An intelligent marketing method and system based on user habits
By generating and optimizing user habitual self-portraits and personalized product recommendations, the problem of mismatch in marketing solutions caused by inaccurate user data is solved, sales efficiency and user experience are improved, and the competitiveness of shopping platforms is enhanced.
Patent Information
- Application Number
- CN202410134875.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-01-30
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-01-30
AI Technical Summary
In the big data environment, users' awareness of personal information protection has increased, resulting in inaccurate user data, affecting the accuracy of marketing plans, and thus affecting product recommendations and user shopping experience.
By collecting users' basic information, behavioral records and social data, generating user habitual self-portraits, optimizing data, performing positioning analysis and updating product marketing recommendation solutions, realizing personalized push and social behavior analysis, and improving sales efficiency and user experience.
It realizes the accurate processing of user habitual data and accurate recommendation of product marketing, improves sales efficiency and user satisfaction, reduces marketing costs, and enhances the competitiveness of shopping platforms.
Smart Images

Figure CN117934086B_ABST
Abstract
Description
Technical Field
[0001] The present invention proposes an intelligent marketing method and system based on user habits, belonging to the technical field of user habit intelligent marketing. Background Art
[0002] At present, in the big data environment, user information is collected more and more comprehensively. User habits are obtained through analysis of user data, and precise marketing is carried out to users based on the habits of users shopping on online platforms. This method increases the success rate of marketing. However, users are becoming more and more aware of protecting personal information. Users consciously do not fill in completely compliant big data information, resulting in inaccurate user data, mismatched marketing plans, inability to carry out precise marketing and poor user shopping experience. Summary of the invention
[0003] The present invention provides an intelligent marketing method and system based on user habits, which realizes accurate processing of user habit data, precise recommendation of product marketing, personalized push of user notifications, personalized layout of product recommendations and in-depth analysis of social behavior, thereby improving the sales efficiency of products and the shopping experience of users.
[0004] The present invention proposes an intelligent marketing method based on user habits, the method comprising:
[0005] Collect user habit data and generate user habit self-portraits;
[0006] Verify user habit self-portraits and optimize user habit data;
[0007] Perform positioning analysis based on optimized user habit data and generate product marketing recommendation plans;
[0008] Update user habit data, adjust product marketing recommendation plans, and obtain new product marketing recommendation plans.
[0009] Further, generating a user habit self-portrait based on the collected user habit data includes:
[0010] Collect the user's age, occupation, gender and region information registered by the user when using the service for the first time to obtain the user's basic information;
[0011] Collect users' search history, browsing history, purchase history, comment history, and sharing history to obtain information about users' behavior habits;
[0012] Collect user product purchase records, obtain product types, brands, price ranges, colors, sizes and models, and obtain user preference information;
[0013] Collect users' social comment records and sharing records to obtain information about their social habits;
[0014] The user's basic information, behavioral habit information, preference habit information and social habit information are integrated to generate a user habit self-portrait.
[0015] Further, based on the verification of the user's habit self-portrait, the user's habit data is optimized, including:
[0016] According to the user's self-portrait, when the user browses, 80% of the products that match the self-portrait are recommended, and 20% of the products that do not match the self-portrait are recommended. The browsing and purchase records of heterogeneous products are collected to obtain the attention data of heterogeneous products.
[0017] Analyze the attention data of heterogeneous products, compare and verify the attention data of heterogeneous products, extract the heterogeneous product information and add it to the user habit self-portrait to obtain optimized user habit data.
[0018] Further, the positioning analysis is performed based on the optimized user habit data to generate a product marketing recommendation plan, including:
[0019] Analyze user search records. If the searched product matches the user's personal information, record the product category, style, and applicable season. At the same time, record the user's search time to generate the user's product purchase timing habits. When the timing arrives, add the same searched product to the first product recommendation area. If it does not match the user, then in the next product cycle, 80% of the same type of searched products and 20% of the same type of searched products will be placed in the first product recommendation area.
[0020] Analyze the user's browsing history, obtain the items that have been browsed for a long time, query the type, brand, price range, color, size and model data of the items, extract similar products from the product library, and add them to the user's second product recommendation area; analyze the user's purchase history, query the various data of the purchased products, and if the purchased products are consumables, extract the products and similar functional products before the consumable period arrives, and add them to the user's third product recommendation area;
[0021] Analyze user review records, obtain user review opinions, extract products similar to those in the opinions, and add them to the user's fourth product recommendation area;
[0022] Integrate various data on user habits, put the products on the first, second, third and fourth product recommendation areas, and cycle the positions of the product recommendation areas to obtain the product recommendation page.
[0023] Further, adjusting the product marketing recommendation plan according to the updated user habit data to obtain a new product marketing recommendation plan includes:
[0024] Track the user's search history. When an active search record occurs, add the searched product attributes to the real self-portrait, and add the same type of products in the product library to the first product recommendation area. If no transaction record is generated, the product will continue to be placed in the first product recommendation area. If a transaction is generated for the product, it will be placed in the third product recommendation area after the usage cycle.
[0025] Track user browsing history, analyze browsing speed, classify browsed products and product attribute factors according to the analysis results, add high-weighted products and product factors to user habits, increase the second product recommendation weight, and update the second product recommendation area;
[0026] Track user purchase records, analyze the purchase records, and place products that are logically related to the use of the purchased products, or products of the same category as the purchased products but with significant differences in price, model, and color, in the third product recommendation area, and update the recommended products in the third product recommendation area;
[0027] Track user evaluation records. When active transaction records occur, combine the transaction product attributes with the evaluation records, and add or subtract product attributes based on the star rating of the evaluation records. When the star rating is low, upgrade the product attribute level, increase the product attribute requirements, update the product requirements to user habits, and update the products in the fourth product recommendation area.
[0028] The backend analyzes the sales performance of each recommendation area after the products are updated. If the number of sales increases after the products are updated, or the sales remain the same or increase, it means that the product marketing direction matches the user habits and the product marketing direction should be continued. If the number of sales decreases and the average sales also decreases, it means that the marketing recommendation direction is wrong, and the marketing direction should be regenerated and the products in the recommendation area should be updated.
