A real-time calculation method for attribution analysis of marketing conversions

Abstract

This invention discloses a real-time calculation method for attribution analysis of marketing conversions, belonging to the field of attribution analysis technology. The method includes: collecting user behavior data through a data tracking SDK, sequentially processing the data via log files and message queues, consuming the data in real time using stream processing technology and performing cleaning, dimension association, path identification and segmentation based on session identifiers and timestamps, calling an attribution model to calculate the contribution of each touchpoint, and finally displaying the results through a microservice interface and a front-end page. This invention can improve the accuracy and efficiency of attribution analysis, achieving comprehensive and in-depth marketing conversion path analysis.

Description

Technical Field

[0001] This invention relates to the field of attribution analysis technology, and in particular to a real-time calculation method for attribution analysis of marketing conversion. Background Technology

[0002] Attribution analysis is a key technology for evaluating the contribution of different marketing channels and touchpoints to user conversion, and it is widely used in e-commerce, advertising, and user growth. Its core lies in quantifying the weighting of each marketing touchpoint on target events (such as order payment and registration conversion) by tracking the complete behavioral journey of users from their first contact to final conversion, thereby supporting marketing budget optimization and strategy adjustment.

[0003] Current attribution analysis techniques primarily employ a batch offline processing model. A typical approach involves collecting user behavior logs via a tracking SDK, periodically transferring these log files to a data warehouse, executing a T+1 scheduled task for data cleaning and path reconstruction based on the data warehouse, and calculating the contribution of each touchpoint using pre-defined attribution rules to generate reports. Regarding path identification, traditional methods typically only sort user behavior based on timestamps, forming a linear time series as the conversion path. In terms of attribution model application, a single first or last touchpoint model is often used for contribution allocation. Finally, result presentation relies mainly on fixed-period statistical reports.

[0004] However, with the increasing demand for real-time marketing decisions, the aforementioned technical solutions suffer from the following shortcomings: First, the offline batch processing mode leads to delays in attribution results, failing to support real-time strategy adjustments by operations personnel. Second, simple sorting based on global timestamps ignores the complex scenarios of multiple user sessions, making it difficult to accurately reconstruct the true business conversion path and lacking an effective mechanism for splitting multiple conversion events along a single path, resulting in insufficient path identification accuracy. Third, the attribution model is rigid and singular, making it difficult to flexibly adapt to the dynamic configuration needs of different marketing scenarios. Fourth, the high coupling between various links in the data collection and result display chain, coupled with the lack of an efficient data flow mechanism, leads to low overall processing efficiency. Fifth, the visualization capabilities are limited, making it difficult to intuitively present the flow of conversion paths and the time trend changes of attribution results, affecting decision-making efficiency. These shortcomings directly limit the effectiveness of attribution analysis in real-time refined operations. Summary of the Invention

[0005] The technical problem to be solved by this invention is to address the shortcomings of existing technologies, specifically by providing a real-time calculation method for attribution analysis of marketing conversion, as detailed below: 1) In a first aspect, the present invention provides a real-time calculation method for attribution analysis of marketing conversion, the specific technical solution of which is as follows: The tracking SDK module collects user behavior data for any delivered advertisement and generates a raw behavior sequence containing user identity, session identity, timestamp and event code, wherein the event code contains the encoding of the target event and the event to be attributed. The original behavior sequence is sequentially written into a log file by the log collection module, and then read from the log file and pushed to a message queue to form a data stream to be processed that is consistent with the order of the original behavior sequence. The attribution analysis calculation module consumes the data stream to be processed from the message queue in real time, performs data cleaning and dimension correlation on the data stream to be processed, and obtains normalized behavioral data with supplementary event business attributes. Based on the session identifier and the timestamp, the normalized behavior data is path identified, behavior sequences containing the target event are filtered, and the path is split according to the target event to obtain a set of conversion paths ending with a single target event. The attribution model is invoked to calculate each path in the conversion path set, and the conversion contribution of the target event is allocated to the event to be attributed according to the allocation rules of the attribution model. An attribution result containing the contribution of each event to be attributed is generated. Based on the attribution result, any ad that has been placed is analyzed. The attribution results are obtained through the data display module, stored in the database, and visualized through microservice interfaces and front-end pages.

