Intelligent marketing potential analysis method applied to new product popularization

Abstract

The invention is suitable for the technical field of intelligent marketing data processing and advertisement putting strategy optimization, and provides an intelligent marketing potential analysis method applied to new product promotion, and the method comprises the steps: obtaining the historical sales data of a target advertisement-driven business place, and the advertisement display content of a to-be-promoted new product; analyzing the historical sales data, and identifying whether there is a reference historical new product which belongs to the same kind of the to-be-popularized new product, is in a new product popularization stage, and has an instant decision-making type purchase behavior exceeding a preset standard in a plurality of night sub-periods. According to the method, a multi-dimensional mapping and correction mechanism of'key advertisement display element-display frequency-expression intensity value 'is introduced for the first time by constructing an intelligent analysis method for the marketing potential based on night sub-period division, and dynamic optimal configuration of new advertisement contents in different sub-periods is realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent marketing data processing and advertisement placement strategy optimization, and particularly relates to a dynamic sales potential intelligent analysis method applied to new product promotion. BACKGROUND

[0002] In the actual store operation process, for the goods in the new product promotion stage, the operator usually arranges concentrated advertisement placement, and displays the advertisement in the visual display area in the form of graphics, video, light box, etc., so as to attract the attention of customers and promote conversion. Especially in the advertisement-driven business places with night business capacity, the advertisement display content often becomes an important influencing factor for customers to make instant purchase decisions. However, due to the great influence of time period, physiological rhythm, environmental illumination and other factors on the behavior characteristics of night customers, the conversion effect of the same advertisement content in different night sub-periods is significantly different. How to match a more suitable advertisement content combination to improve the dynamic sales potential in different time periods becomes the key to the optimization of new product promotion strategy of the store.

[0003] In the prior art, the formulation and display frequency of the advertisement content usually depend on the experience setting of the operator or the unified configuration based on the static template, and there is a lack of accurate identification and adaptation to the reaction difference of customer behavior in different time periods. At the same time, the current method is difficult to dynamically migrate and optimize the new product promotion strategy according to the known high conversion historical samples, especially in the early stage when the new product has not accumulated enough behavior data. It often faces the problems of unreasonable allocation of advertisement resources and poor display effect. In addition, the traditional placement mechanism cannot be refined to the precision of "advertisement element-time period", and cannot effectively utilize the high conversion rules contained in the historical data, resulting in poor timeliness of advertisement response and insufficient promotion accuracy. SUMMARY

[0004] The purpose of the present application is to provide a dynamic sales potential intelligent analysis method applied to new product promotion, which aims to solve the problems raised in the background art.

[0005] The present application is implemented as follows: a dynamic sales potential intelligent analysis method applied to new product promotion, the method comprising:

[0006] obtaining historical sales data of a target advertisement-driven business place and advertisement display content of a new product to be promoted;

[0007] analyzing the historical sales data to identify whether there is a reference historical new product which is the same type of goods as the new product to be promoted and is in the new product promotion stage, and which appears instant decision-making purchase behavior exceeding a preset standard in a plurality of night sub-periods;

[0008] If there is a reference historical new product, its advertisement display content is obtained, and the key advertisement display elements and the corresponding expression intensity values of the new product to be promoted and the reference historical new product in each nightly sub-period are extracted respectively;

[0009] The key advertisement display elements and the historical display frequency of the reference historical new product in each nightly sub-period are converted into the key advertisement display elements and the initial display frequency of the new product to be promoted respectively, and a key advertisement display element set is constructed;

[0010] According to the deviation amplitude of the expression intensity values of the new product to be promoted and the reference historical new product on the corresponding key advertisement display element, the initial display frequency in the key advertisement display element set is corrected, and the optimized advertisement display content of the new product to be promoted is generated.

[0011] As a further limitation of the technical scheme of the embodiment of the application, the advertisement-driven business place refers to a retail place that takes advertisement display as the main commodity promotion means, and has a visual display area for customers, an advertisement delivery device, and a sensing device for collecting customer behavior data.

[0012] As a further limitation of the technical scheme of the embodiment of the application, the division of the several nightly sub-periods refers to dividing the nightly time period into several sub-periods according to a preset time division rule within the nightly business hours of the target advertisement-driven business place, so as to capture the behavior response characteristics of customers to the advertisement display content in different time periods.

