Advertisement content updating method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of advertisement content updating methods, in particular to an advertisement content updating method and device, electronic equipment and a storage medium. The method specifically comprises the following steps: step 1, multi-dimensional data acquisition and standardization: acquiring core data through an advertisement data monitoring system, wherein the core data comprises historical delivery data, real-time user feedback data and industry trend data; step 2, advertisement effect decline degree calculation: constructing an advertisement effect decline degree model based on the standardized historical delivery data and real-time feedback data in the step 1, and quantifying the decline degree D of the current effect of the advertisement; according to the method, firstly, three types of core data of historical delivery, real-time user feedback and industry trend are systematically collected, data authenticity is ensured through abnormal data elimination, a comprehensive and reliable basis is provided for subsequent analysis, and the problem that the updating direction is fuzzy due to single data dimension of a traditional method is solved.

Description

Technical Field

[0001] This invention relates to the technical field of methods for updating advertising content, and in particular to a method, apparatus, electronic device, and storage medium for updating advertising content. Background Technology

[0002] In the field of digital advertising and marketing, timely optimization and updating of ad content is crucial for maintaining advertising competitiveness and improving campaign effectiveness. However, current mainstream ad content update technologies still have shortcomings, failing to meet advertisers' needs for precise and efficient updates, which severely restricts the maximization of advertising and marketing value.

[0003] In determining the timing of updates, traditional methods have long relied on a rigid "fixed-cycle driven" model. Whether it's a fixed update frequency of once a month or once every two weeks, it fails to consider the fluctuations in actual advertising effectiveness. Some ads, whose effectiveness has severely declined within the fixed cycle, continue to be run because it's not time to update, resulting in a significant waste of marketing resources. Conversely, some ads, whose effectiveness is still on the rise, are forced to update, interrupting the positive trend and completely deviating from the actual changing patterns of advertising effectiveness. This makes it impossible to adjust as needed and greatly reduces the cost-effectiveness of advertising. Summary of the Invention

[0004] In response to the technical problems mentioned in the background art, the present invention provides a method, apparatus, electronic device, and storage medium for updating advertising content.

[0005] The technical solution adopted in this invention is: a method for updating advertising content, specifically including the following steps:

[0006] Step 1: Multi-dimensional data collection and standardization: Collect core data through the advertising data monitoring system, including historical campaign data, real-time user feedback data, and industry trend data;

[0007] Step 2: Calculation of Ad Performance Decline: Based on the standardized historical delivery data and real-time feedback data from Step 1, construct an ad performance decline model to quantify the degree of decline D in the current ad performance;

[0008] Step 3: Industry Competitiveness Gap Analysis: Introduce industry benchmark values ​​and competitor data to calculate the current industry competitiveness gap G of the advertisement;

[0009] Step 4: Priority Determination: Combining the effectiveness decay rate D in Step 2 and the competitiveness gap G in Step 3, determine the urgency P of updating the ad content using the priority determination formula to clarify whether an immediate update is needed.

[0010] Step 5: Optimize Content Update Direction: Based on the multi-dimensional data from Step 1, use a contribution analysis model to pinpoint specific update and optimization directions. Calculate the contribution coefficient C of each data dimension to advertising effectiveness;

[0011] Step 6: Update effect prediction: Based on the optimization direction determined in Step 5, construct an update effect prediction model and use the gradient boosting tree algorithm to calculate the expected improvement rate of the updated effect;

[0012] Step 7: Update plan generation and implementation; Based on the optimization direction in Step 5 and the prediction results in Step 6, generate specific update plans through the rule engine.

[0013] In one embodiment, in step 1, outlier data is removed after collection, and the data is mapped to a uniform distribution range using the Z-Score normalization formula to eliminate dimensional differences.

[0014]

[0015] Among them, X std For the standardized data, X is the original data, μ is the mean of the data in this dimension, and σ is the standard deviation.

[0016] In one embodiment, step 2 first calculates the decay rate of the core performance indicators, and then combines this with negative indicators from real-time user feedback to calculate the degree of decay through weighted summation.

