The data is pre-processed, including washing, normalization, etc., prior to application of the matching algorithm to ensure quality and consistency of the data. And selecting a proper matching algorithm according to the difference of the matching target and the data characteristics. Common matching algorithms include rule-based matching, similarity-based matching, machine learning algorithms, and the like. Defining a matching rule if rule-based matching is selected; if a machine learning algorithm is selected, a training model is built. Ensuring that the rules or models adequately capture the relationships between features. For the machine learning algorithm, training and tuning are performed, the model is trained by using historical data, and tuning is performed according to performance so as to improve matching accuracy. And applying the designed matching algorithm to the actual data to perform matching operation. This may involve batch processing of large-scale data sets or streaming of real-time data. And evaluating the matching algorithm, verifying by using the test data set, and checking the matching accuracy and effect. And optimizing according to the evaluation result, and continuously improving the performance of the algorithm. If the matching algorithm needs to be applied in real time, the algorithm is ensured to update new data in real time so as to keep the accuracy and timeliness of matching. Privacy and safety problems are considered in the design of the matching algorithm, proper measures are taken to protect user data, and the matching process is ensured to accord with relevant regulations and privacy standards. Through the steps, the designed matching algorithm can better meet specific business requirements, and personalized service and accurate recommendation effects of the system are improved.

Further, the matching algorithm further comprises an interaction weight matching method, different interaction types (such as praise and comment) are classified and respectively given with weights in an interaction weight definition stage, and the actual influence of the user interaction is reflected more accurately by evaluating the influence of the interaction and adjusting the weight value in real time. In the process of calculating the interaction score, the linear regression model is utilized to calculate the interaction score of the user, and the interaction frequency and the interaction quality are comprehensively considered so as to accurately evaluate the activity degree and the influence of the user on the social platform. The weight threshold setting stage sets the weight threshold based on a statistical method, and simultaneously considers the diversity of the user groups, and adjusts the threshold to be more fit with the characteristics of different user groups. Finally, in the resource allocation logic stage, intelligent resource allocation decisions are carried out according to the user interaction scores and the set threshold value through a decision tree algorithm, and personalized recommendation and service are provided for the user so as to optimize the user experience. The whole flow covers the whole steps from the definition of the interaction weight to the resource allocation logic, and aims to realize more refined user interaction management and personalized service.

Further, the interactive weight matching method comprises weight items and weight proportions, wherein the weight items comprise the following contents: recording content browsed by a user on a platform, including checking commodity details, reading articles, watching videos and the like; the user performs praise, like or express good behaviors on the content; the user can issue comment interactions under the articles, pictures or commodities; the user shares the content on the platform to other social media or sends the content to other users; recording purchase activities completed by a user on a platform, including purchased goods, transaction amounts and the like; the user can track the information such as keywords, search frequency and the like through the search operation performed on the platform; users focus on or by other users, and interactions with fans; users participate in topic discussions on the platform, participate in groups or community activities; the user participates in online or offline activities of the platform organization, such as voting, surveys, sweepstakes, and the like.

Further, the weight ratio of the above items is:

purchasing behavior: 0.25

Purchasing behavior is generally regarded as a key index for user decision, and has a great influence on the economic value of the platform.

Browsing behavior: 0.15

The browsing behavior reflects the interests and the focus of attention of the user and has a certain influence on personalized recommendation.

Praise and like: 0.1

Praise and like show the user's preference for specific content, with some effect on recommendations and social impact.

Comment behavior: 0.1

The comment behavior shows the deep participation and feedback of the user on the content, and is helpful for knowing the opinion and emotion of the user.

Sharing behavior: 0.1

The sharing behavior can expand the spread of content and make a contribution to social impact.

Search behavior: 0.1

The search behavior of the user reflects more specific interests and demands and plays a role in personalized recommendation.

Attention and fan interaction: 0.05

The focus and fan interactions reflect the social relationship of the user, which has an impact on the establishment of the social network.

Participation in topics and discussions: 0.05

The participation topics and discussions indicate the activity degree of users in communities, and have a certain influence on the activity degree of the social platform.