[0013] As a further limitation of the technical scheme of the embodiment of the application, the step of analyzing the historical sales data and identifying whether there is a reference historical new product that belongs to the same commodity category as the new product to be promoted and is in the new product promotion stage, and has instant decision-making purchase behavior exceeding a preset standard in the several nightly sub-periods includes:

[0014] Determining the commodity category to which the new product to be promoted belongs, and screening the sales data of several historical new products of the same commodity category and in the new product promotion stage in the historical sales data;

[0015] For each historical new product, the image information of customers watching the visual display area and the purchase behavior information thereof in each nightly sub-period are extracted;

[0016] The instant decision-making purchase behavior of customers in each nightly sub-period is analyzed, and it is determined whether it exceeds a preset standard. The instant decision-making purchase behavior exceeding the preset standard refers to the proportion of customers who enter the store and complete the purchase after watching the advertisement display of the historical new product to the total number of historical new product purchases in the nightly sub-period exceeding a preset threshold value;

[0017] If there are instant decision-making purchase behaviors exceeding the preset standard in all nighttime sub-periods, then the historical new product is determined to be a reference historical new product.

[0018] As a further limitation of the technical solution of the present invention, if the number of identified reference historical new products exceeds one, then based on the pricing, use and other product attributes of the new product to be promoted, the one that is closest to the new product to be promoted is selected as the final reference historical new product.

[0019] As a further limitation of the technical solution of this invention embodiment, the key advertising display elements and their corresponding expression intensity values ​​refer to:

[0020] Key advertising display elements refer to the key components in advertising display content that have a significant impact on guiding customer attention or forming purchase intentions, including at least one of the main image area, core copy fields, visual focus graphics, color contrast areas, or sound emphasis segments;

[0021] The expression intensity value refers to a numerical parameter used to characterize the prominence of the key advertising display element in the overall advertising display content. It is calculated based on one or more of the following factors: color saturation, image area ratio, font size, volume intensity, animation frequency, or appearance duration.

[0022] As a further limitation of the technical solution of this invention, the step of constructing a set of key advertising display elements by referring to the key advertising display elements and historical display frequencies of historical new products in each nighttime sub-period, respectively, and converting them into key advertising display elements and initial display frequencies of the new product to be promoted, includes:

[0023] Extract the advertising content of historical new products in each nighttime sub-period from historical sales data, and identify the corresponding key advertising elements and historical display frequency;

[0024] The original key advertising display elements of the new products to be promoted in each nighttime sub-period are replaced with the key advertising display elements of the reference historical new products that are consistent through structural hierarchy matching or semantic tag matching, and the historical display frequency of the key advertising display elements of the reference historical new products is used as the initial benchmark value.

[0025] Based on the similarity difference between the key advertising display elements of the referenced historical new products and the original key advertising display elements of the corresponding new products to be promoted, the initial benchmark value is corrected to obtain the initial display frequency of the new products to be promoted in that nighttime sub-period.

[0026] As a further limitation of the technical solution of the present invention, the historical display frequency refers to the number of times or the display duration of the key advertising display element is repeatedly displayed in the corresponding nighttime sub-period, which accounts for the proportion of the total display duration of the advertisement in that period.

[0027] As a further limitation of the technical solution of this invention, the step of generating optimized advertising display content for the new product to be promoted by correcting the initial display frequency of each key advertising display element in the set of key advertising display elements based on the deviation of the expression intensity value of the new product to be promoted and the reference historical new products in the corresponding key advertising display elements includes:

[0028] The deviation of the expression intensity values ​​of the new product to be promoted and the reference historical new products in the corresponding key advertising display elements is quantified in each night sub-period, and the deviation is used as a correction factor to form a set of correction factors;

[0029] By using each correction factor to adjust the initial display frequency of the key advertising display elements of the new product to be promoted, the corrected display frequency can be obtained.

[0030] All the revised display frequencies are recombined with the corresponding key advertising display elements to form the optimized advertising display content for the new product to be promoted. The optimized advertising display content serves as the input basis for subsequent intelligent analysis of the sales potential of the new product to be promoted, and is used to improve the accuracy of advertising response prediction in different nighttime sub-segments.

[0031] As a further limitation of the technical solution of this invention, the expression intensity value of the new product to be promoted on the corresponding element is subtracted from the expression intensity value of the historical new product on the corresponding key advertising display element, and the difference is divided by the expression intensity value of the historical new product to obtain the relative expression intensity deviation rate, which is used as a correction factor.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] This invention constructs an intelligent analysis method for sales potential based on nighttime sub-segments, introducing for the first time a multi-dimensional mapping and correction mechanism of "key advertising display elements - display frequency - expression intensity value," enabling dynamic optimization of new product advertising content across different sub-segments. Compared to traditional methods relying on static templates and experience-based placement strategies, this invention can identify high-conversion behavior characteristics from historical new product data and, combined with structural matching and semantic tag matching, accurately generate an initial set of advertising elements for the new product to be promoted. Furthermore, it corrects the frequency based on the deviation in expression intensity, effectively improving the time-segment adaptability of advertising response and the accuracy of conversion potential identification. This method combines feasibility and intelligent adaptability, making it particularly suitable for nighttime promotion scenarios in advertising-driven retail spaces, and possesses significant commercial application value and promotional prospects. Attached Figure Description