[0017]

[0018] Where D represents the ad performance degradation, ω1 = 0.4 (click-through rate decay), ω2 = 0.4 (conversion rate decay), ω3 = 0.2 (complaint rate weight), and CTR... current For the current period's click-through rate and CR current CTR is the conversion rate for the current period. peak Peak click-through rate and CR in the past 30 days peak For conversion rates over the past 30 days, Com std The complaint rate is the standardized rate.

[0019] In one embodiment, step 3 first involves obtaining average performance metrics for similar advertisements from an industry database, including but not limited to the industry average click-through rate (CTR). ind Average conversion rate CR ind Furthermore, consider the core competitive advantages of competitors' advertising metrics, including but not limited to the popularity of competitors' keywords (Hot). comp Competitiveness is quantified using a gap calculation formula:

[0020]

[0021] Where G represents the industry competitiveness gap, Hot self Based on the current popularity of advertising keywords, Hot comp To boost the popularity of competitor advertising keywords.

[0022] In one embodiment, in step 4, the urgency level P is calculated as follows:

[0023] P = γ × D + (1 - γ) × G;

[0024] Where P is the update priority and γ = 0.6 is the weight of the effect decay.

[0025] In one embodiment, step 5 uses a multiple linear regression formula for calculation:

[0026]

[0027] Where Effect refers to the advertising effect, C i (Contribution coefficient for the i-th data dimension), X std,i Let ε be the i-th standardized data dimension, and let ε be the error term.

[0028] In one embodiment, in step 6, the improvement rate R is calculated as follows:

[0029]

[0030] Where R represents the expected improvement rate, and Effect pred To predict the updated ad performance using the GBRT model, Effect current This represents the current advertising effectiveness.

[0031] In one embodiment, an advertising content updating device includes:

[0032] The data acquisition and processing module is used to collect historical delivery data, real-time user feedback data, and industry trend data from advertisements, and to perform standardized processing.

[0033] The advertising effectiveness decay calculation module is used to construct an advertising effectiveness decay model to quantify the degree of decay of the current advertising effectiveness.

[0034] The industry competitiveness gap analysis module is used to: introduce industry benchmark values ​​and competitor data, and calculate the current industry competitiveness gap of the advertisement.

[0035] The update priority determination module determines the urgency of updating the advertising content through a priority determination formula, thus clarifying whether an immediate update is required.

[0036] The update content optimization direction positioning module uses a contribution analysis model to locate specific update optimization directions and calculates the contribution coefficient of each data dimension to the advertising effect.

[0037] The update effect prediction module constructs an update effect prediction model and uses the gradient boosting tree algorithm to calculate the expected effect improvement rate after the update.

[0038] The update scheme generation and implementation module generates specific update schemes through a rules engine.

[0039] The beneficial effects of this invention are as follows: Compared with existing technologies, this invention first systematically collects three types of core data: historical campaign data, real-time user feedback, and industry trends. Simultaneously, it ensures data authenticity by eliminating outlier data, providing a comprehensive and reliable foundation for subsequent analysis. This solves the problem of unclear update direction caused by the single data dimension in traditional methods. Regarding update timing, the solution abandons the traditional fixed-cycle update mechanism. It quantifies the decline of core performance indicators and real-time negative feedback into a decline degree using an advertising performance decline formula. Combined with an industry competitiveness gap formula, it clarifies the advertisement's positioning within the industry. Then, it dynamically determines the urgency of updates using an update priority formula. Updates are triggered promptly when decline is severe or the competitiveness gap is large, while adjustments are postponed when performance is good. This completely breaks free from the constraints of fixed cycles and effectively avoids the waste of resources from continuously running declining ads or the mistaken updating of high-quality ads. Attached Figure Description

[0040] Figure 1 This is a flowchart illustrating the present invention; Detailed Implementation

[0041] In the description of this invention, it should be noted that the terms "front", "up", "down", "left", "right", "vertical", "horizontal", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0042] To address the problems existing in the background technology, this application proposes the following technical solution: a method for updating advertising content, specifically including the following steps:

[0043] Step 1: Multi-dimensional data collection and standardization;

[0044] Three types of core data are collected through an existing advertising data monitoring system: 1) Historical campaign data (Click-through rate (CTR), Conversion rate (CR), and Impressions (Imp) for the past 30 days); 2) Real-time user feedback data (User sentiment index (Sen), Interaction rate (Inter), and Complaint rate (Com); 3) Industry trend data (Competitor ad update frequency (Fre), Hot topic relevance (Rel), and Industry keyword popularity (Hot). After collection, outlier data (such as click-through rates greater than 50%) is removed, and the data is standardized using the Z-Score formula to map it to a uniform distribution range, eliminating dimensional differences.

[0045]

[0046] Among them, X std For standardized data, X represents the original data, μ represents the mean of the data in this dimension, and σ represents the standard deviation. The purpose of this formula is to transform data of different magnitudes (such as impressions in "tens of thousands" and complaint rates in "%)) into comparable standardized data, laying the foundation for subsequent multi-dimensional effect evaluation and preventing the analysis results from being dominated by excessively large values ​​in one dimension (such as millions of impressions).

[0047] The above technical solution is explained as follows: By systematically collecting three core data types—historical campaign data, real-time user feedback, and industry trends—it comprehensively covers key information in advertising operations. Historical data reflects past advertising effectiveness, real-time feedback captures current user attitudes, and industry trends relate to the market competition environment. The combination of these three transforms data collection from "single and fragmented" to "comprehensive and systematic." Simultaneously, by removing outlier data and filtering out invalid and interfering information, data authenticity is ensured. Standardization eliminates differences between data of different dimensions, making data that was previously incomparable fair and comparable, preventing one dimension's data from dominating analysis due to its size advantage, and laying a solid foundation for subsequent multi-dimensional quantitative evaluation.

[0048] Step 2: Calculate the advertising effectiveness decay rate;

[0049] Based on the standardized historical campaign data and real-time feedback data from Step 1, an advertising effectiveness decay model is constructed to quantify the degree of decline (D) in the current advertising effectiveness. First, the decay rate of core performance metrics (click-through rate, conversion rate) is calculated. Then, combined with negative metrics from real-time user feedback (complaint rate, percentage of negative comments), the decay degree is calculated using a weighted summation formula:

[0050]

[0051] Where D represents the ad performance degradation (range [0,1], the larger the value, the more severe the degradation), ω1=0.4, ω2=0.4, and ω3=0.2 are the weights of click-through rate degradation, conversion rate degradation, and complaint rate, respectively. CTR currentCR current CTR represents the click-through rate and conversion rate for the current period. peak CR peak The peak click-through rate and conversion rate over the past 30 days, Com std This is the standardized complaint rate. The purpose of this formula is to quantify the decline in advertising effectiveness, replacing the traditional fixed-period update mechanism. For example, when D = 0.7 (severe decline), an update warning is triggered, and when D = 0.3 (slight decline), no update is performed, thus achieving dynamic update timing judgment.

[0052] The above technical solution is explained as follows: By employing an innovative decay quantification model, the rigidity of traditional fixed-cycle update mechanisms is resolved, enabling dynamic judgment of ad update timing. Traditional methods update at fixed intervals, forcibly adjusting ads regardless of whether their effectiveness has declined. This can easily lead to well-performing ads being mistakenly updated or declining ads being continuously run, wasting resources. This step combines the decay status of core performance indicators with real-time negative feedback, transforming the ad's decline status into a quantifiable metric that clearly presents the current degree of decline. Using a weighted summation logic, it considers both the decay of core performance indicators such as click-through rate and conversion rate, as well as negative feedback such as complaint rates, making the decay judgment more comprehensive. Through this metric, advertisers can accurately identify whether an ad has entered a decline phase. When decline is severe, an update warning can be triggered promptly; when decline is slight, no adjustment is needed. This completely breaks free from the constraints of fixed cycles, allowing update timing judgments to better align with actual ad performance changes, effectively avoiding resource waste and ensuring that ads maximize their value within their effective period.