[0029] Furthermore, the user's desired products are monitored, and when the desired products meet the user's expectations, an email or SMS notification is sent.
[0030] Furthermore, we record the user's usage time, infer the user's usage patterns, arrange the product recommendation page according to the length of the user's usage time, put the products with high weight that the user is accustomed to at the front of the recommendation, and obtain the product recommendation page that matches the user's browsing time.
[0031] Further, the user's social habit information is analyzed, and the users' social followers are recorded. According to the commodity communication density and purchase status of the followers, the information of the followers on social media is retrieved, and the information with high commodity communication density weight is added to the user's habits, and the recommended commodities are updated and placed in the fourth commodity recommendation area.
[0032] On the other hand, the present application also provides an intelligent marketing system based on user habits, the system comprising:
[0033] The habit data collection module is used to collect user habit data and generate user habit self-portraits;
[0034] The habit data verification module is used to verify the user's habit self-portrait and optimize the user's habit data;
[0035] Product marketing recommendation module, used to perform positioning analysis based on optimized user habit data and generate product marketing recommendation plans;
[0036] Product marketing adjustment module, used to update user habit data, adjust product marketing recommendation plans, and obtain new product marketing recommendation plans;
[0037] The marketing information sending module is used to monitor the products that users are interested in and send emails or SMS notifications when the products meet the users' expectations.
[0038] The duration section recommendation module is used to adjust the product recommendation page according to the length of time the user is using the product, put the products with high weight of the user's habitual products at the front of the recommendation, and obtain the product recommendation page that matches the user's browsing time;
[0039] The social record analysis module is used to analyze user social habit information, extract information about intensive product exchanges and add it to user habits, and update recommended products and put them into the product recommendation area.
[0040] Furthermore, an intelligent marketing system based on user habits includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps of the method are implemented when the processor executes the program.
[0041] The invention has the following beneficial effects: by collecting the basic information registered by the user and the various behavior records of the user on the online shopping platform, these data are integrated and analyzed to obtain the user's habitual self-portrait. Through the user's habitual self-portrait, the user's personality, hobbies, and the type, style and economic affordability of the favorite products can be obtained, and the user's self-portrait can be initially formed, which is conducive to accurately positioning the user, providing the user with the products of interest when browsing the user platform, and conducting precision marketing;
[0042] By adding heterogeneous products to the browsed products, collecting browsing data of heterogeneous products, and analyzing and acquiring the collected results, it is possible to optimize the user's self-portrait, and verify the user's self-registration information through real user behavior data to obtain the user's real habits. By analyzing the user's real habit data and generating a product marketing recommendation plan, it is possible to extract the attributes of the corresponding recorded products based on the user's behavior records, and put the products with the same attributes on the corresponding recommendation area, and cycle the products in the recommendation area to obtain a product marketing plan that meets the user's habits, greatly arouse the user's interest, and many products that meet the user's interests stimulate the user's purchasing ideas and promote the transaction rate of products. By tracking user habit data, it is possible to recommend products to new users in a timely manner based on changes in user behavior records when the user's ideas and needs change, change marketing product recommendations, and seize the user's current By analyzing user habit data, it is possible to mark products that users like the most. When new products are released or the prices are reduced and marketing activities that meet the users’ needs appear, it is possible to send users notifications of their favorite products and complete transactions through precision marketing. By recording the regularity of users’ use of online platforms, it is possible to sort products that meet users’ habits according to their current relative weights, arouse users’ interest to the greatest extent within a limited time, and increase the marketing transaction rate in a short period of time. By analyzing users’ social habit information, it is possible to understand users’ social relationships. Through the communication of products in social relationships and other social followers who frequently communicate, it is possible to obtain users’ potential product needs and their actual consumption level from the followers who frequently communicate. By expanding users’ information, it is possible to adjust users’ marketing strategies and obtain the most suitable marketing methods at the moment. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] Figure 1 is a flow chart of the method of the present invention;
[0044] Figure 2 The system block diagram of the system of the present invention. DETAILED DESCRIPTION
[0045] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.
[0046] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. The embodiments described are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.
[0048] One embodiment of the present invention provides an intelligent marketing method based on user habits, the method comprising:
[0049] Collect user habit data and generate user habit self-portraits;
[0050] Verify user habit self-portraits and optimize user habit data;
[0051] Perform positioning analysis based on optimized user habit data and generate product marketing recommendation plans;
[0052] Update user habit data, adjust product marketing recommendation plans, and obtain new product marketing recommendation plans.
[0053] The working principle of the above technical solution is:
[0054] First, collect user habit data. This step is the basis of the entire intelligent marketing method. Collect user behavior habit data through various channels (such as user search records, browsing records, purchase records, comment records, sharing records and online time, etc.). These data can reflect user needs, preferences, consumption habits and other information. Based on the collected user habit data, generate the user's "habit self-portrait" through data analysis and organization. This self-portrait can be regarded as a digital user model, which describes the user's various habits in detail;
[0055] Then, the user's habit self-portrait is verified, and product recommendations are made based on the generated habit self-portrait. The generated product self-portrait is analyzed and compared with the user's actual behavior. If there is any inconsistency or complete inconsistency, the inconsistent behavior will be supplemented and corrected, or data will be collected again to replace the previous self-portrait data to obtain optimized user habit data.
[0056] After that, positioning analysis is performed based on the optimized user habit data. Based on the optimized user habit data, a deeper analysis is performed. For example, the user's interests can be analyzed and combined with the best-selling products of the season, or the products that are suitable for the user at the moment can be recommended based on the user's age, occupation, physical condition, and psychological state. Then, a product marketing recommendation page is generated based on this information. This page will list in detail the marketing strategies and recommended products for specific users, thereby generating a product marketing recommendation plan.
[0057] Finally, user habits may change over time, so it is necessary to regularly update user habit data and adjust product marketing recommendation plans accordingly to obtain new product marketing recommendation plans.