[0034] Figure 1 A flowchart of the method provided in the embodiments of the present invention;

[0035] Figure 2 This is a flowchart illustrating the process of identifying referenced historical new products in the method provided in this embodiment of the invention;

[0036] Figure 3 This is a flowchart illustrating the process of constructing a set of key advertising display elements in the method provided in this embodiment of the invention;

[0037] Figure 4 This is a flowchart illustrating the method for correcting the display frequency based on the deviation of the expression intensity in an embodiment of the present invention. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0039] Figure 1 A flowchart of the method provided by an embodiment of the present invention is shown.

[0040] Specifically, a method for intelligent analysis of sales potential applied to new product promotion includes the following steps:

[0041] Step S100: Obtain historical sales data of the target advertising-driven retail location, as well as the advertising display content of the new products to be promoted. The advertising-driven retail location refers to a retail location that uses advertising display as the main means of product promotion and has a visual display area for customers, advertising placement devices, and sensing devices for collecting customer behavior data.

[0042] Step S200: Analyze historical sales data to identify whether there are any reference historical new products that belong to the same category as the new product to be promoted, are in the new product promotion stage, and have shown immediate decision-making purchasing behavior exceeding preset standards in several nighttime sub-periods. The division of the several nighttime sub-periods refers to dividing the nighttime period into several sub-periods within the nighttime business hours of the target advertising-driven business venue according to preset time division rules, in order to capture the behavioral response characteristics of customers to the advertising content displayed in different time periods.

[0043] Specifically, Figure 2 A flowchart is shown to illustrate the process of identifying new products based on historical data.

[0044] The process of analyzing historical sales data to identify reference products that are similar to the new product to be promoted, are in the new product promotion stage, and have shown immediate decision-making purchasing behavior exceeding preset standards in several nighttime sub-periods includes the following steps:

[0045] Step S201: Determine the product category of the new product to be promoted, and filter the sales data of several historical new products that have the same product category and are in the new product promotion stage from the historical sales data;

[0046] Step S202: For each historical new product, extract the image information of customers viewing the visual display area and their purchase behavior information during each nighttime sub-period.

[0047] Step S203: Analyze the instant decision-making purchasing behavior of customers in each nighttime sub-period and determine whether it exceeds the preset standard. The instant decision-making purchasing behavior that exceeds the preset standard refers to the proportion of customers who enter the store and complete the purchase after watching the historical new product advertisement display to the total number of historical new product purchases in that nighttime sub-period that exceeds the preset threshold.

[0048] Step S204: If there are instant decision-making purchase behaviors exceeding the preset standard in all nighttime sub-periods, then the historical new product is determined to be a reference historical new product.

[0049] If more than one historical new product is identified as a reference, then based on the pricing, usage, and other product attributes of the new product to be promoted, the one that is closest to the new product to be promoted will be selected as the final historical new product to be referenced.

[0050] In this embodiment of the invention, the focus of the research scenario is the new product promotion activities carried out by target advertising-driven business venues in the advertising display environment during the nighttime period. The aim is to achieve intelligent identification and optimization of the sales potential of new products by analyzing the correlation between advertising response and purchasing behavior during this specific period.

[0051] This scenario exhibits the following significant characteristics: nighttime hours are often accompanied by faster customer decision-making, shorter dwell times, fewer sources of advertising interference, and greater sensitivity to emotional states. This is especially true when the business environment is advertising-driven, where the response patterns to advertising stimuli are more representative. Therefore, this invention, by analyzing customers' immediate decision-making purchasing behavior during nighttime sub-segments, can more accurately identify new products with advertising response-driven characteristics, and based on this, construct advertising display strategies with greater sales potential, thereby improving the conversion efficiency of new products to be promoted.

[0052] Advertising-driven retail locations can include convenience stores, chain pharmacies, supermarkets operating at night, and temporary retail areas in airports or train stations. Their common characteristics are that they rely on advertising content (rather than salesperson guidance) to stimulate customer attention and decision-making behavior, and they have relatively standardized advertising display devices and behavior monitoring systems.

[0053] Advertising content can include: product video ads displayed on electronic screens, lightbox posters, promotional animations, automated voice announcements, interactive advertising terminal interfaces, etc.