Activity participation: 0.05

The activity participation displays the interest of the user in the platform for holding the activity, and has a certain effect on the user participation degree.

Account settings and personal information updates: 0.05

The account settings and the updating of personal information can reflect the personalized needs and changes of the user, and have a certain effect on the customized service.

This is just one possible weight distribution example, and in practical applications, the adjustment needs to be performed according to specific service situations and platform targets.

5. Personalized marketing services

User-related data including browsing history, purchase records, search behavior, favorites labels, etc. are collected to create a comprehensive and accurate representation of the user. And cleaning the collected user data to ensure the accuracy and the integrity of the data. Deep analysis is performed by the data analysis tool to identify user behavior patterns, preferences, and trends. Based on the analysis results, the users are divided into different sub-divided groups to better understand their characteristics and needs. This may be achieved by means of cluster analysis, behavioral analysis, etc. By means of the personalized recommendation engine, customized products, services or contents are recommended to the user according to the historical behaviors and preferences of the user, and user experience and purchase conversion rate are improved. Personalized promotion activities are formulated, and special offers, customized services or activity participation qualification are provided for the users according to the purchase history, the attention field and other information of the users so as to excite the purchase desire. The consistency of personalized services in different marketing channels is ensured, and similar personalized experience can be provided whether an online platform, social media or physical shops are realized. And a real-time interaction mechanism is introduced, and communication with the user is kept through pushing, mail, short messages and the like. User feedback is collected, and personalized service schemes are timely adjusted to meet the change of user demands. And comparing different personalized service schemes by using an A/B test method, analyzing user response and effect, optimizing the scheme and continuously improving. In the process of establishing the personalized service scheme, the user privacy is fully respected, related regulations and compliance requirements are met, and a transparent privacy policy is established. And continuously monitoring the effect of the personalized service scheme by using a monitoring tool and an index system, wherein the effect comprises user participation, conversion rate, sales and the like, and adjusting and optimizing according to the monitoring result.

By the method, enterprises can better know the demands of users and provide personalized services and recommendations, so that the satisfaction degree of the users is enhanced, the loyalty of the users is improved, and the sales growth is promoted.

The main functions of the invention are as follows:

1. establishing and recording a platform database: by establishing a comprehensive platform database, various activity data of visitors on an e-commerce platform are recorded, including commodity browsing, searching records, purchase history, interaction records and the like, and a data base is provided for subsequent personalized services and accurate analysis.

2. User portrayal analysis: carrying out customer and employee portrayal analysis on the visitor, and establishing a customer portrayal by deeply analyzing the activity data of the customer visitor, wherein the customer portrayal comprises browsing tracks, purchasing behavior, interests and hobbies and the like so as to understand the customer requirements; meanwhile, portrait analysis is carried out on staff and visitors, and influence and interaction level of staff on a platform are known.

3. And (3) designing a matching algorithm: and a matching algorithm is designed to divide visitors into different groups so as to better understand the characteristics and behaviors of the visitors, help a platform to locate user types more accurately, and provide a basis for personalized service and accurate marketing.

4. Personalized service formulation: and (3) formulating a personalized service scheme by utilizing the result of the matching algorithm, wherein the personalized service scheme comprises a primary marketing scheme and a secondary marketing scheme based on staff and visitors matched with the client, and providing personalized product recommendation, exclusive preference and the like so as to improve the satisfaction degree and purchase conversion rate of the client.

SCRM interaction weight definition and resource allocation: the method comprises the steps of defining weights of different interaction types, adjusting weight values in consideration of interaction influence, calculating user interaction scores by using a linear regression model, setting weight thresholds, and carrying out resource allocation by using a decision tree algorithm so as to realize intelligent personalized recommendation and service.

The invention aims to provide a comprehensive e-commerce marketing method, which realizes more accurate user portraits and personalized services by deeply analyzing user behaviors and interactions so as to promote user experience and sales growth.

In view of the above, after reading the present document, those skilled in the art should make various other corresponding changes without creative mental effort according to the technical scheme and the technical conception of the present invention, which are all within the scope of the present invention.