[0058] The effects of the above technical solution are:
[0059] By acquiring user habit data and generating user habit self-portraits, and through in-depth analysis of user habit data, we can accurately understand user needs and preferences, thereby providing users with more accurate product recommendations, which helps improve product sales and also helps improve users' shopping experience;
[0060] By optimizing user habit data, target users can be more accurately located, thus reducing unnecessary computing power investment, advertising and promotion costs, and reducing marketing costs. At the same time, by optimizing user habit data, product recommendations can be made more efficiently, further improving marketing efficiency.
[0061] By optimizing user habit data through positioning analysis and generating product marketing recommendation plans, we can better grasp market trends and user needs, and thus conduct more targeted product marketing, which helps to increase the market share of products and enhance the competitiveness of shopping platforms.
[0062] By continuously tracking and analyzing user habits, we can continuously optimize and adjust product recommendation plans to ensure that we provide users with more personalized services. This personalized service can better meet user needs and further improve user satisfaction. At the same time, by collecting, analyzing and using user data to formulate marketing strategies, we can ensure that recommendation plans are always highly matched with user habits, thereby better responding to market changes and improving the resilience of shopping platforms.
[0063] In one embodiment of the present invention, generating a user habit self-portrait based on the collected user habit data includes:
[0064] Collect the user's age, occupation, gender and region information registered by the user when using the service for the first time to obtain the user's basic information;
[0065] Collect users' search history, browsing history, purchase history, comment history, and sharing history to obtain information about users' behavior habits;
[0066] Collect user product purchase records, obtain product types, brands, price ranges, colors, sizes, and models, and obtain user preference information;
[0067] Collect users' social comment records and sharing records to obtain information about their social habits;
[0068] The user's basic information, behavioral habit information, preference habit information and social habit information are integrated to generate a user habit self-portrait.
[0069] The working principle of the above technical solution is:
[0070] First, by collecting basic information about users, such as age, occupation, gender, and region, we can gain a preliminary understanding of their background and characteristics. This basic information can provide basic data for subsequent analysis of user habits.
[0071] Secondly, user behavior information, such as search history, browsing history, purchase history, comment history, and sharing history, can reveal users’ needs, interests, and consumption habits in more depth, and obtain information about users’ behavior habits;
[0072] Furthermore, by collecting the user's product purchase records, we can further obtain information such as product type, brand, price range, color, size and model. This specific product information can help to more accurately judge the user's preference habits and obtain preference behavior information;
[0073] In addition, by collecting users' social comment records and sharing records, we can provide additional clues about users' social behaviors and speeches, and obtain information about their social habits. For example, users' social speeches and shared content can reflect their values, attitudes, and interests.
[0074] Finally, by combining the above basic information, behavioral habit information, preference habit information and social habit information, a complete user portrait can be formed. This portrait not only includes the basic characteristics of the user, but also deeply reflects their behavioral habits, consumption preferences and social behaviors.
[0075] The effects of the above technical solution are:
[0076] By collecting basic user information, we can provide basic data for subsequent analysis of user habits; by collecting behavioral habit information, we can reveal users' needs, interests and consumption habits; by collecting preference behavior information, we can more accurately judge users' preference habits; by collecting social habit information, we can reflect values, attitudes and interests, and help discover users' potential habits; by collecting multi-faceted user data, we can fully understand users' personal information, behavioral habits, consumption preferences and social habits. This comprehensive understanding helps to more accurately grasp users' needs and characteristics, and provide strong support for subsequent intelligent marketing;
[0077] By integrating the user's basic information, behavioral habit information, preference habit information and social habit information, a more accurate user habit self-portrait can be generated. This accurate user portrait can better reflect the user's real needs and preferences, and further improve the accuracy of smart marketing;
[0078] Based on the user's habit self-portrait, we can provide users with more personalized services and recommendations. This personalized service can better meet user needs, enhance user satisfaction, and further promote the improvement of user stickiness.
[0079] In one embodiment of the present invention, the verifying of the user's habit self-portrait and optimizing the user's habit data includes:
[0080] According to the user's self-portrait, when the user browses, 80% of the products that match the self-portrait are recommended, and 20% of the products that do not match the self-portrait are recommended. The browsing and purchase records of heterogeneous products are collected to obtain the attention data of heterogeneous products.
[0081] Analyze the attention data of heterogeneous products, compare and verify the attention data of heterogeneous products, extract the information of heterogeneous products and add it to the user's habit self-portrait to obtain and optimize the user's habit data
[0082] The working principle of the above technical solution is:
[0083] First, based on the user's habitual self-portrait, a certain proportion of products are recommended to the user. For example, if the user is a college student under the age of 23, electronic entertainment products, sports and leisure clothing or interview suits, employment guidance and postgraduate entrance examination tutoring will be recommended. Among them, 80% of the products are the same as the self-portrait, and the remaining 20% are heterogeneous products that match the self-portrait. The purpose of this is to test and verify the accuracy of the user's habitual self-portrait; for the case where the user browses heterogeneous matching products, the online shopping backend system will collect the browsing records and purchase records of these products. These data are called heterogeneous product attention data, which reflects the user's interest and attention to heterogeneous products; record the total number of views M and the total number of views m of heterogeneous products used by the user on the shopping platform at a single time, the total browsing time of the user on the shopping platform at a single time is T, the total browsing time of heterogeneous products is t, the heterogeneous product attention data is L, and each time the user enters the shopping platform, the platform usage count is counted once, and the platform usage count is n.
[0084] Next, we conduct an in-depth analysis of these heterogeneous product attention data L. If the heterogeneous product attention data is 0, it means that the user has no interest or attention in the recommended heterogeneous products, indicating that the user's habit self-portrait has relatively accurately reflected the user's real habits;
[0085] However, if the heterogeneous product attention data is not 0, it does not mean that the user is interested in heterogeneous products. It may be a mistaken click or a curious click. Analyze the heterogeneous product attention data.
[0086] When 0≤L<0.05, it means that the heterogeneous product is a late click by the user or does not arouse the user's interest.