[0054] Visual display areas can include: advertising screens above shelves, visual flow charts around the checkout area, and light boxes at the store entrance;

[0055] Advertising devices can include: LCD / LED advertising screens, electronic price tag terminals, voice broadcasters, AR advertising terminals, etc.

[0056] Sensing devices used to collect customer behavior data can include: cameras (for analyzing gaze patterns and pedestrian flow), infrared sensors (for detecting dwell time), RFID or Wi-Fi probes (for tracking customer movement), and POS terminals (for recording purchasing behavior).

[0057] Historical sales data should include at least the following data types: product number and category, sales timestamp, sales quantity, sales amount, customer identification information (such as whether they are members), corresponding advertising display records (such as advertising time period and display frequency), and customer behavior records (such as whether they watched the advertisement, viewing duration, entry time and path).

[0058] The specific implementation process of each sub-step in step S200 is as follows:

[0059] Step S201 is achieved through product category tag matching technology, that is, automatically classifying new products and historical products based on the standard product classification system;

[0060] Step S202 uses image processing and behavior trajectory recognition technology to identify the advertising content that customers are looking at using camera video streams, and combines timestamps and POS system data to extract purchasing behavior;

[0061] Step S203 uses a behavioral ratio analysis model to calculate the ratio of the number of customers who immediately made a purchase after watching the advertisement to the total number of purchases during that period, and compares it with a set threshold.

[0062] Step S204 uses the full sub-period traversal logic to determine whether all time periods meet the preset behavior criteria. If they do, the historical new product is identified as the reference historical new product.

[0063] The criteria for defining the new product promotion stage can be: the date when the product was first listed is no more than a set number of days away from the current time (such as 30 days), and there is no history of large-scale distribution in the current location, or it is marked as a product in the promotion period.

[0064] The preset time division rules can be: dividing the nighttime business hours (such as 18:00~24:00) into equal-length time segments (such as one hour as a sub-period), or dynamically adjusting according to changes in customer traffic or advertising display density, such as a three-segment division of 18:00–19:30, 19:30–21:00, and 21:00–24:00.

[0065] The significance of identifying immediate decision-making purchasing behavior exceeding preset standards lies in its ability to reflect the immediate ability of the advertisement to stimulate purchase intentions during a specific sub-period, demonstrating the high compatibility and sales potential between the advertisement and the product. Based on this behavioral indicator, this invention identifies historically significant new products with strong responses, thereby enhancing the targeting and effectiveness of subsequent advertising configurations.

[0066] The setting of preset thresholds can be comprehensively judged by combining historical data statistical analysis and the actual needs of the promotion scenario to ensure its operability and representativeness. In specific implementation, the proportion of instant decision-making purchase behavior of similar new products in each nighttime sub-time period in the target advertising-driven business venues can be counted first. After removing outliers, a valid sample set is formed. Conversion rate distribution analysis is performed on this sample set to obtain statistical parameters such as median, mean, and standard deviation, forming basic reference values.

[0067] Building upon this, to more accurately identify historical samples with a significant purchase-driving effect in nighttime advertising environments, a preset threshold can be set at a value significantly higher than the baseline conversion level. For example, this could be 1.3 times the median, the average plus one standard deviation, or the 25th percentile of the conversion rate distribution as the identification threshold. These settings can be dynamically adjusted based on factors such as location traffic characteristics, product price sensitivity, and advertising intensity to adapt to the different screening needs for historical new products in various promotional scenarios.

[0068] Referencing historical new products is significant for the following reasons: they consistently achieved high conversion rates for immediate decision-making purchases across multiple sub-periods, demonstrating the strong driving effect of their advertising display strategies, visual element structures, and pacing. Using them as templates not only reduces the trial-and-error costs of new products but also provides empirically validated guidelines for setting display elements and frequencies, thereby guiding the formation of display strategies for new products to be promoted. This approach has significant practical value and promotional adaptability.

[0069] Furthermore, the intelligent analysis method for sales potential applied to new product promotion also includes the following steps:

[0070] In step S300, if there is a reference historical new product, obtain its advertising display content, and extract the key advertising display elements and corresponding expression intensity values ​​of the new product to be promoted and the reference historical new product in each nighttime sub-period.

[0071] The key advertising display elements and their corresponding expression intensity values ​​refer to:

[0072] Key advertising display elements refer to the key components in advertising display content that have a significant impact on guiding customer attention or forming purchase intentions, including at least one of the main image area, core copy fields, visual focus graphics, color contrast areas, or sound emphasis segments;

[0073] The expression intensity value refers to a numerical parameter used to characterize the prominence of the key advertising display element in the overall advertising display content. It is calculated based on one or more of the following factors: color saturation, image area ratio, font size, volume intensity, animation frequency, or appearance duration.