[0006] The beneficial effects of the real-time calculation method for attribution analysis of marketing conversion provided by this invention are as follows: By using session identifiers and timestamps to accurately identify and segment normalized behavioral data, a set of conversion paths ending at a single target event is obtained. The attribution model is then invoked to accurately allocate the conversion contribution of the target event to each event to be attributed according to the allocation rules. This enables more accurate identification and quantification of the contribution of different marketing touchpoints to user conversion, reducing misjudgments and biases. The sequential flow mechanism of log files and message queues combined with the real-time computing capabilities of stream processing technology significantly improves data processing speed and attribution analysis efficiency, shortens the analysis cycle, and rapidly responds to marketing optimization needs. Furthermore, the use of the event tracking SDK module to collect user behavior data from advertising channels and correlate dimensions to supplement event business attributes enables comprehensive capture and in-depth analysis of marketing influencing factors, revealing potential patterns in user conversion paths and providing more comprehensive information support for advertising effectiveness evaluation and marketing decisions.

[0007] Based on the above solution, the present invention can be further improved as follows.

[0008] Furthermore, the process of forming a data stream to be processed that is consistent with the original behavioral sequence specifically includes: The original sequence of behaviors is written sequentially to a log file; The Flume Agent monitors changes to the log files, reads them line by line, pushes the data to the Kafka cluster, and uses the user's identity as the partition key to ensure the data order of the same user.

[0009] Furthermore, the process of obtaining the set of transformation paths ending with a single target event is as follows: The normalized behavior data is grouped according to the session identifier and the user identity identifier; The grouped data is sorted in ascending order according to the timestamps to form user behavior paths; Identify the target event in the user behavior path, and use the target event as an anchor point to trace back the time range within the attribution window period to form the conversion path set. When a single path contains multiple target events, the path is split into multiple independent conversion paths with each target event as the endpoint.

[0010] Furthermore, the attribution models include the first contact attribution model, the last contact attribution model, the linear attribution model, and the location attribution model; The first-touch attribution model assigns 100% of the conversion contribution to the first event to be attributed in the path; The last-touch attribution model assigns 100% of the conversion contribution to the event to be attributed at the end of the path. The linear attribution model distributes the transformation contribution equally among all attributable events in the path.

[0011] 2) Secondly, the present invention also provides a real-time calculation system for attribution analysis of marketing conversion, the specific technical solution of which is as follows: The event tracking SDK module is used to: collect user behavior data for any delivered advertisement, and generate a raw behavior sequence containing user identity, session identity, timestamp and event code, wherein the event code contains the encoding of the target event and the event to be attributed; The log collection module is used to: write the original behavior sequence sequentially into a log file, read from the log file and push to a message queue to form a data stream to be processed that is consistent with the original behavior sequence. The attribution analysis calculation module is used to: consume the data stream to be processed from the message queue in real time; perform data cleaning and dimension correlation on the data stream to be processed to obtain normalized behavioral data supplemented with event business attributes; perform path identification on the normalized behavioral data based on the session identifier and the timestamp; filter behavioral sequences containing the target event; and split the path according to the target event to obtain a set of conversion paths ending with a single target event; call the attribution model to calculate each path in the set of conversion paths; allocate the conversion contribution of the target event to the event to be attributed according to the allocation rules of the attribution model; generate attribution results containing the contribution of each event to be attributed; and analyze any ad that has been placed based on the attribution results. The data display module is used to: obtain the attribution results, store the attribution results in the database, and display them visually through microservice interfaces and front-end pages.

[0012] Based on the above solution, the present invention can be further improved as follows.

[0013] Furthermore, the process of forming a data stream to be processed that is consistent with the original behavioral sequence specifically includes: The original sequence of behaviors is written sequentially to a log file; The Flume Agent monitors changes to the log files, reads them line by line, pushes the data to the Kafka cluster, and uses the user's identity as the partition key to ensure the data order of the same user.

[0014] Furthermore, the process of obtaining the set of transformation paths ending with a single target event is as follows: The normalized behavior data is grouped according to the session identifier and the user identity identifier; The grouped data is sorted in ascending order according to the timestamps to form user behavior paths; Identify the target event in the user behavior path, and use the target event as an anchor point to trace back the time range within the attribution window period to form the conversion path set. When a single path contains multiple target events, the path is split into multiple independent conversion paths with each target event as the endpoint.

