A marketing activity page development system and method

Abstract

The present application discloses a system and method for developing a marketing campaign page, which is used to solve the problem in the prior art that the marketing effect of the marketing campaign page cannot be effectively estimated before the marketing campaign page is officially launched, and the actual marketing effect does not meet the expected goal. The development method of the marketing campaign page of the present application includes: obtaining multiple preset page components, setting the multiple preset page components to generate a first preset marketing campaign page; previewing the first preset marketing campaign page to obtain a first preview interface, predicting user behavior on the first preview interface to obtain expected marketing effect data; when the expected marketing effect data is greater than a preset first threshold, the first preview interface is used as the actual marketing campaign page for external release. Through preview and prediction algorithms based on user behavior data, the effect of the marketing campaign can be more accurately estimated. The marketing effect is estimated before the page is officially launched, effectively reducing the risk of marketing strategy failure.

Description

Technical Field

[0001] The present application relates to the technical field of scenic spot navigation methods and systems, and specifically to a system and method for developing a marketing activity page. Background Art

[0002] In the field of computer network technology, a variety of technologies and tools have been developed for creating and editing H5 pages. Among them, the more practical solution is the so-called "What You See Is What You Get" (WYSIWYG) editor. This editor can add, arrange and edit HTML elements in a graphical way, and users can preview the display effect of the interface in the browser after editing. It not only simplifies the process of web page development, but also improves development efficiency, making web page design more intuitive and easy to operate.

[0003] Many WYSIWYG editors on the market, such as Wix and WordPress page editors, provide users with the ability to design and create web pages by dragging and dropping components. These tool technologies are quite mature and have been widely used in website construction and content management platforms. Their core advantage is that users can easily create attractive and functional web pages without programming knowledge.

[0004] Among them, the application publication number is: CN112241257A, and the application name is: A component-based activity page development system and method. The invention discloses a component-based activity page development method, including a data source and an editor. The editor obtains data from the data source. The editor includes a basic component module, a combined component module, a property editing module, an event editing module, a state management module, and a renderer module. The basic component module is used to generate basic elements of the page, the combined component module is used to combine the components, the property editing module is used to edit component properties, the event editing module is used to edit the event occurrence of the overall component, the state management module is used to modify the component state, and the renderer module is used to render the text field into a component.

[0005] The application publication number is: CN114168117A, and the application name is: A designer-based low-code development tool and storage device, including: an application management module, a component management module, a data management module and a designer; the application management module is used to: manage various applications implemented by the designer; the component management module is used to: provide the designer with workflow configuration and automatically generate code for various components; the data management module is used for: interface registration and interface management; the designer is used to: provide different components for designing applications corresponding to different business scenarios.

[0006] The application publication number is: CN11 6974558A, and the application name is: A marketing billing activity H5 page development platform and development method, the platform includes a design and development module and a parsing and running module; the design and development module specifically includes a style designer, an event designer, a process designer, an object editor and a page visual editor; the parsing and running module specifically includes a syntax parser, a code generator and a cloud service adapter.

[0007] The problems with the above patent are: when developing the page, the prediction mechanism of user behavior and marketing effect was not integrated, and the marketing effect could not be effectively estimated before the page was officially launched. At the same time, in the early stage of the marketing activity, the evaluation and optimization mechanism of the page performance was not established, resulting in the actual marketing effect failing to achieve the expected goal. Summary of the invention

[0008] To this end, the present application provides a system and method for developing a marketing campaign page to solve the problem in the prior art that the marketing effect of the marketing campaign page cannot be effectively estimated before it is officially launched, and the actual marketing effect fails to meet the expected goal.

[0009] In order to achieve the above objectives, this application provides the following technical solutions:

[0010] A marketing activity page development system, including a page editing end and a page output end;

[0011] The page editing end is used to obtain multiple preset page components, set the multiple preset page components to generate a first preset marketing activity page, preview the first preset marketing activity page to obtain a first preview interface, perform user behavior prediction on the first preview interface to obtain expected marketing effect data, and use the first preview interface as the actual marketing activity page when the expected marketing effect data is greater than a preset first threshold;

[0012] The page output end is used to publish the actual marketing activity page externally.