[0074] In this embodiment of the invention, the key advertising display elements and their corresponding expression intensity values ​​are an important indicator system for characterizing the internal structure of advertising content. They have clear definition standards and executable quantification methods, and can be implemented in actual advertising material analysis, with a mature technical foundation.

[0075] Key advertising display elements refer to visual or auditory units in the advertising display content that play a crucial guiding role in attracting customers' visual attention, forming their cognition, and stimulating their purchase intention. These elements typically include, but are not limited to, the following types: (1) main image area, such as close-up images of products, guiding images of people, and scene application images; (2) core copywriting fields, such as text content that highlights the brand, promotional activities, or core selling points; (3) visually focusing graphics, such as arrows that guide the eye, highlighted borders, or emphasis logos; (4) color contrast areas, i.e., color saturation or brightness that is significantly different from the surrounding areas, used to create a focal area for visual attention; and (5) sound emphasis segments, such as background sound effects and keyword voice prompts used to attract customers' attention. The above elements have been widely identified as important visual or auditory driving factors influencing customer behavior in advertising design and human-computer interaction research.

[0076] The expression intensity value is a numerical indicator used to quantify the "prominence" of each key advertising element within the overall advertising content. Its core purpose is to provide a unified and easily comparable metric to support subsequent tasks such as similarity assessment, frequency correction, and strategy optimization. This value can be obtained based on the following known and widely used image processing and audio signal analysis techniques:

[0077] (1) Image elements can be quantified using indicators such as color saturation, brightness contrast, area ratio, and edge sharpness. For example, after image segmentation, the area ratio of a certain graphic to the entire advertising area can be extracted, or the average saturation value of the color channel can be calculated;

[0078] (2) The expressive intensity of text elements can be determined based on font size, word count ratio, and positional weight (such as whether it is located in a prominent position);

[0079] (3) Dynamic elements can be modeled by their intensity of expression through animation frequency, number of flashes, or duration of appearance;

[0080] (4) Audio elements can be extracted using parameters such as volume peak, speech repetition rate, and rhythm beat density.

[0081] In practical applications, the extraction of these expression intensity values ​​relies on existing computer vision and audio recognition technologies, such as open-source libraries like OpenCV, TensorFlow, and FFmpeg. These are combined with deep learning models for semantic segmentation, object detection, and audio frame analysis, providing a mature implementation path and a broad engineering foundation. Therefore, this technical solution not only possesses theoretical support and engineering feasibility but also supports various intelligent decision-making and sales potential analysis tasks in the subsequent reconstruction of nighttime sub-segment advertising strategies.

[0082] Furthermore, the intelligent analysis method for sales potential applied to new product promotion also includes the following steps:

[0083] Step S400 involves referencing the key advertising display elements and historical display frequencies of new products in each nighttime sub-period, and converting them into key advertising display elements and initial display frequencies for the new products to be promoted, thereby constructing a set of key advertising display elements.

[0084] Specifically, Figure 3 A flowchart is shown for constructing a set of key advertising display elements.

[0085] Specifically, the key advertising display elements and historical display frequencies of new products in each nighttime sub-period will be referenced and converted into key advertising display elements and initial display frequencies for the new products to be promoted. The construction of the key advertising display element set includes the following steps:

[0086] Step S401: Extract the advertising display content of new products in each nighttime sub-period from historical sales data, and identify the corresponding key advertising display elements and historical display frequency.

[0087] Step S402: Replace the original key advertising display elements of the new products to be promoted in each nighttime sub-period with the key advertising display elements of the reference historical new products that are consistent with the key advertising display elements determined by structural hierarchy matching or semantic tag matching, and use the historical display frequency of the key advertising display elements of the reference historical new products as the initial benchmark value.

[0088] Step S403: Based on the similarity difference between the key advertising display elements of the referenced historical new products and the original key advertising display elements of the corresponding new products to be promoted, the initial benchmark value is corrected to obtain the initial display frequency of the new products to be promoted in the nighttime sub-period.

[0089] The historical display frequency refers to the number of times or the display duration of the key advertising element is repeatedly displayed in the corresponding nighttime sub-period, which is the proportion of the total display duration of the advertisement in that period.