[0015] Furthermore, the attribution models include the first contact attribution model, the last contact attribution model, the linear attribution model, and the location attribution model; The first-touch attribution model assigns 100% of the conversion contribution to the first event to be attributed in the path; The last-touch attribution model assigns 100% of the conversion contribution to the event to be attributed at the end of the path. The linear attribution model distributes the transformation contribution equally among all attributable events in the path.

[0016] 3) In a third aspect, the present invention also provides an electronic device, the electronic device including a processor coupled to a memory, the memory storing at least one computer program, the at least one computer program being loaded and executed by the processor to enable the electronic device to perform any of the above methods.

[0017] 4) In a fourth aspect, the present invention also provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to enable a computer to perform any of the above methods.

[0018] It should be noted that the beneficial effects of the technical solutions of the second to fourth aspects of the present invention and their corresponding possible implementations can be found in the above description of the technical effects of the first aspect and its corresponding possible implementations, and will not be repeated here. Attached Figure Description

[0019] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a flowchart illustrating a real-time calculation method for attribution analysis of marketing conversion according to an embodiment of the present invention. Figure 2 This is an example diagram of attribution analysis configuration parameters for a real-time calculation method of attribution analysis for marketing conversion according to an embodiment of the present invention. Figure 3 This is a sample image of user behavior path identification data, representing a real-time calculation method for attribution analysis of marketing conversion according to an embodiment of the present invention. Figure 4 This is a schematic diagram of path decomposition for a real-time calculation method of attribution analysis for marketing conversion according to an embodiment of the present invention. Figure 5 This is a schematic diagram illustrating the calculation rules of a real-time calculation method for attribution analysis of marketing conversion according to an embodiment of the present invention. Figure 6 This is a schematic diagram of the attribution analysis results of a real-time calculation method for attribution analysis of marketing conversion according to an embodiment of the present invention. Figure 7 This is a schematic diagram of the architecture of a real-time calculation method for attribution analysis of marketing conversion according to an embodiment of the present invention. Figure 8 This is a structural framework diagram of an electronic device according to the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0021] like Figure 1 as well as Figure 7 As shown, an embodiment of the present invention provides a real-time calculation method for attribution analysis of marketing conversion, comprising the following steps: S1, collect user behavior data for any delivered advertisement through the event tracking SDK module, and generate an original behavior sequence containing user identity identifier, session identifier, timestamp and event code, wherein the event code contains the encoding of the target event and the event to be attributed; S2, the original behavior sequence is sequentially written into a log file by the log collection module, and read from the log file and pushed to a message queue to form a data stream to be processed that is consistent with the order of the original behavior sequence; S3, the attribution analysis calculation module consumes the data stream to be processed from the message queue in real time, performs data cleaning and dimension correlation on the data stream to be processed, and obtains normalized behavioral data with supplementary event business attributes; S4. Based on the session identifier and the timestamp, perform path identification on the normalized behavior data, filter out behavior sequences containing the target event, and split the path according to the target event to obtain a set of conversion paths ending with a single target event. S5, call the attribution model to calculate each path in the conversion path set, allocate the conversion contribution of the target event to the event to be attributed according to the allocation rules of the attribution model, generate attribution results containing the contribution of each event to be attributed, and analyze any ad that has been placed based on the attribution results. S6. Obtain the attribution results through the data display module, store the attribution results in the database, and display them visually through the microservice interface and the front-end page.

[0022] The beneficial effects of the real-time calculation method for attribution analysis of marketing conversion provided by this invention are as follows: By using session identifiers and timestamps to accurately identify and segment normalized behavioral data, a set of conversion paths ending at a single target event is obtained. The attribution model is then invoked to accurately allocate the conversion contribution of the target event to each event to be attributed according to the allocation rules. This enables more accurate identification and quantification of the contribution of different marketing touchpoints to user conversion, reducing misjudgments and biases. The sequential flow mechanism of log files and message queues combined with the real-time computing capabilities of stream processing technology significantly improves data processing speed and attribution analysis efficiency, shortens the analysis cycle, and rapidly responds to marketing optimization needs. Furthermore, the use of the event tracking SDK module to collect user behavior data from advertising channels and correlate dimensions to supplement event business attributes enables comprehensive capture and in-depth analysis of marketing influencing factors, revealing potential patterns in user conversion paths and providing more comprehensive information support for advertising effectiveness evaluation and marketing decisions.