[0087] When 0.05≤L<0.1, it means that heterogeneous products arouse the user's interest, and the attributes of heterogeneous products are added to the end of user habits.
[0088] When 0.1≤L<0.3, it means that heterogeneous products can be accepted by users, and the attributes of heterogeneous products are changed to the same products.
[0089] When 0.3≤L<0.5, it means that heterogeneous products are well-liked by users, and the weight of heterogeneous product attributes in user habits is increased.
[0090] When 0.5≤L<1, it means that heterogeneous products are most in line with the current user preferences, and the habit weight of the corresponding homogeneous products is reduced.
[0091] By analyzing the product attention data results, the detailed information of these products is added to the user's real habit self-portrait, and the types of heterogeneous products are updated to optimize the user's habit data accordingly.
[0092] The effects of the above technical solution are:
[0093] Through the user's habit self-portrait, we can preliminarily generate a product marketing recommendation plan for the online shopping platform, put recommended products on the shelves when the user browses freely, and also recommend heterogeneous products that are contrary to the user's self-portrait. By analyzing the heterogeneous product attention data, we can verify whether the user's habit self-portrait data is complete and correct. The lower the heterogeneous product attention data, the more accurate the user's habit self-portrait is. We can analyze that the user's behavior is more frank, and the collected user's habit information is less misleading. Through slight adjustments to the user's self-portrait, we can provide users with accurate marketing recommendations and increase merchants' interest and transaction rates on the shopping platform. By analyzing the user's heterogeneous product attention data, the higher the heterogeneous product attention level, the more popular the attributes of the heterogeneous product are with users. According to the heterogeneous product attention level, we can add, modify and weight the user's habit self-portrait content in different product attention level ranges, and optimize the user's habit data to different degrees.
[0094] An embodiment of the present invention performs positioning analysis based on optimized user habit data to generate a product marketing recommendation plan, including:
[0095] Analyze user search records. If the searched product matches the user's personal information, record the product category, style, and applicable season. At the same time, record the user's search time to generate the user's product purchase timing habits. When the timing arrives, add the same searched product to the first product recommendation area. If it does not match the user, then in the next product cycle, 80% of the same type of searched products and 20% of the same type of searched products will be placed in the first product recommendation area.
[0096] Analyze the user's browsing history, obtain the items that have been browsed for a long time, query the type, brand, price range, color, size and model data of the items, extract similar products from the product library, and add them to the user's second product recommendation area;
[0097] Analyze user purchase records and query various data of purchased goods. If the purchased goods are consumables, extract the goods and similar functional goods before the consumable period arrives and add them to the user's third product recommendation area;
[0098] Integrate various data on user habits, put the products on the first, second, third and fourth product recommendation areas, and cycle the positions of the product recommendation areas to obtain the product recommendation page.
[0099] Based on various data on user habits, products are put on the first, second, third and fourth product recommendation areas. The positions of the product recommendation areas are not fixed but alternate in a cycle.
[0100] The working principle of the above technical solution is:
[0101] First, by analyzing the user's search records, the products searched by the user are divided into self-use and other-use. For self-use products, the category, style, and applicable season of the products are counted through the search records. By knowing whether the user searches before the arrival or after the season, the user's product purchase timing habits are understood. When the time comes, the products suitable for the moment are placed in the first product recommendation area, and the product recommendation booth is placed in front. For other-use products, the current search time is for gifts, and the types and search times of other-use products are counted. Before the same time of the next year or before the consumption cycle of other-use products arrives, 7 days before the time arrives, similar products are placed in the first product recommendation area, and the product recommendation booth is placed in front.
[0102] Then, by analyzing the user's browsing history, we obtain the items that the user has browsed for a long time, query the type, brand, price range, color, size and model data of the items, extract similar products from the product library, add them to the user's second product recommendation area, and put the category attributes of the items that have been browsed for a long time into the user's habits. We also weight the category attributes of the items based on the length of time, increasing the weight of items that have been browsed for a long time. As time changes, the weight of the browsed product attributes also changes continuously, thereby continuously updating the products in the second product recommendation area.
[0103] Secondly, by analyzing the user's purchase records, the attributes of the purchased goods are extracted and added to the user's habits. The time of consumables is estimated through big data. Before the wear date arrives, the same goods are extracted from the product library and added to the user's third product recommendation area for pre-recommendation. If the user does not purchase the goods within the wear date, the product recommendation order is lowered. When the user purchases the same product again, the date and product quantity are analyzed and calculated to obtain the product's usage efficiency and calculate the next wear date.
[0104] After that, by analyzing the user's comments on the product, extract the user's approval and disapproval opinions on the product, and obtain the product framework recognized by the user, whether it focuses on comprehensive functions, quality first, product details, additional services, or follows the public's choice, so as to classify the products that meet the user's self-portrait in detail, obtain the decisive factors for the user to buy the product, and extract the products that meet the user's detailed classification in the product library, and sort the product booths with more weight and add them to the fourth product recommendation area;
[0105] Finally, the first, second, third, and fourth product recommendation areas will be cycled on the shopping platform booth. During the cycle, the user's current shopping platform behavior will be analyzed to obtain products in the four product recommendation areas in real time, so that users can quickly obtain the required products when browsing on the shopping platform.
[0106] The effects of the above technical solution are:
[0107] By analyzing user search records and displaying products in the first product recommendation area, we can accurately analyze customers' product purchasing habits, narrow the scope of user recommendations, and accurately target users for marketing when the time comes, thereby increasing marketing success rates, reducing marketing costs, and reducing data backend computing power.