[0090] In this embodiment of the invention, historical display frequency is used to describe the specific exposure intensity of a key advertising element within a corresponding nighttime sub-period, reflecting its actual dissemination weight within the target time period. The statistical object of historical display frequency is the advertising content of a reference historical new product within a single nighttime sub-period, including only the total advertising display duration of the reference historical new product itself, excluding advertising content from other products. Therefore, "total display duration" refers to the total advertising playback duration of the reference historical new product within that sub-period, rather than the cumulative duration of all product advertisements. This setting ensures that the frequency value remains consistent with the attribution logic of specific advertising elements, avoiding normalization bias caused by mixed statistical methods.

[0091] Within each nighttime sub-period, the ad creatives are derived from specific promotional strategies for historical new products. Their internal structure is stable, and the display elements are fixed. This means that all repeatedly played ads within the same sub-period use the same set of key display elements, ensuring the consistency and traceability of statistical data. The key ad display elements do not change with each individual play, allowing the accumulated display frequency to reflect the degree of long-term user exposure to these elements, thus providing a reliable basis for subsequent replacements and optimizations.

[0092] In step S402, although the original key advertising elements of the new product to be promoted have independent designs, in this invention, they are identified with high similarity to the display elements of a reference historical new product through "structural hierarchy matching or semantic tag matching" before their corresponding positions are replaced. Structural hierarchy matching refers to aligning the positions and functions of corresponding elements in two advertising contents based on the structured expression of the advertising content (such as image area → text field → color hierarchy → dynamic overlay); semantic tag matching uses a natural language processing model or image content recognition model to extract and compare whether the intent or meaning carried by the two sets of advertising elements is consistent. Only when the two achieve a high degree of consistency in both structural and semantic dimensions are they determined to be "consistent" to ensure the rationality and adaptability of the replacement process.

[0093] By replacing the original elements through the above matching process and referencing the historical display frequency of new products as the initial benchmark, it is possible to extract the most effective display combination for customer perception from samples with excellent actual sales performance. This combination can then be used to empower the advertising strategy of new products to be promoted, thereby reducing trial and error costs and increasing the hit rate of the first display optimization.

[0094] In step S403, the initial display frequency is further adjusted by introducing the "similarity difference" between the new product to be promoted and the reference historical new products in terms of the corresponding key advertising display elements. The core significance is that even if the structural and semantic similarity meets the standard, due to factors such as product image, brand background, or visual style, the same elements of the new product to be promoted may still have perceptual deviations in actual presentation. To avoid "overfitting" caused by directly using historical frequencies, this step makes compensatory adjustments by introducing expression difference weights.

[0095] Calculating similarity differences can employ various mature technical approaches. For example, structural similarity index (SSIM), cosine similarity of feature vectors, or spatial distance of deep embeddings can be used between image features; Euclidean distance after BERT embedding or Jaccard coefficients can be used for text fields; and dynamic elements can be weighted and fused by combining playback rhythm, duration, and animation style. Ultimately, the degree of deviation forms a multiplicative correction relationship with the baseline frequency, i.e.:

[0096] The initial display frequency can be obtained by adjusting the historical display frequency of key advertising elements in the reference historical new products by multiplying the historical display frequency by a reduction factor related to the expression difference of the corresponding display element in the new product to be promoted. The reduction factor is equal to "1 minus the product of the difference and the sensitivity factor", where the difference is used to characterize the relative difference between the two display elements in terms of structure, semantics or visual expression, and the sensitivity factor is used to adjust the overall impact of the adjustment.

[0097] The sensitivity factor can be adjusted based on historical samples, and the difference is the normalized inverse of the similarity index mentioned above. This approach maintains the reference value of empirical frequencies while adapting to the expressive characteristics of the new product's content, ensuring that the initial display frequency has a stronger personalized adaptability.

[0098] Furthermore, the intelligent analysis method for sales potential applied to new product promotion also includes the following steps:

[0099] Step S500: Based on the deviation of the expression intensity values ​​of the new product to be promoted and the reference historical new products in the corresponding key advertising display elements, adjust the initial display frequency of each key advertising display element in the set of key advertising display elements, and generate optimized advertising display content for the new product to be promoted.

[0100] Specifically, Figure 4 A flowchart is shown to adjust the display frequency based on the deviation of the expression intensity.

[0101] The process of generating optimized ad display content for the new product to be promoted, based on the deviation between the expression intensity values ​​of the new product to be promoted and historical new products in the corresponding key ad display elements, includes the following steps:

[0102] Step S501: Quantify the deviation of the expression intensity values ​​of the new product to be promoted and the reference historical new products in the corresponding key advertising display elements in each night sub-period, and use the deviation as a correction factor to form a set of correction factors;

[0103] Step S502: Use each correction factor to correct the initial display frequency of the key advertising display elements of the new product to be promoted, and obtain the corrected display frequency.