[0023] In another embodiment of this solution, when collecting user behavior data for any delivered advertisement through the tracking SDK module, the tracking SDK module provides a unified tracking interface for the web, mobile, and mini-program ends. When a user interacts with the delivered advertisement, the tracking SDK captures the user behavior in real time and generates a raw behavior sequence. Each record in the raw behavior sequence contains four core attributes: user identity, session identity, timestamp, and event code. The event code is used to identify the behavior type and distinguish between target events and events to be attributed (target events refer to the final conversion behavior configured by the operations personnel that needs to be analyzed, such as payment orders; events to be attributed refer to the pre-marketing touchpoint behaviors configured by the operations personnel that need to be evaluated for contribution, such as browsing product details or clicking on advertisements). The raw behavior sequence also includes the collection time and other business attribute fields for subsequent dimension association.

[0024] It should be noted that the event tracking SDK refers to a software development kit (SDK) integrated on the web, mobile (App) and mini-program platforms. It is used to automatically capture and report user behavior data when users interact with the displayed advertisements or application interfaces.

[0025] In another embodiment of this solution, when the log collection module processes the original behavior sequence, the log collection module sequentially writes the original behavior sequence uploaded by the event tracking SDK into the local log file through the log receiving service to ensure that the data is not lost. The Flume Agent continuously monitors the changes in the log file, reads the log content line by line, and pushes the data to the Kafka message queue in real time. During the push process, the user identity is used as the partition key to ensure that the behavior data belonging to the same user maintains the same order as the original behavior sequence in the message queue, thereby forming a data stream to be processed.

[0026] The log receiving service is a service component used to receive user behavior data uploaded by the event tracking SDK and write it sequentially to local log files for persistent storage. The log receiving service is located between the SDK and the log files, and plays the role of data buffering and disk writing, ensuring that the original behavior sequence is not lost before being transmitted to Kafka.

[0027] Flume Agent is a component used to monitor changes in log files and push data to Kafka. It is a data collection and execution unit within the Apache Flume framework. The log collection module first collects the data from the monitoring points into a file, and then from the file into Kafka. Flume Agent is responsible for forwarding the data from the file to the message, reading the log file content written by the log receiving service line by line, and transmitting the data to the Kafka cluster in real time.

[0028] In another embodiment of this solution, when the attribution analysis calculation module consumes the data stream to be processed from the message queue in real time, the attribution analysis calculation module uses stream processing technology (such as Flink) to process the data in real time, performs data cleaning operations to remove duplicate and invalid records, associates event codes with dimension tables to supplement event names, and associates user identity identifiers with user profile tables to supplement business attributes such as channel source and user level, to obtain normalized behavioral data (the normalized behavioral data refers to a set of structured behavioral records that have been standardized in format, deduplicated, and supplemented with business attribute descriptions).

[0029] In another embodiment of this solution, when performing path identification on normalized behavioral data based on session identifiers and timestamps, the attribution analysis calculation module groups the data according to session identifiers and user identity identifiers, aggregates the behavior of the same user within the same session, sorts each group of data in ascending order by timestamp to form user behavior paths, filters out sequences containing target events from the user behavior paths, and uses each target event as an anchor point to trace back the time range of the attribution window to extract the path. When a single path contains multiple target events, the path is split into multiple independent conversion path sets according to the target events, with each path ending at a single target event.

[0030] The target event is the unique final conversion behavior specified by operations personnel when configuring attribution analysis. It serves as the endpoint anchor for attribution calculations, and its conversion contribution must be allocated to the attribution event according to the attribution model's allocation rules. Examples include core business actions completed by users, such as payment orders (PayOrder).

[0031] In another embodiment of this solution, when the attribution model is invoked to calculate each path in the conversion path set, the attribution analysis calculation module executes the corresponding allocation rules according to the attribution model type (including first touchpoint attribution, last touchpoint attribution, linear attribution, and location attribution) configured by the operations personnel. (The allocation rules refer to the preset proportion logic used by each attribution model to allocate the conversion contribution of the target event to the event to be attributed. For example, the location attribution model allocates 40% to the first event to be attributed, 40% to the last event to be attributed, and the remaining 20% ​​is evenly allocated to the middle events to be attributed.) After calculation, an attribution result containing the contribution of each event to be attributed is generated. This attribution result is used to analyze the actual conversion contribution of different marketing touchpoints in any ad that has been placed.