[0108] By analyzing browsing records, we can extract products that users care about based on browsing time, add product attributes to user habits, and obtain user habit weight preferences through user browsing time analysis. According to user weights, we can sort products that meet user habit preferences in the second product recommendation area, so that users can get the product information of their habit preferences first, and improve the marketing success conversion rate;
[0109] By analyzing purchase records, products can enter the third product recommendation area before they are worn out, and the recommendation interface can be placed in front according to the time urgency, which can remind users to replenish products, increase users' demand for products, and promote the probability of product transactions. In addition, through the second purchase records of similar products, the user's loss period calculation can be optimized, the loss date calculation can be made more accurate, the user's loss frequency can be accurately located, and user needs can be further grasped, and the importance of products can be increased to promote transactions;
[0110] By analyzing user review records, we can obtain the framework of products recognized by users, understand the underlying logic of users' purchases, and obtain the decisive factors for users to purchase products. By adding the decisive factors to the user's habitual self-portrait, we can extract products containing the details of the decisive factors from the products and add them to the fourth product recommendation area. When users browse, we can quickly obtain products that meet the details that users care about, improve users' satisfaction with recommended products, thereby promoting product transactions and completing marketing tasks;
[0111] By not fixing the display area of the four product recommendation areas and displaying them in a cycle, users will not feel bored or tired when browsing due to repeated products. In addition, the products in the recommendation area are updated during the cycle display to make the products more in line with the user's current shopping psychological needs, thereby achieving shopping intentions and completing marketing goals.
[0112] In one embodiment of the present invention, updating user habit data, adjusting a product marketing recommendation plan, and obtaining a new product marketing recommendation plan includes:
[0113] Track the user's search history. When an active search record occurs, add the searched product attributes to the real self-portrait, and add the same type of products in the product library to the first product recommendation area. If no transaction record is generated, the product will continue to be placed in the first product recommendation area. If a transaction is generated for the product, it will be placed in the third product recommendation area after the usage cycle.
[0114] Track user browsing history, analyze browsing speed, classify browsed products and product attribute factors according to the analysis results, add high-weighted products and product factors to user habits, increase the second product recommendation weight, and update the second product recommendation area;
[0115] Track user purchase records, analyze the purchase records, and place products that are logically related to the use of the purchased products, or products of the same category as the purchased products but with significant differences in price, model, and color, in the third product recommendation area, and update the recommended products in the third product recommendation area;
[0116] Track user evaluation records. When active transaction records occur, combine the transaction product attributes with the evaluation records, and add or subtract product attributes based on the star rating of the evaluation records. When the star rating is low, upgrade the product attribute level, increase the product attribute requirements, update the product requirements to user habits, and update the products in the fourth product recommendation area.
[0117] The backend analyzes the sales performance of each recommendation area after the products are updated. If the number of sales increases after the products are updated, or the sales remain the same or increase, it means that the product marketing direction matches the user habits and the product marketing direction should be continued. If the number of sales decreases and the average sales also decreases, it means that the marketing recommendation direction is wrong, and the marketing direction should be regenerated and the products in the recommendation area should be updated.
[0118] The working principle of the above technical solution is:
[0119] First, track the user's search history, extract the attributes of the product that the user has newly searched for or the attributes of the product that have not appeared before, add the extracted product attributes to the user's habitual real self-portrait, and put the same type of products that meet the new product attributes into the first product recommendation area through search. If the user does not make a transaction in the second browsing, the product will be placed in the first product recommendation area for 3 days. If a transaction is made within 3 days, the loss of the product is estimated through big data, and the product is placed in the third product recommendation area after the use cycle. If no transaction is made, the order of the product in the first product recommendation area is gradually reduced;
[0120] Then, we track the user's browsing history and analyze the browsing speed. The total browsing length of a single use of the shopping platform is Q, and the total browsing time is T. The average browsing speed is V = Q / T.
[0121] Assume that the length of a product page in a single browsing process of the shopping platform is Q1, the browsing time of the product page is T1, the average browsing speed of the product is V1, the sections of the page are q1, q2, q3, ... qn, the browsing time of the section is t1, t2, t3, ... tn, and the browsing speed of each section of a product is v1, v2, v3, ... vn.
[0122] Then the average browsing speed of a product is V1, V1=Q1 / T1,
[0123] When V ≥ V1, the user is interested in a certain product and the browsing speed of each module of the product is compared.
[0124] The browsing speed of each module of the product is v1=q1 / t1,v2=q2 / t2,v3=q3 / t3,....,vn=qn / tn,
[0125] Ratio v1, v2, v3, ...vn to V1 in turn, the ratio sn = vn / V1,
[0126] If sn = 1, it means that the browsing speed of the article section is consistent with the average browsing speed, and the attribute weight of the product section does not change;
[0127] If sn>1, it means that the browsing speed of the article section is greater than the average browsing speed, and the attribute weight of the product section is reduced;
[0128] If sn<1, it means that the browsing speed of the article section is lower than the average browsing speed, and the attribute weight of the product section is increased;
[0129] Add the newly increased product attributes to the user's habits, or update the product attributes of the user's habits, and put the products that meet the user's habit weights into the second product recommendation area, so as to update the recommended products according to the browsing habits;
[0130] After that, the user's purchase record is tracked, and the purchased goods are searched using logical association, or the purchased goods are searched for the same type of goods with different styles, models, and designs, and are placed in the third product recommendation area as an extended product of the purchased goods, as a purchase record random recommendation area, to expand the user's product demand range;
[0131] Secondly, track user evaluation records. When users comment on the products that have generated transactions, conduct a comprehensive analysis of the attributes of the products and the comment records. When the comment record is satisfied or very satisfied, maintain the current user habits, indicating that the recommended products meet the user's expectations. When the comment record is dissatisfied and points out the specific reasons, extract the product attributes of the specific reasons and update them to the user's habits, increase the user's attention to the product attributes of the habits, and update the recommended products that meet the user's habits into the fourth product recommendation area for update;
[0132] Finally, analyze the sales performance after the change of the product recommendation area. If the product weight changes, the product recommendation order is changed, and the number of sales increases, it means that the weight matches the user habits, the marketing direction is correct, and it is continuously updated to ensure that products with the same product attributes are focused on; if the number of sales decreases, but the sales revenue remains the same or increases, it means that the marketing direction is correct. The quality of the product may be excellent and the price may be relatively increased, but the user's purchasing power is limited, so a decrease in the number of product sales is normal; if the number of sales decreases and the average sales revenue also decreases, it means that the marketing recommendation direction is incorrect. The product marketing recommendation direction returns to the previous product recommendation, and the user self-portrait is re-portraitd to obtain the marketing plan again.