[0013] Optionally, the method for generating an actual marketing activity page at the page editing end further includes:

[0014] When the expected marketing effect data is less than a preset first threshold, multiple new preset page components are re-acquired, and the new multiple preset page components are set to generate a second preset marketing activity page, the second preset marketing activity page is previewed to obtain a second preview interface, user behavior prediction is performed on the second preview interface to obtain the expected marketing effect data, and it is determined whether the expected marketing effect data is greater than the preset first threshold; if the expected marketing effect data is less than the preset first threshold, it is iterated in sequence until the expected marketing effect data of the new preview interface is greater than the preset first threshold, and the last new preview interface is used as the actual marketing activity page.

[0015] Optionally, the method for generating the first preset marketing activity page at the page editing end includes:

[0016] Obtain multiple preset page components associated with the marketing campaign in the database, edit the parameters and styles of the preset page components, the preset page components include basic components and custom components, the basic components are composed of one or more page elements such as navigation bar, video, picture, text, form, carousel, music, multimedia player, data acquirer, rich text, share, make a call, coupon, button and pop-up window, drag multiple basic components to create a custom component in a free container, change the content, function timing, hierarchy, size and color of the page elements in the custom component and basic component to complete the editing of the parameters and styles of multiple preset page components, and generate a first preset marketing campaign page after arranging the position and execution order of the preset page components;

[0017] Or, multiple templates consisting of the preset page components are searched in a database through keywords of the marketing campaign, and one of the templates is selected as the first preset marketing campaign page.

[0018] Optionally, the method of previewing the first preset marketing activity page on the page editing end to obtain a first preview interface includes:

[0019] Creating a virtual DOM structure and a real DOM structure of the first preset marketing campaign page, editing the preset page components, and recording changes made to the first preset marketing campaign page in the virtual DOM structure;

[0020] A diff algorithm is used to identify and compare the differences between the virtual DOM structure and the real DOM structure. According to the comparison result, the real DOM structure is updated through the changed part of the virtual DOM structure, and the real DOM structure is rendered and displayed as the first preview interface.

[0021] Optionally, the method for performing user behavior prediction on the first preview interface at the page editing end to obtain expected marketing effect data includes:

[0022] One or more classification tags are selected for the first preview interface, user historical behavior data under different classification tags are obtained, and the user historical behavior data is predicted using a matrix clustering algorithm or a user journey mapping algorithm to obtain expected marketing effect data.

[0023] Optionally, the method of using a matrix clustering algorithm to predict the user historical behavior data to obtain expected marketing effect data includes:

[0024] Obtaining user IDs and web page addresses in user historical behavior data, constructing an association matrix of behavioral user IDs and web page addresses, and adding one or more user historical behavior data including number of visits, dwell time, number of clicks, form filling information, and number of component interactions to the association matrix;

[0025] Taking the association matrix as a starting point, clustering is performed alternately in the user identification dimension and the web page address dimension to extract a dense submatrix, wherein the dense submatrix contains a collection of user identifications and web page addresses with similar behavior patterns;

[0026] The similarity values ​​of the preset page components in the first preview interface and the page components of the web page addresses in the dense sub-matrix are compared, and the expected marketing effect data is obtained after weighted calculation.

[0027] Optionally, the page output end records user behavior data on the actual marketing activity page, and utilizes the user behavior data to iteratively optimize the actual marketing activity page.

[0028] Optionally, the method for iteratively optimizing the actual marketing activity page includes:

[0029] A tracking component is set in the actual marketing activity page, and the tracking component captures and records each event information of registration, login, click, edit, delete, browse, share, favorite and comment in real time;

[0030] Storing and analyzing the event information in a database, and extracting each key data including timestamp, user identification, behavior type, target element, and stay duration;

[0031] After the key data are collected, duplicate data are eliminated, data are classified and data are labeled on the cloud server, the preset page components are evaluated, and after the preset page components are changed on the page editing end, the iterative optimization of the actual marketing activity page is completed.

[0032] Optionally, the page output end obtains one or more feedback information of user satisfaction rating, feedback comments and page design rating on the actual marketing activity page, and utilizes the feedback information to iteratively optimize the actual marketing activity page.

[0033] A method for developing a marketing activity page includes the following steps:

[0034] Acquire multiple preset page components, and set the multiple preset page components to generate a first preset marketing activity page;

[0035] Previewing the first preset marketing activity page to obtain a first preview interface, and performing user behavior prediction on the first preview interface to obtain expected marketing effect data;

[0036] When the expected marketing effect data is greater than a preset first threshold, the first preview interface is used as the actual marketing activity page to be published externally.