[0032] In another embodiment of this solution, when obtaining attribution results through the data display module, the module stores the calculated attribution results in a result table of a MySQL database. The microservice layer provides a RESTful API interface for the front-end to call. The front-end page displays the contribution percentage of each event to be attributed using a pie chart, the user conversion path flow using a Sankey diagram, and the time trend change of contribution using a line graph. At the same time, the effective conversion click-through rate, the total number of target events, and the effective conversion clicks are presented on the same page with the contribution, realizing a visual display of the attribution results.

[0033] Furthermore, the process of forming a data stream to be processed that is consistent with the original behavioral sequence specifically includes: The original sequence of behaviors is written sequentially to a log file; The Flume Agent monitors changes to the log files, reads them line by line, pushes the data to the Kafka cluster, and uses the user's identity as the partition key to ensure the data order of the same user.

[0034] Furthermore, the process of obtaining the set of transformation paths ending with a single target event is as follows: The normalized behavior data is grouped according to the session identifier and the user identity identifier; The grouped data is sorted in ascending order according to the timestamps to form user behavior paths; Identify the target event in the user behavior path, and use the target event as an anchor point to trace back the time range within the attribution window period to form the conversion path set. When a single path contains multiple target events, the path is split into multiple independent conversion paths with each target event as the endpoint.

[0035] Furthermore, the attribution models include the first contact attribution model, the last contact attribution model, the linear attribution model, and the location attribution model; The first-touch attribution model assigns 100% of the conversion contribution to the first event to be attributed in the path; The last-touch attribution model assigns 100% of the conversion contribution to the event to be attributed at the end of the path. The linear attribution model distributes the transformation contribution equally among all attributable events in the path.

[0036] Example 1: The collected user behavior data includes core attributes: timestamp, cookie, session, event code, collection time, and other business attributes.

[0037] Operations personnel can configure attribution analysis in real time, including: target event (only one item), events to be attributed (multiple items), filter criteria, attribution model, attribution window, and data time range. For example... Figure 2 As shown.

[0038] like Figure 3 as well as Figure 4 As shown, user behavior path identification is performed through session. After the same session is sorted in ascending order of timestamp, it is identified as the path of a user's one complete visit from beginning to end.

[0039] The target event is: PayOrder The first touchpoint event in the pending attribution events: commodityDetail The last touchpoint event in the pending attribution events: buyNow When there are multiple target events in a path, the path is split according to the target event path.

[0040] like Figure 5 As shown, the attribution models include four types: first-contact attribution, last-contact attribution, linear attribution, and location attribution. The calculation rules for each model are as follows: First Touchpoint Attribution: During the retrospective period, the first touchpoint's conversion contribution is allocated 100%. All other touchpoints are allocated 0%.

[0041] Last touchpoint attribution: During the retrospective period, the last touchpoint is assigned 100% of the conversion contribution, and the remaining touchpoints are assigned 0%.

[0042] Linear attribution: During the retrospective period, a conversion is evenly distributed across all touchpoints. For example, if a user's conversion involves 5 touchpoint events, then each of the 5 touchpoints is assigned 20%.

[0043] Location attribution: During the retrospective period, the first contact was allocated 40%, the last contact was allocated 40%, and the remaining intermediate contacts were allocated an average of 20%.

[0044] like Figure 6As shown, the attribution analysis results include: total clicks, effective conversion clicks, effective conversion click-through rate, total number of target events, and contribution.

[0045] This invention improves the accuracy of attribution analysis: by introducing advanced algorithms (such as deep learning, ensemble learning, and causal inference) and combining them with multi-dimensional data fusion technology, it can more accurately identify and quantify the contribution of different factors to a particular event or outcome. This helps reduce misjudgments and biases, and improves the accuracy and reliability of attribution analysis.

[0046] This invention improves the efficiency of attribution analysis: by optimizing data processing workflows, utilizing high-performance computing resources, and introducing automated and intelligent tools, it significantly enhances the speed and efficiency of attribution analysis. This helps shorten analysis cycles, quickly respond to business needs, and provide timely and effective support for decision-making.

[0047] This invention enables comprehensive and in-depth attribution analysis: it integrates data from different channels, formats, and dimensions, including user behavior data, market environment data, and social media sentiment analysis, thereby achieving a comprehensive capture and in-depth analysis of influencing factors. This helps to reveal the underlying patterns and trends hidden behind complex data, providing more comprehensive information support for business decisions.