[0053] Step 3: Industry competitiveness gap analysis;

[0054] By introducing industry benchmarks and competitor data, the current industry competitiveness gap (G) of the advertisement is calculated. First, the average performance metrics (such as the industry average click-through rate, CTR) of similar advertisements are obtained from an industry database. ind Average conversion rate CR ind ), and then combine the core competitive advantages of competitors' advertising (such as the popularity of competitors' keywords). comp ), quantifying competitiveness through a gap calculation formula:

[0055]

[0056] Where G represents the industry competitiveness gap (range [0,1], the larger the value, the larger the gap), Hot self Based on the current popularity of advertising keywords, Hot compThe keyword popularity of competitor ads (all data are standardized after step 1). The purpose of this formula is to clarify the competitive positioning of the ad in the industry. For example, when G = 0.6, it indicates that the ad lags behind the industry average in click-through rate, conversion rate, and keyword popularity. It is necessary to prioritize updating to narrow the gap and solve the problem of traditional methods neglecting industry comparison.

[0057] The above technical solution is explained as follows: By introducing industry benchmarks and competitor data, and through comprehensive calculations, the gap between the advertisement and the industry average and competitors is quantified. From core performance indicators to detailed dimensions such as keyword popularity, a comprehensive comparative analysis is conducted, allowing advertisers to clearly identify where they lag behind the industry and where the gaps exist with competitors. This comparative analysis is no longer a vague "inferior to competitors," but rather precisely identifies the areas of gap, providing a clear benchmarking direction for subsequent updates. Through this step, advertisers can avoid closed-door updates, ensuring that the update direction closely aligns with industry trends and competitor advantages, making the optimized advertisement more competitive in the market and effectively narrowing the gap with the industry's best practices.

[0058] Step 4: Update priority determination;

[0059] Combining the effectiveness decay rate D in step 2 with the competitiveness gap G in step 3, the urgency P of updating the ad content is determined using a priority judgment formula to clarify whether an immediate update is necessary.

[0060] P=γ×D+ ( 1-γ)×G

[0061] Where P is the update priority (value range [0,1], P≥0.5 indicates immediate update is required, P<0.5 indicates postponement of update), and γ=0.6 is the weight of the effect decay degree (prioritizing the current effect decay risk). The purpose of this formula is to dynamically determine the update timing, replacing the traditional fixed period mechanism. For example, if an ad has D=0.8 (severe decay) and G=0.4 (small competitive gap), the calculated P=0.64 indicates that it needs to be updated immediately, avoiding the continued running of decaying ads due to the fixed period not being reached.

[0062] The above technical solution is explained as follows: By integrating the degree of performance degradation with the gap in industry competitiveness, a scientific prioritization logic is constructed, solving the problems of chaotic update timing and lack of clear judgment standards in traditional methods. Traditional technologies either update according to a fixed cycle or rely on subjective human judgment, often resulting in delays for urgently needed updates and priority adjustments for ads that do not require urgent updates. This step uses performance degradation as the core, combined with the gap in industry competitiveness, to clarify the urgency of updates through weighted calculations. Prioritizing the risk of performance degradation of the ads themselves, while also taking into account industry competitive pressure, a clear priority standard is formed. If the urgency threshold is reached, updates are made immediately; otherwise, they are postponed, making update decisions systematic. This mechanism ensures that advertisers can focus their energy and resources on the ads that most need updating, avoiding the dispersion of resources into non-urgent update needs, improving update efficiency, and ensuring that each update addresses key issues specifically. This avoids resource waste due to delayed updates and prevents unnecessary costs from excessive updates.

[0063] Step 5: Update content and optimize positioning;

[0064] Based on the multi-dimensional data from step 1, a contribution analysis model is used to pinpoint specific directions for updates and optimizations. The contribution coefficient C of each data dimension to advertising effectiveness (click-through rate, conversion rate) is calculated using a multiple linear regression formula.