[0037] Compared with the prior art, this application has at least the following beneficial effects:

[0038] 1. Improved marketing effectiveness: Through real-time preview and function debugging as well as prediction algorithms based on user behavior data, the effectiveness of marketing campaign pages can be more accurately estimated.

[0039] 2. Risk reduction: By estimating the marketing effect of a page before it is officially launched, the risk of marketing strategy failure can be effectively reduced.

[0040] 3. Decision support: Provide data-supported decision-making tools for marketing teams to help them make more informed marketing strategy choices, collect and analyze user behavior data, help users better understand the performance of their pages, and adjust and optimize page designs in a targeted manner, thereby improving marketing effectiveness.

[0041] 4. Powerful functions: It provides a variety of page elements and marketing-related modules for users to freely drag and combine, such as carousels, forms, etc., to create a rich variety of marketing pages.

[0042] 5. No programming knowledge required: Users do not need to master programming languages ​​such as HTML and CSS. They can create professional H5 marketing pages by simply dragging and setting.

[0043] 6. Flexible and diverse templates: Provide users with a series of preset page templates that users can quickly select and customize without having to design from scratch.

[0044] 7. Continuous updating and optimization: It has good scalability and maintainability, and can continuously update and optimize the system based on user feedback and market demand to provide more functions and better user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more intuitively illustrate the prior art and the present application, exemplary drawings are given below.

[0046] Figure 1A schematic diagram of a system for developing a marketing campaign page according to an embodiment of the present application. DETAILED DESCRIPTION

[0047] The present application is further described below in detail through specific embodiments in conjunction with the accompanying drawings.

[0048] In the description of this application: unless otherwise specified, "plurality" means two or two of them. The terms "first", "second", "third", etc. in this application are intended to distinguish the objects referred to, and do not have special meanings in terms of technical connotations (for example, they should not be understood as emphasizing the importance or order, etc.). Expressions such as "including", "comprising", "having", etc. also mean "not limited to" (certain units, components, materials, steps, etc.).

[0049] Examples, such as Figure 1 As shown, a development system for a marketing activity page includes a page editing end and a page output end;

[0050] The page editing end is used to obtain multiple preset page components, set the multiple preset page components to generate a first preset marketing activity page, preview the first preset marketing activity page to obtain a first preview interface, perform user behavior prediction on the first preview interface to obtain expected marketing effect data, and use the first preview interface as the actual marketing activity page when the expected marketing effect data is greater than a preset first threshold;

[0051] The page output end is used to publish the actual marketing activity page externally.

[0052] The method for generating an actual marketing activity page at the page editing end further includes:

[0053] When the expected marketing effect data is less than a preset first threshold, multiple new preset page components are re-acquired, and the new multiple preset page components are set to generate a second preset marketing activity page, the second preset marketing activity page is previewed to obtain a second preview interface, user behavior prediction is performed on the second preview interface to obtain the expected marketing effect data, and it is determined whether the expected marketing effect data is greater than the preset first threshold; if the expected marketing effect data is less than the preset first threshold, it is iterated in sequence until the expected marketing effect data of the new preview interface is greater than the preset first threshold, and the last new preview interface is used as the actual marketing activity page.

[0054] In an optional implementation manner, the method for generating a first preset marketing activity page at the page editing end includes:

[0055] Acquire multiple preset page components associated with the marketing campaign in a database, edit parameters and styles of the preset page components, and generate a first preset marketing campaign page after arranging positions and execution orders of the preset page components;

[0056] For example, let’s create a product promotion page:

[0057] Choose the pre-set page components you want, such as product display cards, customer testimonials areas, sale timers, and order buttons.

[0058] Edit component parameters and styles: Product display card, choose an attractive picture for each product, write a concise product description, and set appropriate fonts and colors; Customer review area, design an area to display customer reviews, adjust the font size and style to ensure easy reading; Promotion timer, set a countdown timer to show the remaining time of the promotion, use eye-catching colors and large fonts; Order button, design an eye-catching order button, use eye-catching colors and size.

[0059] Arrange component positions and execution order: Place product display cards at the top of the page so that users see them first; place customer review areas below the product display to increase credibility; set promotion timers in prominent locations on the page to increase marketing effectiveness; place order buttons next to each product display card to facilitate users to purchase quickly.