[0048] In the above embodiments, although the steps are numbered S1, S2, etc., they are only specific embodiments given by the present invention. Those skilled in the art can adjust the execution order of S1, S2, etc. according to the actual situation, and these situations are also within the protection scope of the present invention. It can be understood that in some embodiments, some or all of the above embodiments may be included.

[0049] This invention also provides a real-time calculation system for attribution analysis of marketing conversion, the specific technical solution of which is as follows: The event tracking SDK module is used to: collect user behavior data for any delivered advertisement, and generate a raw behavior sequence containing user identity, session identity, timestamp and event code, wherein the event code contains the encoding of the target event and the event to be attributed; The log collection module is used to: write the original behavior sequence sequentially into a log file, read from the log file and push to a message queue to form a data stream to be processed that is consistent with the original behavior sequence. The attribution analysis calculation module is used to: consume the data stream to be processed from the message queue in real time; perform data cleaning and dimension correlation on the data stream to be processed to obtain normalized behavioral data supplemented with event business attributes; perform path identification on the normalized behavioral data based on the session identifier and the timestamp; filter behavioral sequences containing the target event; and split the path according to the target event to obtain a set of conversion paths ending with a single target event; call the attribution model to calculate each path in the set of conversion paths; allocate the conversion contribution of the target event to the event to be attributed according to the allocation rules of the attribution model; generate attribution results containing the contribution of each event to be attributed; and analyze any ad that has been placed based on the attribution results. The data display module is used to: obtain the attribution results, store the attribution results in the database, and display them visually through microservice interfaces and front-end pages.

[0050] It should be noted that the beneficial effects of the real-time calculation system for attribution analysis of marketing conversion provided in the above embodiments are the same as those of the real-time calculation method for attribution analysis of marketing conversion described above, and will not be repeated here. Furthermore, the system provided in the above embodiments is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the system can be divided into different functional modules according to the actual situation to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process is detailed in the method embodiments, and will not be repeated here.

[0051] like Figure 8 As shown, an electronic device 300 according to an embodiment of the present invention includes a processor 320 coupled to a memory 310. The memory 310 stores at least one computer program 330, which is loaded and executed by the processor 320 to enable the electronic device 300 to implement any of the above-mentioned methods. Specifically: The electronic device 300 can vary considerably due to differences in configuration or performance. It may include one or more processors 320 (Central Processing Units, CPUs) and one or more memories 310. The one or more memories 310 store at least one computer program 330, which is loaded and executed by the one or more processors 320 to enable the electronic device 300 to implement the real-time calculation method for attribution analysis of marketing conversion provided in the above embodiment. Of course, the electronic device 300 may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. The electronic device 300 may also include other components for implementing device functions, which will not be elaborated upon here.

[0052] An embodiment of the present invention provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to enable a computer to implement any of the above-described methods.

[0053] Alternatively, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, and an optical data storage device, etc.

[0054] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform any of the methods described above.

[0055] It should be noted that the terms "first" and "second" in the specification and claims of this application are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0056] Those skilled in the art will recognize that this invention can be implemented as a system, method, or computer program product. Therefore, this disclosure can be specifically implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, the invention can also be implemented as a computer program product contained in one or more computer-readable media, which includes computer-readable program code.

[0057] Any combination of one or more computer-readable media may be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0058] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A real-time calculation method for attribution analysis of marketing conversion, characterized in that, include: The tracking SDK module collects user behavior data for any delivered advertisement and generates a raw behavior sequence containing user identity, session identity, timestamp and event code, wherein the event code contains the encoding of the target event and the event to be attributed. The original behavior sequence is sequentially written into a log file by the log collection module, and then read from the log file and pushed to a message queue to form a data stream to be processed that is consistent with the order of the original behavior sequence. The attribution analysis calculation module consumes the data stream to be processed from the message queue in real time, performs data cleaning and dimension correlation on the data stream to be processed, and obtains normalized behavioral data with supplementary event business attributes. Based on the session identifier and the timestamp, the normalized behavior data is path identified, behavior sequences containing the target event are filtered, and the path is split according to the target event to obtain a set of conversion paths ending with a single target event. The attribution model is invoked to calculate each path in the conversion path set, and the conversion contribution of the target event is allocated to the event to be attributed according to the allocation rules of the attribution model. An attribution result containing the contribution of each event to be attributed is generated. Based on the attribution result, any ad that has been placed is analyzed. The attribution results are obtained through the data display module, stored in the database, and visualized through microservice interfaces and front-end pages.