[0133] The effects of the above technical solution are:
[0134] By tracking user search records, we can timely capture the user's current product interests, update similar products in the product recommendation area in a timely manner, and recommend products when the user has the strongest purchase intention, promote user product demand, and increase the transaction rate of marketing plans. At the same time, we can estimate the loss of the sold products and recommend products again when the loss period arrives, so as to encourage users to repurchase and increase the repurchase rate of products.
[0135] By tracking the user's browsing history and analyzing the browsing speed, we can obtain the user's attention to products and product sections, so as to shift the weight of product attributes and segment the importance of product attributes. In this way, we can refine and locate user habits through browsing history, and make personalized product recommendations to users based on their habits, thereby increasing user satisfaction, reducing user screening time, and increasing the user's transaction volume per unit time.
[0136] By tracking user purchase records, we can obtain extended products that are positively and reversely related to the purchased products, expand the user's demand range, ensure that the recommended products are new, and prevent the products from not being updated for a long time and reducing the user's purchasing interest. In addition, by recommending extended products based on user records, we can explore more product attributes that users are interested in, optimize and supplement user habit data, increase user stickiness to the shopping platform, gain user trust, and ensure the continuity of marketing plans.
[0137] By tracking user evaluation records, it is possible to analyze user comments and the products that users have traded, obtain user satisfaction, adjust the attributes of purchased products according to the satisfaction, obtain the attributes of products that users are satisfied with, extract and search for the attributes of products that users are satisfied with, and add the searched products to the fourth product recommendation area, which can increase the accuracy of intelligent marketing product selection. By obtaining more products that are more suitable for users, the conversion rate of marketing plans can be improved, thereby improving the sales performance of the platform;
[0138] By tracking and analyzing sales performance, we can determine the correctness of product selection in intelligent marketing plans. Changes in the number of sales and sales volume can intuitively show the results of marketing plans. In addition, products sold on online shopping platforms are mainly daily fast-moving consumer goods with short marketing cycles. Therefore, sales plans can be adjusted based on changes in performance, thereby improving the transaction efficiency of recommended products, reducing user browsing time, and improving platform marketing performance.
[0139] One embodiment of the present invention monitors the user's desired products and sends an email or SMS notification when the desired products meet the user's expectations.
[0140] The working principle of the above technical solution is:
[0141] Analyze user habit data and monitor products that users browse repeatedly or collect through their product browsing data. If such products are in line with user habits but do not meet user purchase expectations due to product timing or price / performance ratio reasons, when the monitored products are on sale and may meet user expectations, the user will be notified via email or SMS.
[0142] The effects of the above technical solution are:
[0143] By monitoring and obtaining products that meet user expectations, we can send notification text messages and emails to customers, and complete personalized marketing recommendations for users. At the same time, accurate product recommendations can improve user favorability without causing disgust due to frequent advertisements, increase user stickiness and shopping convenience, complete sales through precision marketing, reduce marketing advertising costs, and enhance marketing efficiency.
[0144] An embodiment of the present invention records user usage time, infers user usage patterns, arranges product recommendation pages according to the length of user usage time, places products with high user habit weights at the front of the recommendation, and changes the product circulation speed to obtain product recommendation pages that match the user's browsing time.
[0145] The working principle of the above technical solution is:
[0146] Statistics are collected on the total browsing time of a user on a shopping platform in a single visit, and the average usage time of the user is calculated. During the user's usage time, the products in the product recommendation area are laid out. If the browsing time is longer than the platform recommended product display time, the user's products are laid out normally, and the cycle speed of the product recommendation page is consistent with the original system preset. If the browsing time is less than the platform recommended time, the products with higher weights in the user's habits are placed at the front of the product recommendation area, and the product cycle time is compressed to ensure that a product stays in the field of vision for at least 3 seconds. The cycle time of the product recommendation area is changed to obtain a product recommendation page that matches the user's browsing time.
[0147] The effects of the above technical solution are:
[0148] By obtaining product recommendation pages that match the user's browsing time, the product recommendation pages are made more consistent with the user's browsing habits, thereby improving user satisfaction. By understanding the user's habits and obtaining user-recommended products, the user's browsing rhythm can be customized to ensure that the user browses the recommended products that meet the user's habits during the use of the platform, thereby arousing the user's interest and completing the transaction of the user's accustomed products.
[0149] One embodiment of the present invention analyzes user social habit information, records the users' social followers, retrieves information on the users based on the commodity communication density and purchase status of the followers, adds information with high commodity communication density weight to the user's habits, and updates recommended commodities to be placed in a fourth commodity recommendation area.
[0150] The working principle of the above technical solution is:
[0151] Analyze the user's social habit information, record the people the user follows in his social circle, retrieve the information of the people he follows on social media based on the intensity of product communication and purchase behavior of the people he follows, count the interaction frequency of the people he follows, count the top 6 people with higher interaction frequency, count the products with intensive communication, count the product attributes of products with intensive communication, extract the products and product attributes with intensive communication, reduce the weight of the extracted attributes by one-sixth and add them to the user's habits, update the recommended products, and put them in the fourth product recommendation area.
[0152] The effects of the above technical solution are:
[0153] By analyzing the people followed by the user in the social circle, we can obtain the people with high interaction rate in the circle of followers, and by counting their product habit attributes, we process the product habit attributes and add them to the user habits to update the recommended products. We can obtain the hidden common habits of users through the circle of followers, expand the scope of recommended products that meet user habits, and mine user habits to guide users to expand the scope of product demands. On the basis of satisfying user habits, we can expand the diversification of product attributes, enhance user interest, prevent user loss, and make product recommendations more in line with users' social needs and trends, thereby increasing the chances of product marketing success.