[0060] Or, multiple templates consisting of the preset page components are searched in a database through keywords of the marketing campaign, and one of the templates is selected as the first preset marketing campaign page.

[0061] For example, let’s create a campaign page for a fashion brand to promote a new season’s products:

[0062] Keyword search templates. On the page editing side, use keywords such as "fashion", "product display", and "quarterly promotion" to search for preset templates; determine the priority of keywords. For example, in the "fashion product promotion" scenario, "fashion" is the primary keyword, which is directly related to the core image of the brand; use logical operators (such as "AND" and "OR") to combine multiple keywords. For example, using "fashion AND product display" will return templates that meet the characteristics of both "fashion" and "product display"; try different keyword combinations to obtain different search results. For example, "fashion OR quarterly promotion" will return a wider selection of templates, including fashion and other types of promotional activity templates.

[0063] Based on the search results provided, I selected a template that had a striking visual effect and matched the brand style, for example, a template with a large image slider, a product display area, a customer feedback section, and an order button.

[0064] Customize the template, upload high-definition images of the new quarter's products, set up autoplay and user manual control options, and display detailed information about each new product, including images, descriptions, and prices, in predefined locations; display real customer reviews to increase product credibility; place a clear "Buy Now" button next to each product, with colors and styles that are consistent with the design style of the entire page.

[0065] The method for editing the parameters and styles of the plurality of preset page components includes:

[0066] The first preset page component includes basic components and custom components. The basic components are composed of one or more page elements including navigation bar, video, picture, text, form, carousel, music, multimedia player, data acquirer, rich text, sharing, making phone calls, coupons, buttons and pop-up windows. Drag multiple basic components to create custom components in a free container, change the content, function timing, hierarchy, size and color of the page elements in the custom components and basic components to complete the editing of the parameters and styles of multiple first preset page components.

[0067] Create a sample marketing campaign page and select the required components from the basic component library.

[0068] Navigation bar component: Choose a navigation bar design that fits the overall style of the webpage. To enhance the user experience, add a drop-down menu function to list product categories, colors, font sizes, and styles, all of which can be adjusted according to the page theme.

[0069] Image component: Select high-quality product images and place them in a prominent position on the page. You can adjust the image size and border style to make it coordinate with the overall design of the page.

[0070] Text component: Add descriptive text next to the image to provide detailed information about the product. Choose a font and size that is easy to read, and emphasize important information by using bold or italics.

[0071] Slideshow component: Create a product slideshow that can be set to play automatically or allow users to switch manually, adjust the slideshow speed and transition effects to ensure a smooth user experience.

[0072] Form component: Design a form for user feedback or consultation, including fields such as name, email address, message area, etc. Set up a submit button and make sure the form design is simple and easy to fill out.

[0073] Button component: Add a “Learn More” or “Buy Now” button to each product. Use an eye-catching color and size to make sure the button is prominent and easy for users to operate.

[0074] Custom combinations: Combine the above components into a coordinated and unified custom component, such as combining an image, text description, and purchase button into a single product display unit. Adjust the position of each element to ensure that the layout is beautiful and the information is clearly conveyed.

[0075] In an optional implementation manner, the method of previewing the first preset marketing activity page on the page editing end to obtain a first preview interface includes:

[0076] Creating a virtual DOM structure (Virtual DOM) and a real DOM structure (the English name of the real DOM structure is "Real DOM", which refers to the document object model actually existing on the web page and rendered by the browser, corresponding to the virtual DOM structure) of the first preset marketing campaign page, editing the preset page components, and recording the changes generated by the first preset marketing campaign page in the virtual DOM structure;

[0077] A diff algorithm (difference calculation algorithm) is used to identify and compare the differences between the virtual DOM structure and the real DOM structure. According to the comparison result, the real DOM structure is updated through the changed part of the virtual DOM structure, and the real DOM structure is rendered and displayed as the first preview interface.

[0078] For example, let’s create a promotion page for an e-commerce website:

[0079] Build a virtual DOM structure that includes all preset page components (such as product display area, special offer information, customer reviews, etc.); edit each component of the page, change the layout of the product display area, adjust the font size of the special offer information, or add new customer reviews; all these changes are first made on the virtual DOM without directly affecting the real DOM structure. For example, if the color of the special offer information is changed, this change is first implemented on the virtual DOM, and the diff algorithm is used to compare the differences between the virtual DOM and the real DOM. The algorithm determines the color change of the special offer information; based on the comparison results of the diff algorithm, the system only updates the changed part of the real DOM, that is, the color of the special offer information. After the update is completed, the color change of the special offer information on the page is immediately reflected, and the actual page change effect can be seen.