[0065]

[0066] Where Effect refers to the advertising effect (such as the standardized value of click-through rate), C i X represents the contribution coefficient of the i-th data dimension (a positive value indicates that the dimension is positively correlated with the effect, and the larger the absolute value, the stronger the influence). std,i Let ε be the i-th standardized data dimension (e.g., keyword popularity, comment sentiment), and let ε be the error term. This formula is used to precisely pinpoint the optimization direction. For example, if the Hot contribution coefficient of keyword popularity C = 0.8 (strong positive correlation), and the current ad is Hot... std =0.2 (too low), so the optimization direction is clearly "increase the popularity of advertising keywords", which solves the problem of the general update direction of traditional methods.

[0067] The above technical solution is explained as follows: By achieving precise positioning of the update and optimization direction, it completely solves the problem of traditional methods having a vague update direction and lacking specific guidance. Based on multi-dimensional data, regression analysis is used to calculate the degree of influence of each data dimension on advertising performance, clarifying which dimensions are strongly correlated with advertising performance and which dimensions have shortcomings. By analyzing the contribution coefficient of each dimension, the key factors affecting advertising performance are accurately identified, thereby determining specific optimization directions—no longer the vague "optimize keywords," but clearly indicating the need to increase keyword popularity; no longer the ambiguous "improve user feedback," but precisely targeting the adjustment of copy tone. This gives advertisers a clear goal for their update work, eliminating the need for blind attempts and allowing them to directly address key shortcomings, significantly improving the accuracy and efficiency of updates, ensuring that every optimization directly addresses the core issues and effectively improves advertising performance.

[0068] Step 6: Update the predicted effects;

[0069] Based on the optimization direction determined in step 5, a prediction model for the update effect is constructed, and the expected improvement rate R after the update is calculated using the Gradient Boosting Tree (GBRT) algorithm, as shown in the following formula:

[0070]

[0071] Where R is the expected improvement rate (positive value indicates improved effect, negative value indicates possible decrease in effect), and Effect pred To predict the updated ad performance (e.g., click-through rate) using the GBRT model, Effect current This represents the current advertising performance. The purpose of this formula is to assess the potential effect of an update plan in advance. For example, if the predicted R = 30% (click-through rate increases by 30%), the update can be executed with confidence; if R = -5% (click-through rate decreases by 5%), the optimization direction needs to be readjusted to reduce the risk of trial and error in the update.

[0072] The above technical solution is explained as follows: By introducing a performance prediction model, the potential effects of the update plan are assessed in advance, solving the problems of high trial-and-error risk and inability to avoid ineffective updates in advance in traditional methods. Traditional updates often follow a "execute first, then observe the results" approach. If the update is ineffective, it not only wastes time and resources but may also lead to missed market opportunities. This step, based on a defined optimization direction, uses professional algorithms to predict the improvement in performance after the update, transforming the abstract update plan into quantifiable performance expectations. If the predicted effect is good, the update can be confidently implemented; if the predicted effect is poor or even declining, the optimization direction can be adjusted in time to avoid ineffective updates. This mechanism allows advertisers to have a clear expectation of the effect before the update is implemented, avoiding risks in advance and reducing trial-and-error costs. At the same time, performance prediction also provides a basis for adjusting the update plan, helping advertisers continuously improve the plan before implementation, ensuring that the final updated plan maximizes advertising effectiveness and avoids losses caused by blind updates.

[0073] Step 7: Update and implement the solution;

[0074] Based on the optimization direction in step 5 and the prediction results in step 6, a specific update plan is generated through the rule engine. For example, if the optimization direction is "increase keyword popularity" and the prediction effect is good, the generated plan is: "Replace one low-popularity keyword (e.g., replace "high-quality" with "bestseller"), and add two new high-popularity industry keywords (e.g., 2024 new products)." If the optimization direction is "improve user emotional feedback," the generated plan is: "Adjust the copywriting tone from 'aggressive promotion' to 'warm recommendation,' and delete the 'must-buy for a limited time' expression that triggered negative comments." Simultaneously, a post-update performance monitoring period (e.g., 7 days) is set. If the actual performance improvement rate deviates from the predicted value by more than 10%, a secondary optimization mechanism is triggered, and the update plan is adjusted. The purpose of this step is to transform the analysis results into a feasible update strategy, achieving a closed loop of "analysis-prediction-implementation-monitoring," ensuring that the advertising update effect is controllable.