Claims (9)

1. An electronic commerce marketing method based on SCRM is characterized by comprising the following steps:

step one: establishing a platform database and recording the activity data of the visitor;

step two: cleaning the data to finish user duplication removal;

step three: user portrait analysis is carried out on the visitor;

step four: matching is carried out according to a matching algorithm;

step five: establishing personalized service:

(1) first-level marketing scheme formulated according to user portrayal of customer

The platform establishes a primary marketing scheme by comprehensively analyzing the information of interests, purchasing behaviors, geographical positions and the like of the clients, wherein the primary marketing scheme comprises the steps of providing personalized product recommendation, regular exclusive preference and customized service for the clients so as to improve the satisfaction degree and purchasing conversion rate of the clients;

(2) making a secondary marketing scheme according to the user portraits of staff and visitors matched by the clients;

based on the user portraits of employee visitors to which the clients are matched, the platform formulates a secondary marketing scheme, including establishing more intimate relationships with the employee visitors, providing personalized services, and introducing more clients through recommendations of the employee visitors.

2. The method for marketing the electronic commerce based on the SCRM according to claim 1, wherein the specific method for establishing the platform database in the first step is as follows: using a database management system (DBMS), designing a database structure, creating a physical database, selecting appropriate data types, implementing data security measures, importing initial data, formulating periodic backup and restore policies, optimizing database performance, documenting database structure and operating procedures.

3. The method for electronic commerce marketing based on SCRM according to claim 1, wherein the method for recording the activity data of the visitor in the first step is as follows: the method for recording the activity data of the visitor comprises the steps of implementing page tracking, and recording page browsing behaviors and clicking events of the visitor through browsing data; recording search behaviors including search keywords and search result clicking conditions; recording purchase history, including order information and transaction details; recording interaction behaviors, commenting, scoring and participating in topics; performing user identification to correlate activity data to a particular visitor; and introducing a real-time data updating technology to ensure the timeliness of the data.

4. The method for electronic commerce marketing based on SCRM according to claim 1, wherein the method for cleaning data in the second step is as follows: the method comprises the steps of processing missing values and abnormal values, unifying data formats, adopting a deduplication operation to clean based on key fields and time windows by using database tools and scripts, and verifying logical relations and uniqueness.

5. The method for conducting user portrait analysis on the visitor in the third step is as follows: the user portrait is divided into a client user portrait and an employee user portrait, and the client portrait data comprises browsing tracks, purchasing behaviors, interests, participation topics and interaction records; the employee portrait data comprises subdivision fields, content creation, interaction records and attention numbers.

6. The method according to claim 1, wherein the matching algorithm in the fourth step is designed as follows: and extracting key features from the interaction data of the clients and the staff, including interaction type and frequency, evaluating the similarity between the clients and the staff by using a similarity calculation method, judging whether the matching is successful, and triggering personalized service or recommendation.

7. The SCRM-based e-commerce marketing method of claim 6, wherein the matching algorithm design further comprises: based on the interaction weight value between the customer and the employee account, the system can record and update the interaction weight value in real time in the interaction process, reflect the activity degree and influence of the user on the platform, and when the interaction weight value between the customer and the employee reaches a set matching success threshold, the system judges that the matching is successful and triggers corresponding personalized service and marketing strategies; otherwise, if the weight value does not reach the set threshold, the client remains in the unmatched state and the system waits for the next matching opportunity.

8. The method for electronic commerce marketing based on the SCRM of claim 7, wherein the method for designing the interactive weight value comprises the following steps: the method comprises the steps of weighing items and weight proportion, wherein the weighing items comprise browsing behaviors, praise and like, comment behaviors, sharing behaviors, purchasing behaviors, searching behaviors, attention and fan interaction, participation in topics and discussions, activity participation and account setting and personal information updating.

9. The method of claim 8, wherein the weight ratio is assigned as: purchase behavior 0.25, browsing behavior 0.15, praise and like 0.1, comment behavior 0.1, share behavior 0.1, search behavior 0.1, attention and fan interaction 0.05, participation in topics and discussions 0.05, activity participation 0.05, account setting and personal information updating 0.05.