[0080] See the impact of each change on the page in real time while avoiding unnecessary full-page refreshes, improving development efficiency and performance.

[0081] In an optional implementation manner, the method for performing user behavior prediction on the first preview interface at the page editing end to obtain expected marketing effect data includes:

[0082] One or more classification tags are selected for the first preview interface, user historical behavior data under different classification tags are obtained, and the user historical behavior data is predicted using a matrix clustering algorithm or a user journey mapping algorithm to obtain expected marketing effect data.

[0083] Classification labels can be defined based on different features or attributes to better understand users' historical behaviors and predict expected marketing effects.

[0084] Examples of classification labels are as follows: User attributes: Classify based on user personal information and characteristics, such as gender, age, geographic location, occupation, etc.

[0085] Interests and preferences: Users are divided into different interest groups based on their interests, hobbies, purchase history, browsing history, etc.

[0086] Behavioral history: Users are divided into different behavioral groups based on their behaviors on the website, such as clicks, purchases, comments, sharing, etc.

[0087] Purchase history: Based on their purchase records and purchase frequency, users are classified into different buyer types, such as high-value customers, ordinary customers, etc.

[0088] Channel source: Users are divided into different source groups based on the channel sources from which they visit the website, such as search engines, social media, email, etc.

[0089] Purchase cycle: According to the frequency and periodicity of users’ purchases, users are divided into different purchasing behavior types, such as quarterly buyers, occasional buyers, etc.

[0090] Interaction frequency: Users are divided into active users and inactive users based on how often they interact with the website.

[0091] The classification tags you choose depend on the goals and strategies of the preset marketing activities. By using these classification tags, you can obtain the historical behavior data of users under different classification tags, apply the user journey mapping algorithm, and thus predict and obtain the expected marketing effect data, so as to better conduct marketing and optimization strategies for different user groups.

[0092] Identify and define key indicators that affect the user journey, such as click-through rate (CTR), page browsing time, conversion rate, etc., assign a weight to each indicator, which reflects the relative importance of the indicator to the user journey, use website analysis tools (such as Google Analytics or Baidu Analytics) to collect user historical behavior data, which should include page visits, user stay time, number of clicks, etc., and establish a scoring system to convert user behavior data into a user journey score UJS. Each key indicator corresponds to a score.

[0093] Use the formula to calculate UJS: UJS = ∑(KJI i ×W i )

[0094] KJI i is the actual value of the i-th key user journey indicator, W i is the weight of the indicator.

[0095] Use a linear regression model to analyze the relationship between the user journey score and the expected marketing effect. Based on the linear regression model, predict the marketing effect corresponding to different user journey scores. The marketing effect can be the expected conversion rate, sales, or other related indicators.

[0096] Assume the following indicators and their weights: click-through rate (40%), page browsing time (30%), conversion rate (30%). If a user's historical behavior data is: click-through rate 0.2, page browsing time 300 seconds, conversion rate 0.05, then the user journey score is calculated as follows:

[0097] UJS=(0.2×0.4)+(300×0.3)+(0.05×0.3)

[0098] The scores are input into a linear regression model, which predicts the corresponding marketing effects based on historical data.

[0099] The linear regression model is a statistical method that can be used to predict the relationship between two or more variables. It is usually used to predict the target variable (dependent variable) based on one or more predictor variables (independent variables). In the use scenario of the marketing campaign page, the linear regression model is used to predict the marketing effect (dependent variable) based on the user journey score (independent variable). The expected marketing effect data calculation formula is as follows:

[0100] EME=β0+β1×UJS

[0101] EME (Expected Marketing Effectiveness) represents the score of expected marketing effectiveness data predicted based on the user journey score.

[0102] UJS is the User Journey Score.

[0103] β0 is the intercept of the linear regression model and represents the baseline effect when UJS is zero.

[0104] β1 is the coefficient of UJS in the linear regression model, representing the change in EME for each unit change in UJS.