[0075] The above technical solution is explained as follows: By transforming the analysis results into specific and feasible update solutions and establishing an effect monitoring and secondary optimization mechanism, a complete closed loop of "analysis-prediction-implementation-monitoring" is achieved, which solves the problems of traditional methods in that update solutions are difficult to implement and lack follow-up support.

[0076] In summary, this invention first systematically collects three core data types: historical campaign data, real-time user feedback, and industry trends. Combined with the Z-Score standardization formula, it eliminates the differences between data of different dimensions, making previously incomparable data fair and comparable. This prevents one dimension's data from dominating the analysis due to its magnitude advantage. Simultaneously, outlier removal ensures data authenticity, providing a comprehensive and reliable foundation for subsequent analysis. This solves the problem of unclear update direction caused by the single data dimension in traditional methods. Regarding update timing, the solution abandons the traditional fixed-cycle update mechanism. It quantifies the decline of core performance indicators and real-time negative feedback into a decline degree using an advertising performance decline formula. Combined with an industry competitiveness gap formula, it clarifies the advertisement's positioning within the industry. Then, an update priority formula dynamically determines the urgency of the update. Updates are triggered promptly when decline is severe or the competitiveness gap is large, while adjustments are postponed when performance is good. This completely breaks free from the constraints of fixed cycles, effectively preventing the waste of resources from continuously running declining ads or the accidental updating of high-quality ads. It solves the pain point of traditional methods' inability to flexibly adjust update timing. In terms of optimization direction and performance assurance, the solution uses a multiple linear regression formula to calculate the contribution coefficient of each data dimension to advertising performance, accurately identifying key factors affecting results and clarifying specific optimization directions (such as increasing keyword popularity or adjusting copy tone). This replaces traditional, general optimization suggestions, giving advertisers clear goals for their updates. Simultaneously, it introduces a gradient boosting tree algorithm to build an update performance prediction model. The feasibility of the solution is assessed in advance using a formula for expected performance improvement rate, allowing for timely adjustments if results are unsatisfactory, significantly reducing trial-and-error risks. Furthermore, the solution generates specific, actionable update plans through a rule engine and sets performance monitoring cycles and a secondary optimization mechanism. If the actual results deviate significantly from the predictions, secondary optimization is triggered to ensure controllable results, avoiding the problem of uncontrolled performance due to lack of tracking after updates, as is common with traditional methods. This not only significantly improves the efficiency and effectiveness of ad updates, helping advertisers accurately adjust keywords, copy, and other content, achieving substantial increases in core metrics such as click-through rate without repeated trial and error, but also effectively reduces marketing resource waste, lowers trial-and-error costs, and enhances the competitiveness of advertising in the industry.

[0077] Although embodiments of the invention have been shown and described, the scope of the invention will be defined by the appended claims and their equivalents by those skilled in the art.

Claims

1. A method for updating advertising content, characterized in that, Specifically, the following steps are included: Step 1: Multi-dimensional data collection and standardization: Collect core data through the advertising data monitoring system, including historical campaign data, real-time user feedback data, and industry trend data; Step 2: Calculation of Ad Performance Decline: Based on the standardized historical delivery data and real-time feedback data from Step 1, construct an ad performance decline model to quantify the degree of decline D in the current ad performance; Step 3: Industry Competitiveness Gap Analysis: Introduce industry benchmark values ​​and competitor data to calculate the current industry competitiveness gap G of the advertisement; Step 4: Priority Determination: Combining the effectiveness decay rate D in Step 2 and the competitiveness gap G in Step 3, determine the urgency P of updating the ad content using the priority determination formula to clarify whether an immediate update is needed. Step 5: Update Content Optimization Direction: Based on the multi-dimensional data from Step 1, use the contribution analysis model to pinpoint specific update optimization directions; calculate the contribution coefficient C of each data dimension to the advertising effect; Step 6: Update effect prediction: Based on the optimization direction determined in Step 5, construct an update effect prediction model and use the gradient boosting tree algorithm to calculate the expected improvement rate of the updated effect; Step 7: Update plan generation and implementation; Based on the optimization direction in Step 5 and the prediction results in Step 6, generate specific update plans through the rule engine.