[0154] One embodiment of the present invention is a user habit-based intelligent marketing system, the system comprising:
[0155] The habit data collection module is used to collect user habit data and generate user habit self-portraits;
[0156] The habit data verification module is used to verify the user's habit self-portrait and optimize the user's habit data;
[0157] Product marketing recommendation module, used to perform positioning analysis based on optimized user habit data and generate product marketing recommendation plans;
[0158] Product marketing adjustment module, used to update user habit data, adjust product marketing recommendation plans, and obtain new product marketing recommendation plans;
[0159] The marketing information sending module is used to monitor the products that users are interested in, and send emails or SMS notifications when the products meet the user's expectations;
[0160] The duration section recommendation module is used to adjust the product recommendation page according to the length of time the user is using the product, put the products with high weight that the user is used to at the front of the recommendation, change the product circulation speed, and obtain the product recommendation page that matches the user's browsing time;
[0161] The social record analysis module is used to analyze the user's social habits information, extract the information of intensive commodity exchanges and add it to the user's habits, and update the recommended commodities and put them into the commodity recommendation area;
[0162] One embodiment of the present invention is a user habit-based intelligent marketing system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of any one of the methods described above when executing the program.
[0163] The working principle of the above technical solution is:
[0164] First, use the habit data collection module to collect user habit data and generate user habit self-portraits; then, use the habit data verification module to verify the user habit self-portraits and optimize the user habit data; then, use the product marketing recommendation module to perform positioning analysis based on the optimized user habit data and generate a product marketing recommendation plan; then, use the product marketing adjustment module to update the user habit data, adjust the product marketing recommendation plan, and obtain a new product marketing recommendation plan; then, use the marketing information sending module to monitor the user's intended products, and when the intended products meet the user's expectations, send an email or SMS notification to the user; then, use the time section recommendation module to adjust the product recommendation page according to the length of user usage time, place products with high user habit weights at the front end of the recommendation, change the product circulation speed, and obtain a product recommendation page that meets the user's browsing time; finally, use the social record analysis module to extract the attribute information of products with intensive communication to user habits, update the recommended products, and put them into the product recommendation area.
[0165] The effects of the above technical solution are:
[0166] By using the habit data collection module and the habit data verification module, the system can automatically and quickly collect users' habit data and verify the accuracy of this data, which improves the speed and efficiency of data processing, enabling the system to provide users with more accurate personalized recommendations, reduce unnecessary computing power investment, advertising and promotion costs, and reduce marketing costs;
[0167] The product marketing recommendation module performs positioning analysis based on optimized user habit data and generates product marketing recommendation plans, making product recommendations more accurate, improving product sales efficiency and user shopping experience, and better grasping market trends and user needs, which helps to increase the market share of products and enhance the competitiveness of shopping platforms;
[0168] Through the product marketing adjustment module, the shopping platform can update user habit data in real time based on user feedback and behavior, and adjust the product marketing recommendation plan so that the recommendation plan always maintains a high degree of match with user habits, thereby better responding to market changes and improving the shopping platform's adaptability;
[0169] The marketing information sending module can send notifications to users via email or SMS when items that highly meet user expectations appear, which increases the shopping convenience of users and thus increases user stickiness, improves the reach and conversion rate of marketing information, reduces marketing advertising costs, and enhances marketing efficiency;
[0170] The duration section recommendation module can adjust the product recommendation page according to the length of time the user is using the product, put the products that meet the user's habits and have high weights at the front of the recommendation, and change the circulation speed of the products according to the user's usage habits. This makes the product recommendation page more in line with the user's browsing habits, improves user satisfaction, and ensures that the user browses all the recommended products within the browsing time as much as possible, displays products that meet the user's habits, arouses the user's interest, and encourages the user to buy the products;
[0171] The social record analysis module can analyze the user's social habit information, extract the commodity attribute information with intensive communication into the user's habits, and update the recommended commodities to add to the commodity recommendation area. It can guide users to expand the scope of commodity demand, expand the diversity of commodity attributes on the basis of satisfying user habits, enhance user interest, prevent user loss, and make commodity recommendations more in line with users' social needs and trends, thereby increasing the sales opportunities of commodities.
[0172] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. An intelligent marketing method based on user habits, characterized in that: The method comprises: Collect user habit data and generate user habit self-portraits; Verify user habit self-portraits and optimize user habit data; Perform positioning analysis based on optimized user habit data and generate product marketing recommendation plans; Update user habit data, adjust product marketing recommendation plans, and obtain new product marketing recommendation plans; Verify the user's habit self-portrait and optimize the user's habit data, including: According to the user's self-portrait, when the user browses, 80% of the products that match the self-portrait are recommended, and 20% of the products that do not match the self-portrait are recommended. The browsing and purchase records of heterogeneous products are collected to obtain the attention data of heterogeneous products. Analyze the heterogeneous product attention data, compare and verify the heterogeneous product attention data, extract the heterogeneous product information and add it to the user habit self-portrait to obtain optimized user habit data; among them, the heterogeneous product attention data L is obtained by the following formula: Among them, the total number of views M and the total number of views of heterogeneous products used by the user in a single use on the shopping platform are recorded, the total browsing time of the user in a single use on the shopping platform is T, the total browsing time of heterogeneous products is t, and the number of platform uses is n; When 0≤L<0.05, the heterogeneous products are either missed by the user or do not arouse the user's interest. When 0.05≤L<0.1, heterogeneous products arouse user interest, and the attributes of heterogeneous products are added to the end of user habits. When 0.1≤L<0.3, heterogeneous products can be accepted by users, and the attributes of heterogeneous products are changed to the same product. When 0.3≤L<0.5, the heterogeneous products are well-liked by users, and the weight of the attributes of heterogeneous products in user habits is increased. When 0.5≤L<1, heterogeneous products are most in line with the current user preferences, and the habit weight of the corresponding homogeneous products is reduced; Update user habit data, adjust product marketing recommendation plans, and obtain new product marketing recommendation plans, including: Track the user's search history. When an active search record occurs, add the searched product attributes to the real self-portrait, and add the same type of products in the product library