[0105] If the model estimates β0 = 0.1, β1 = 0.5, then the EME is:

[0106] EME=0.1+0.5×UJS

[0107] The score based on the expected marketing effect data is compared with a preset first threshold. When the score is less than the first threshold, the marketing team further optimizes a preset marketing activity page to improve the expected marketing effect.

[0108] In an optional implementation manner, the method of using a matrix clustering algorithm to predict the user historical behavior data to obtain expected marketing effect data includes:

[0109] Obtaining user identifiers and web page addresses in user historical behavior data, constructing an association matrix of behavioral user identifiers and web page addresses, and adding one or more user historical behavior data including number of visits, dwell time, number of clicks, form filling information, and number of component interactions to the association matrix;

[0110] Taking the association matrix as a starting point, clustering is performed alternately in the user identification dimension and the web page address dimension to extract a dense submatrix, wherein the dense submatrix contains a collection of user identifications and web page addresses with similar behavior patterns;

[0111] The similarity values ​​of the preset page components in the first preview interface and the preset page components of the web page addresses in the dense sub-matrix are compared, and the expected marketing effect data is obtained after weighted calculation.

[0112] For example, the collected user historical behavior data includes: user C visited page M 10 times and stayed for 200 seconds; user D visited page N 8 times and stayed for 220 seconds.

[0113] Create a matrix with rows representing users (C, D), columns representing web pages (M, N), and values ​​representing the number of visits or dwell time.

[0114] Calculate the distance between users, for example, use the visit times and stay time data of user C and user D to calculate the Euclidean distance. Use the hierarchical clustering algorithm to process these data, and it may be found that user C and user D belong to the same cluster.

[0115] Hierarchical clustering is an algorithm for identifying similar patterns without specifying the number of clusters in advance. First, the distance between each user ID is calculated, which is achieved through the Euclidean distance. Starting with each user as a separate cluster, iteratively merge the two closest clusters to build a cluster tree; the dendrogram is cut according to specific criteria (such as distance threshold) to obtain the final clustering result.

[0116] Euclidean distance formula example:

[0117]

[0118] Where u and v are the behavior vectors of two different users.

[0119] Identify dense areas. In the clustering results, identify sets of users and web pages with high visit times and high interaction levels. Based on these sets, build a submatrix that only includes the corresponding users and web pages. Identify pages with high visit times and dwell time for users C and D, and extract the dense submatrix.

[0120] Cosine similarity calculation formula:

[0121]

[0122] A and B are the behavior vectors of two web page addresses that need to be compared.

[0123] According to the preset page component characteristics of the first preview interface, determine the indicators for evaluating similarity, such as layout similarity, color scheme similarity, interactive element similarity, etc.; for each similarity indicator, use the cosine similarity formula to calculate the single similarity score between the first preview interface and each web page in the dense submatrix. Determine the weights of each similarity indicator based on the degree of influence it has on user behavior. For example, assuming that layout similarity has a greater impact on user experience, give it a higher weight; multiply each similarity score by its corresponding weight, and then sum them up to get the total similarity score; perform correlation analysis on the weighted total similarity score and historical marketing effect data, and predict the expected marketing effect data through a linear regression model.

[0124] In an optional implementation manner, the page output end records user behavior data on the actual marketing activity page, and utilizes the user behavior data to iteratively optimize the actual marketing activity page.

[0125] The method for iteratively optimizing the actual marketing activity page includes:

[0126] A tracking component is set in the actual marketing activity page, and the tracking component captures and records each event information of registration, login, click, edit, delete, browse, share, favorite and comment in real time;

[0127] Storing and analyzing the event information in a database, and extracting each key data including timestamp, user identification, behavior type, target element, and stay duration;

[0128] After the key data are collected, duplicate data are eliminated, data are classified and data are labeled on the cloud server, the preset page components are evaluated, and after the preset page components are changed on the page editing end, the iterative optimization of the actual marketing activity page is completed.

[0129] In an optional implementation manner, the page output end obtains one or more feedback information of user satisfaction rating, feedback comments and page design rating on the actual marketing activity page, and uses the feedback information to iteratively optimize the actual marketing activity page.