[0104] Step S503: Recombine all the corrected display frequencies with the corresponding key advertising display elements to form the optimized advertising display content of the new product to be promoted; the optimized advertising display content serves as the input basis for subsequent intelligent analysis of the sales potential of the new product to be promoted, and is used to improve the accuracy of advertising response prediction in different nighttime sub-segments.

[0105] The relative expression intensity deviation rate is obtained by subtracting the expression intensity value of the new product to be promoted from the expression intensity value of the corresponding key advertising display elements of the reference historical new product, and dividing the difference by the expression intensity value of the reference historical new product. This difference is used as a correction factor.

[0106] In this embodiment of the invention, the calculation method of using the deviation of the expression intensity value as a correction factor is of great significance. By subtracting the corresponding expression intensity value of the new product to be promoted from the expression intensity value of the historical new product on the corresponding key advertising display elements, and using the expression intensity value of the historical new product as the denominator, the resulting relative expression intensity deviation rate can effectively characterize the difference between the visual or semantic prominence of the new product to be promoted on the corresponding advertising elements and that of successful cases. This relative difference not only reflects the possible weakening of the advertising expression effect, but also provides a quantitative basis for adjusting the display frequency, thereby more accurately simulating the successful dissemination pattern of the historical new product.

[0107] When the expression intensity value of a key advertising element of a reference new product is 0, in order to avoid division by zero error, you can choose to skip the element and not participate in the correction, or set its expression intensity value to a preset minimal constant to complete the calculation, thereby ensuring the stability and feasibility of the correction process.

[0108] In addition to the methods mentioned above, cosine similarity, Euclidean distance normalized difference, or vector angles based on semantic embedding from neural networks can also be used to quantify differences in expression intensity. These methods can be flexibly selected according to specific implementation environments to improve matching accuracy and adaptability to practical applications.

[0109] The deviation essentially reflects the degree of similarity between the new product to be promoted and a reference historical new product in terms of visual, copywriting, or graphic design aspects of advertising content. Adjusting accordingly can improve the fit of the display strategy, thereby effectively enhancing the actual response potential of the new product to be promoted in the same advertising environment and avoiding the waste of advertising resources due to weak expression. Especially when customer attention, behavioral reaction speed, and target preferences differ across different nighttime sub-segments, this mechanism can achieve personalized adaptation at the sub-segment level, greatly enhancing the dynamism and accuracy of the advertising display strategy.

[0110] The process of generating optimized ad display content in step S503 is a crucial step in the entire intelligent analysis of sales potential. Its function is to combine the previously revised display frequency with corresponding key ad display elements into a feasible ad configuration plan, forming a comprehensive configuration set including parameters such as display content, display ratio, and display order. This optimized content not only serves as the input basis for subsequent simulation evaluation and prediction of the sales potential of the new product to be promoted, but can also be directly deployed in real advertising scenarios, further iterating and optimizing strategy parameters using actual response data. Therefore, this step not only improves the accuracy and responsiveness of the prediction model, but also provides a real-time adaptable execution template for advertising placement decisions, demonstrating excellent practical application prospects.

[0111] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0112] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0113] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0114] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

[0115] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent analysis of sales potential applied to new product promotion, characterized in that, The method includes: Obtain historical sales data for target advertising-driven retail locations, as well as advertising display content for new products to be promoted; Analyze historical sales data to identify whether there are any reference historical new products that are similar to the new products to be promoted, are in the new product promotion stage, and have shown real-time decision-making purchase behaviors exceeding the preset standards in several nighttime sub-periods. If there are referenced historical new products, obtain their advertising display content, and extract the key advertising display elements and corresponding expression intensity values ​​of the new product to be promoted and the referenced historical new products in each nighttime sub-period. The key advertising display elements and historical display frequencies of new products in each nighttime sub-period will be converted into key advertising display elements and initial display frequencies for the new products to be promoted, thus constructing a set of key advertising display elements; Based on the deviation of the expression intensity values ​​of the new product to be promoted from those of historical new products in the corresponding key advertising display elements, the initial display frequency of each key advertising display element set is adjusted to generate optimized advertising display content for the new product to be promoted.

2. The intelligent analysis method for sales potential applied to new product promotion according to claim 1, characterized in that, The aforementioned advertising-driven retail space refers to a retail space that uses advertising displays as the primary means of product promotion and is equipped with visual display areas for customers, advertising placement devices, and sensing devices for collecting customer behavior data.