to the first product recommendation area. If no transaction record is generated, the product will continue to be placed in the first product recommendation area. If a transaction is generated for the product, it will be placed in the third product recommendation area after the usage cycle. Track user browsing history and analyze browsing speed. The total browsing length of a shopping platform is Q, and the total browsing time is T. The average browsing speed is V = Q / T. Assume that the length of a product page in a single browsing process of the shopping platform is Q1, the browsing time of the product page is T1, the average browsing speed of the product is V1, the sections of the page are q1, q2, q3, ... qn, the browsing time of the section is t1, t2, t3, ... tn, and the browsing speed of each section of a product is v1, v2, v3, ... vn. Then the average browsing speed of a product is V1, V1=Q1 / T1, When V ≥ V1, the user is interested in a certain product and the browsing speed of each module of the product is compared. The browsing speed of each module of the product is v1=q1 / t1,v2=q2 / t2,v3=q3 / t3,....,vn=qn / tn, Ratio v1, v2, v3, ...vn to V1 in turn, the ratio sn = vn / V1, If sn = 1, it means that the browsing speed of the article section is consistent with the average browsing speed, and the attribute weight of the product section does not change; If sn>1, it means that the browsing speed of the article section is greater than the average browsing speed, and the attribute weight of the product section is reduced; If sn<1, it means that the browsing speed of the article section is lower than the average browsing speed, and the attribute weight of the product section is increased; Add the newly increased product attributes to the user's habits, or update the product attributes of the user's habits, and put the products that meet the user's habit weights into the second product recommendation area, so as to update the recommended products according to the browsing habits; Track user purchase records, analyze the purchase records, and place products that are logically related to the use of the purchased products, or products of the same category as the purchased products but with significant differences in price, model, and color, in the third product recommendation area, and update the recommended products in the third product recommendation area; Track user evaluation records. When active transaction records occur, combine the transaction product attributes with the evaluation records, and add or subtract product attributes based on the star rating of the evaluation records. When the star rating is low, upgrade the product attribute level, increase the product attribute requirements, update the product requirements to user habits, and update the products in the fourth product recommendation area. The backend analyzes the sales performance of each recommendation area after the products are updated. If the number of sales increases after the products are updated, or the sales remain the same or increase, it means that the product marketing direction matches the user habits and the product marketing direction should be continued. If the number of sales decreases and the average sales also decreases, it means that the marketing recommendation direction is wrong, and the marketing direction should be regenerated and the products in the recommendation area should be updated.
2. According to claim 1, a smart marketing method based on user habits is characterized in that: Collect user habit data and generate user habit self-portraits, including: Collect the user's age, occupation, gender and region information registered by the user when using the service for the first time to obtain the user's basic information; Collect users' search history, browsing history, purchase history, comment history, and sharing history to obtain information about users' behavior habits; Collect user product purchase records, obtain product types, brands, price ranges, colors, sizes and models, and obtain user preference information; Collect users' social comment records and sharing records to obtain information about their social habits; The user's basic information, behavioral habit information, preference habit information and social habit information are integrated to generate a user habit self-portrait.
3. According to claim 1, the intelligent marketing method based on user habits is characterized in that: Perform positioning analysis based on optimized user habit data and generate product marketing recommendation plans, including: Analyze user search records. If the searched product matches the user's personal information, record the product category, style, and applicable season. At the same time, record the user's search time to generate the user's product purchase timing habits. When the timing arrives, add the same searched product to the first product recommendation area. If it does not match the user, then in the next product cycle, 80% of the same type of searched products and 20% of the same type of searched products will be placed in the first product recommendation area. Analyze the user's browsing history, obtain the items that have been browsed for a long time, query the type, brand, price range, color, size and model data of the items, extract similar products from the product library, and add them to the user's second product recommendation area; Analyze user purchase records and query various data of purchased goods. If the purchased goods are consumables, extract the goods and similar functional goods before the consumable period arrives and add them to the user's third product recommendation area; Analyze user review records, obtain user review opinions, extract products similar to those in the opinions, and add them to the user's fourth product recommendation area; Integrate various data on user habits, put the products on the first, second, third and fourth product recommendation areas, and cycle the positions of the product recommendation areas to obtain the product recommendation page.
4. The intelligent marketing method based on user habits according to claim 1, characterized in that: Monitor users' preferred products and send email or SMS notifications when the preferred products meet users' expectations.
5. The intelligent marketing method based on user habits according to claim 1, characterized in that: Record user usage time, infer user usage patterns, arrange product recommendation pages according to the length of user usage time, place products with high weight that users are accustomed to at the forefront of recommendations, and obtain product recommendation pages that match user browsing time.
6. The intelligent marketing method based on user habits according to claim 2, characterized in that: Analyze the user's social habit information, record the users' social followers, retrieve the information of the users based on the intensity of product communication and purchase of the users, extract the information of the intensive product communication and add it to the user's habits, and update the recommended products and put them into the fourth product recommendation area.
7. A system for implementing the intelligent marketing method based on user habits as claimed in claim 1, characterized in that: The system comprises: The habit data collection module is used to collect user habit data and generate user habit self-portraits; The habit data verification module is used to verify the user's habit self-portrait and optimize the user's habit data; Product marketing recommendation module, used to perform positioning analysis based on optimized user habit data and generate product marketing recommendation plans; Product marketing adjustment module, used to update user habit data, adjust product marketing recommendation plans, and obtain new product marketing recommendation plans; The marketing information sending module is used to monitor the products that users are interested in, and send emails or SMS notifications when the products meet the user's expectations; The duration section recommendation module is used to adjust the product recommendation page according to the length of time the user uses the product, put the products with high weight of the user's habitual products at the front of the recommendation, and obtain the user's habitual product browsing marketing recommendation page; The social record analysis module is used to analyze user social habit information, extract information about intensive product exchanges and add it to user habits, and update recommended products and put them into the product recommendation area.
8. An intelligent marketing system based on user habits, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the method according to any one of claims 1 to 6 are implemented.