[0130] This application collects user visit, click and other behavioral data, performs data analysis on user behavior, such as conversion rate, dwell time, bounce rate, etc., and then presents these data in an intuitive way (such as various charts) to help the marketing team optimize page design to improve marketing effectiveness. The marketing team may not have professional design capabilities, so it is critical to provide pre-designed templates. These templates cover various marketing scenarios (such as product display, event promotion, etc.), and users can quickly create high-quality pages by selecting and customizing these templates. Use an intuitive graphical interface, import easy-to-understand editing concepts (for example: drag and drop), provide clear guidance and feedback, simplify operation processes, optimize loading speed, etc. The front-end page editor requires the support of back-end services, including user account management, template and project saving, preview and publishing, etc. Use java to provide an API interface to interact with the front end to save or edit the marketing page designed by the user.

[0131] The technical implementation of this application will use front-end foundations such as HTML, CSS, JavaScript, etc., use the Vue front-end framework to manage the user interface and interaction, use Java back-end technology, and databases such as Redis, MongoDB, MySQL to store data, and combine various Web APIs for interaction and function implementation.

[0132] The data collection, storage and analysis of this application will comply with relevant data security and privacy regulations when designed and implemented.

[0133] The technical features of the embodiments can be combined arbitrarily (as long as there is no contradiction in the combination of these technical features). In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described; these embodiments that are not explicitly written should also be considered to be within the scope of this specification.

[0134] The above is a relatively specific and detailed description of the present application through general description and specific embodiments. It should be understood that, based on the technical concept of the present application, several conventional adjustments or further innovations can be made to these specific embodiments; however, as long as they do not deviate from the technical concept of the present application, the technical solutions obtained by these conventional adjustments or further innovations also fall within the scope of protection of the claims of the present application.

Claims

1. A system for developing a marketing activity page, characterized in that: Including page editing end and page output end; The page editing end is used to obtain multiple preset page components, set the multiple preset page components to generate a first preset marketing activity page, preview the first preset marketing activity page to obtain a first preview interface, perform user behavior prediction on the first preview interface to obtain expected marketing effect data, and use the first preview interface as the actual marketing activity page when the expected marketing effect data is greater than a preset first threshold; The page output terminal is used to publish the actual marketing activity page externally; The method for performing user behavior prediction on the first preview interface at the page editing end to obtain expected marketing effect data includes: Selecting one or more classification tags for the first preview interface, obtaining user historical behavior data under different classification tags, and using a matrix clustering algorithm to predict the user historical behavior data to obtain expected marketing effect data; The method of using a matrix clustering algorithm to predict the user's historical behavior data to obtain expected marketing effect data includes: Obtaining user IDs and web page addresses in user historical behavior data, constructing an association matrix of behavioral user IDs and web page addresses, and adding one or more user historical behavior data including number of visits, dwell time, number of clicks, form filling information, and number of component interactions to the association matrix; Taking the association matrix as a starting point, clustering is performed alternately in the user identification dimension and the web page address dimension to extract a dense submatrix, wherein the dense submatrix contains a collection of user identifications and web page addresses with similar behavior patterns; Comparing similarity values ​​between the preset page components in the first preview interface and the page components of the web page addresses in the dense submatrix, and obtaining expected marketing effect data after weighted calculation; Methods for obtaining expected marketing effect data after weighted calculation include: For each similarity index, the cosine similarity formula is used to calculate the single similarity score between the first preview interface and each web page in the dense submatrix. The weights of each similarity index are determined according to the degree of influence of each similarity index on user behavior. Each similarity score is multiplied by its corresponding weight, and then the sum is calculated to obtain the total similarity score. The weighted total similarity score is correlated with the historical marketing effect data, and the expected marketing effect data is predicted through a linear regression model. The similarity indicators include: layout similarity, color scheme similarity, and interactive element similarity.

2. A marketing activity page development system according to claim 1, characterized in that: The method for generating an actual marketing activity page at the page editing end further includes: When the expected marketing effect data is less than a preset first threshold, multiple new preset page components are re-acquired, and the new multiple preset page components are set to generate a second preset marketing activity page, the second preset marketing activity page is previewed to obtain a second preview interface, user behavior prediction is performed on the second preview interface to obtain the expected marketing effect data, and it is determined whether the expected marketing effect data is greater than the preset first threshold; if the expected marketing effect data is less than the preset first threshold, it is iterated in sequence until the expected marketing effect data of the new preview interface is greater than the preset first threshold, and the last new preview interface is used as the actual marketing activity page.