3. The intelligent analysis method for sales potential applied to new product promotion according to claim 2, characterized in that, The division of the nighttime sub-periods refers to dividing the nighttime period into several sub-periods within the nighttime operating hours of the target advertising-driven business venue according to a preset time division rule, in order to capture the behavioral response characteristics of customers to the advertising content displayed in different time periods.

4. The intelligent analysis method for sales potential applied to new product promotion according to claim 3, characterized in that, The steps for analyzing historical sales data to identify reference products that are similar to the new product to be promoted, are in the new product promotion phase, and have shown immediate decision-making purchasing behavior exceeding preset standards in several nighttime sub-periods include: Determine the product category of the new product to be promoted, and filter the sales data of several historical new products that have the same product category and are in the new product promotion stage from the historical sales data; For each historical new product, extract image information of customers viewing the visual display area and their purchasing behavior information during each nighttime sub-period. Analyze the instant decision-making purchasing behavior of customers in each nighttime sub-period and determine whether it exceeds the preset standard. The instant decision-making purchasing behavior that exceeds the preset standard refers to the proportion of customers who enter the store and complete the purchase after viewing the historical new product advertisement to the total number of historical new product purchases in that nighttime sub-period that exceeds the preset threshold. If there are instant decision-making purchase behaviors exceeding the preset standard in all nighttime sub-periods, then the historical new product is determined to be a reference historical new product.

5. The intelligent analysis method for sales potential applied to new product promotion according to claim 4, characterized in that, If more than one historical new product is identified as a reference, then based on the pricing, usage, and other product attributes of the new product to be promoted, the one that is closest to the new product to be promoted will be selected as the final historical new product to be referenced.

6. The intelligent analysis method for sales potential applied to new product promotion according to claim 1, characterized in that, The key advertising display elements and their corresponding expression intensity values ​​refer to: Key advertising display elements refer to the key components in advertising display content that have a significant impact on guiding customer attention or forming purchase intentions, including at least one of the main image area, core copy fields, visual focus graphics, color contrast areas, or sound emphasis segments; The expression intensity value refers to a numerical parameter used to characterize the prominence of the key advertising display element in the overall advertising display content. It is calculated based on one or more of the following factors: color saturation, image area ratio, font size, volume intensity, animation frequency, or appearance duration.

7. The intelligent analysis method for sales potential applied to new product promotion according to claim 1, characterized in that, The steps for constructing a set of key advertising display elements include: referencing the key advertising display elements and historical display frequencies of historical new products in each nighttime sub-hour period, and converting them into key advertising display elements and initial display frequencies for the new product to be promoted. Extract the advertising content of historical new products in each nighttime sub-period from historical sales data, and identify the corresponding key advertising elements and historical display frequency; The original key advertising display elements of the new products to be promoted in each nighttime sub-period are replaced with the key advertising display elements of the reference historical new products that are consistent through structural hierarchy matching or semantic tag matching, and the historical display frequency of the key advertising display elements of the reference historical new products is used as the initial benchmark value. Based on the similarity difference between the key advertising display elements of the referenced historical new products and the original key advertising display elements of the corresponding new products to be promoted, the initial benchmark value is corrected to obtain the initial display frequency of the new products to be promoted in that nighttime sub-period.

8. The intelligent analysis method for sales potential applied to new product promotion according to claim 7, characterized in that, The historical display frequency refers to the number of times or the display duration of the key advertising element is repeatedly displayed in the corresponding nighttime sub-period, which is the proportion of the total display duration of the advertisement in that period.

9. The intelligent analysis method for sales potential applied to new product promotion according to claim 7, characterized in that, The steps for generating optimized ad content for the new product to be promoted, based on the deviation between the expression intensity values ​​of the new product to be promoted and historical new products in the corresponding key ad display elements, include: The deviation of the expression intensity values ​​of the new product to be promoted and the reference historical new products in the corresponding key advertising display elements is quantified in each night sub-period, and the deviation is used as a correction factor to form a set of correction factors; By using each correction factor to adjust the initial display frequency of the key advertising display elements of the new product to be promoted, the corrected display frequency can be obtained. All the revised display frequencies are recombined with the corresponding key advertising display elements to form the optimized advertising display content for the new product to be promoted. The optimized advertising display content serves as the input basis for subsequent intelligent analysis of the sales potential of the new product to be promoted, and is used to improve the accuracy of advertising response prediction in different nighttime sub-segments.

10. The intelligent analysis method for sales potential applied to new product promotion according to claim 9, characterized in that, The relative expression intensity deviation rate is obtained by subtracting the expression intensity value of the new product to be promoted from the expression intensity value of the corresponding key advertising display elements of the reference historical new product, and dividing the difference by the expression intensity value of the reference historical new product. This difference is used as a correction factor.

Images