2. The method for updating advertising content according to claim 1, characterized in that, In step 1, outlier data is removed after collection, and the data is mapped to a uniform distribution interval using the Z-Score standardization formula to eliminate dimensional differences. Among them, X std For the standardized data, X is the original data, μ is the mean of the data in this dimension, and σ is the standard deviation.

3. The method for updating advertising content according to claim 2, characterized in that, In step 2, the decay rate of the core performance indicators is first calculated, and then the degree of decay is calculated by weighted summation, taking into account the negative indicators from real-time user feedback. Where D represents the ad performance degradation, ω1 = 0.4 (click-through rate decay), ω2 = 0.4 (conversion rate decay), ω3 = 0.2 (complaint rate weight), and CTR... current For the current period's click-through rate and CR current CTR is the conversion rate for the current period. peak Peak click-through rate and CR in the past 30 days peak For conversion rates over the past 30 days, Com std The standardized complaint rate.

4. The method for updating advertising content according to claim 3, characterized in that, In step 3, the first step is to obtain the average performance metrics of similar advertisements through an industry database, including but not limited to the industry average click-through rate (CTR). ind Average conversion rate CR ind Furthermore, consider the core competitive advantages of competitors' advertising metrics, including but not limited to the popularity of competitors' keywords (Hot). comp Competitiveness is quantified using a gap calculation formula: Where G represents the industry competitiveness gap, Hot self Based on the current popularity of advertising keywords, Hot comp To boost the popularity of competitor advertising keywords.

5. The method for updating advertising content according to claim 4, characterized in that, In step 4, the urgency level P is calculated as follows: P = γ × D + (1 - γ) × G; Where P is the update priority and γ = 0.6 is the weight of the effect decay.

6. The method for updating advertising content according to claim 5, characterized in that, In step 5, the multiple linear regression formula is used for calculation: Where Effect refers to the advertising effect, C i (Contribution coefficient for the i-th data dimension), X std,i Let ε be the i-th standardized data dimension, and let ε be the error term.

7. The method for updating advertising content according to claim 6, characterized in that, The formula is as follows: In step 6, the improvement rate R is calculated as follows: Where R represents the expected improvement rate, and Effect pred To predict the updated ad performance using the GBRT model, Effect current This represents the current advertising effectiveness.

8. An advertising content updating device, characterized in that, include: The data acquisition and processing module is used to collect historical delivery data, real-time user feedback data, and industry trend data from advertisements, and to perform standardized processing. The advertising effectiveness decay calculation module is used to construct an advertising effectiveness decay model to quantify the degree of decay of the current advertising effectiveness. The industry competitiveness gap analysis module is used to: introduce industry benchmark values ​​and competitor data, and calculate the current industry competitiveness gap of the advertisement. The update priority determination module determines the urgency of updating the advertising content through a priority determination formula, thus clarifying whether an immediate update is required. The update content optimization direction positioning module uses a contribution analysis model to locate specific update optimization directions and calculates the contribution coefficient of each data dimension to the advertising effect. The update effect prediction module constructs an update effect prediction model and uses the gradient boosting tree algorithm to calculate the expected effect improvement rate after the update. The update scheme generation and implementation module generates specific update schemes through a rules engine.

9. An electronic device, characterized in that, The electronic device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for updating advertising content as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for updating advertising content as described in any one of claims 1 to 7.

Images