3. A marketing activity page development system according to claim 1, characterized in that: The method for generating a first preset marketing activity page at the page editing end includes: Obtain multiple preset page components associated with the marketing campaign in the database, edit the parameters and styles of the preset page components, the preset page components include basic components and custom components, the basic components are composed of one or more page elements such as navigation bar, video, picture, text, form, carousel, music, multimedia player, data acquirer, rich text, share, make a call, coupon, button and pop-up window, drag multiple basic components to create a custom component in a free container, change the content, function timing, hierarchy, size and color of the page elements in the custom component and basic component to complete the editing of the parameters and styles of multiple preset page components, and generate a first preset marketing campaign page after arranging the position and execution order of the preset page components; Or, multiple templates consisting of the preset page components are searched in a database through keywords of the marketing campaign, and one of the templates is selected as the first preset marketing campaign page.

4. A marketing activity page development system according to claim 1, characterized in that: The method for obtaining a first preview interface by previewing the first preset marketing activity page at the page editing end includes: Creating a virtual DOM structure and a real DOM structure of the first preset marketing campaign page, editing the preset page components, and recording changes made to the first preset marketing campaign page in the virtual DOM structure; A diff algorithm is used to identify and compare the differences between the virtual DOM structure and the real DOM structure. According to the comparison result, the real DOM structure is updated through the changed part of the virtual DOM structure, and the real DOM structure is rendered and displayed as the first preview interface.

5. A marketing activity page development system according to claim 1, characterized in that: The page output end records the user behavior data on the actual marketing activity page, and iteratively optimizes the actual marketing activity page using the user behavior data.

6. A marketing activity page development system according to claim 5, characterized in that: The method for iteratively optimizing the actual marketing activity page includes: A tracking component is set in the actual marketing activity page, and the tracking component captures and records each event information of registration, login, click, edit, delete, browse, share, favorite and comment in real time; Storing and analyzing the event information in a database, and extracting each key data including timestamp, user identification, behavior type, target element, and stay duration; After the key data are collected, duplicate data are eliminated, data are classified and data are labeled on the cloud server, the preset page components are evaluated, and after the preset page components are changed on the page editing end, the iterative optimization of the actual marketing activity page is completed.

7. A marketing activity page development system according to claim 1, characterized in that: The page output end obtains one or more feedback information of user satisfaction rating, feedback comments and page design rating on the actual marketing activity page, and uses the feedback information to iteratively optimize the actual marketing activity page.

8. A method for developing a marketing activity page, characterized in that: The steps include: Acquire multiple preset page components, and set the multiple preset page components to generate a first preset marketing activity page; Previewing the first preset marketing activity page to obtain a first preview interface, and performing user behavior prediction on the first preview interface to obtain expected marketing effect data; When the expected marketing effect data is greater than a preset first threshold, the first preview interface is used as the actual marketing activity page to be published externally; The method for performing user behavior prediction on the first preview interface to obtain expected marketing effect data includes: Selecting one or more classification tags for the first preview interface, obtaining user historical behavior data under different classification tags, and using a matrix clustering algorithm to predict the user historical behavior data to obtain expected marketing effect data; The method of using a matrix clustering algorithm to predict the user's historical behavior data to obtain expected marketing effect data includes: Obtaining user IDs and web page addresses in user historical behavior data, constructing an association matrix of behavioral user IDs and web page addresses, and adding one or more user historical behavior data including number of visits, dwell time, number of clicks, form filling information, and number of component interactions to the association matrix; Taking the association matrix as a starting point, clustering is performed alternately in the user identification dimension and the web page address dimension to extract a dense submatrix, wherein the dense submatrix contains a collection of user identifications and web page addresses with similar behavior patterns; Comparing similarity values ​​between the preset page components in the first preview interface and the page components of the web page addresses in the dense submatrix, and obtaining expected marketing effect data after weighted calculation; Methods for obtaining expected marketing effect data after weighted calculation include: For each similarity index, the cosine similarity formula is used to calculate the single similarity score between the first preview interface and each web page in the dense submatrix. The weights of each similarity index are determined according to the degree of influence of each similarity index on user behavior. Each similarity score is multiplied by its corresponding weight, and then the sum is calculated to obtain the total similarity score. The weighted total similarity score is correlated with the historical marketing effect data, and the expected marketing effect data is predicted through a linear regression model. The similarity indicators include: layout similarity, color scheme similarity, and interactive element similarity.

Images