2. The real-time calculation method for attribution analysis of marketing conversion according to claim 1, characterized in that, The process of forming a data stream to be processed that is consistent with the original sequence of behaviors specifically includes: The original sequence of behaviors is written sequentially to a log file; The Flume Agent monitors changes to the log files, reads them line by line, pushes the data to the Kafka cluster, and uses the user's identity as the partition key to ensure the data order of the same user.

3. The real-time calculation method for attribution analysis of marketing conversion according to claim 1, characterized in that, The process of obtaining a set of transformation paths ending with a single target event is as follows: The normalized behavior data is grouped according to the session identifier and the user identity identifier; The grouped data is sorted in ascending order according to the timestamps to form user behavior paths; Identify the target event in the user behavior path, and use the target event as an anchor point to trace back the time range within the attribution window period to form the conversion path set. When a single path contains multiple target events, the path is split into multiple independent conversion paths with each target event as the endpoint.

4. The real-time calculation method for attribution analysis of marketing conversion according to claim 1, characterized in that, The attribution models include the first contact attribution model, the last contact attribution model, the linear attribution model, and the location attribution model; The first-touch attribution model assigns 100% of the conversion contribution to the first event to be attributed in the path; The last-touch attribution model assigns 100% of the conversion contribution to the event to be attributed at the end of the path. The linear attribution model distributes the transformation contribution equally among all attributable events in the path.

5. A real-time calculation system for attribution analysis of marketing conversion, characterized in that, include: The event tracking SDK module is used to: collect user behavior data for any delivered advertisement, and generate a raw behavior sequence containing user identity, session identity, timestamp and event code, wherein the event code contains the encoding of the target event and the event to be attributed; The log collection module is used to: write the original behavior sequence sequentially into a log file, read from the log file and push to a message queue to form a data stream to be processed that is consistent with the original behavior sequence. The attribution analysis calculation module is used to: consume the data stream to be processed from the message queue in real time, perform data cleaning and dimension correlation on the data stream to be processed, and obtain normalized behavioral data with supplementary event business attributes; Based on the session identifier and the timestamp, the normalized behavior data is path identified, behavior sequences containing the target event are filtered, and the path is split according to the target event to obtain a set of conversion paths ending with a single target event. The attribution model is invoked to calculate each path in the conversion path set, and the conversion contribution of the target event is allocated to the event to be attributed according to the allocation rules of the attribution model, generating an attribution result containing the contribution of each event to be attributed, and analyzing any ad that has been placed based on the attribution result. The data display module is used to: obtain the attribution results, store the attribution results in the database, and display them visually through microservice interfaces and front-end pages.

6. A real-time calculation system for attribution analysis of marketing conversion according to claim 5, characterized in that, The process of forming a data stream to be processed that is consistent with the original sequence of behaviors specifically includes: The original sequence of behaviors is written sequentially to a log file; The Flume Agent monitors changes to the log files, reads them line by line, pushes the data to the Kafka cluster, and uses the user's identity as the partition key to ensure the data order of the same user.

7. A real-time calculation system for attribution analysis of marketing conversion according to claim 5, characterized in that, The process of obtaining a set of transformation paths ending with a single target event is as follows: The normalized behavior data is grouped according to the session identifier and the user identity identifier; The grouped data is sorted in ascending order according to the timestamps to form user behavior paths; Identify the target event in the user behavior path, and use the target event as an anchor point to trace back the time range within the attribution window period to form the conversion path set. When a single path contains multiple target events, the path is split into multiple independent conversion paths with each target event as the endpoint.

8. A real-time calculation system for attribution analysis of marketing conversion according to claim 5, characterized in that, The attribution models include the first contact attribution model, the last contact attribution model, the linear attribution model, and the location attribution model; The first-touch attribution model assigns 100% of the conversion contribution to the first event to be attributed in the path; The last-touch attribution model assigns 100% of the conversion contribution to the event to be attributed at the end of the path. The linear attribution model distributes the transformation contribution equally among all attributable events in the path.

9. An electronic device, characterized in that, The electronic device includes a processor coupled to a memory storing at least one computer program, which is loaded and executed by the processor to enable the electronic device to perform the method as described in any one of claims 1 to 4.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which is loaded and executed by a processor to enable the computer to perform the method as described in any one of claims 1 to 4.

Images