Commodity marketing strategy optimization system and method based on multi-source data analysis

Abstract

The invention discloses a commodity marketing strategy optimization system and method based on multi-source data analysis, and relates to the technical field of commodity marketing strategy optimization, and the method comprises the steps: S10, carrying out the analysis of an effective browsing frame track of a customer, and obtaining the real-time marketing data of a target commodity under a current marketing strategy; s20, selecting marketing content to be optimized in the current marketing strategy of the target commodity; s30, determining an associated frame of the to-be-optimized marketing content of each frame; and S40, determining an optimization scheme of the current marketing strategy of the target commodity. According to the method, the to-be-optimized condition of the current marketing strategy of the target commodity is determined based on the marketing data, frame-by-frame analysis of the marketing strategy is achieved, and the optimization effect of the marketing strategy is improved. The marketing strategy obtained according to the optimization scheme can be highly accepted by customers.

Description

Technical Field

[0001] The present invention relates to the technical field of commodity marketing strategy optimization, and in particular to a commodity marketing strategy optimization system and method based on multi-source data analysis. Background Art

[0002] As the e-commerce market continues to expand, consumer shopping behavior has gradually become more diversified and personalized. Therefore, e-commerce companies are seeking to use data analysis technology to tap into potential business opportunities, enhance user experience, and optimize marketing strategies.

[0003] At present, the main e-commerce marketing strategies include search engine marketing, social media marketing, content marketing, live streaming marketing and community marketing. For live streaming marketing and content marketing, due to the rapid change trend of product sales process data, existing technologies are unable to adaptively adjust and optimize the product marketing strategies in a short period of time. The determined optimization plan is only achieved by increasing or decreasing the marketing of corresponding marketing content, which fails to fundamentally optimize the marketing strategy and thus cannot guarantee the customer acceptance of the optimized marketing strategy. Summary of the Invention

[0004] The purpose of the present invention is to provide a product marketing strategy optimization system and method based on multi-source data analysis to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing product marketing strategies based on multi-source data analysis, the method comprising:

[0006] S10: Sorting out the sales process data of the target product during the marketing period to obtain the effective browsing frame trajectory of each customer. Analyzing the effective browsing frame trajectory of each customer to obtain the real-time marketing data of the target product under the current marketing strategy. The marketing data includes customer retention rate, marketing content reach rate, and marketing content value density.

[0007] S20: Analyze the optimization index of the current marketing strategy of the target product based on the marketing data, and select the marketing content to be optimized in the current marketing strategy of the target product according to the analysis results;

[0008] S30: Matrixing the real-time marketing content value index of each frame of marketing content in the marketing video, analyzing the correlation between the marketing content of each frame, and determining the associated frames of the marketing content to be optimized for each frame based on the analysis results;

[0009] S40: Determine an optimization plan for the current marketing strategy of the target product.

[0010] Furthermore, the S10 includes:

[0011] S101: Generate a marketing video based on the current marketing strategy of the target product and distribute the marketing video on various marketing channels. Obtain the marketing time period [t, T] of the target product under the current marketing strategy. Collect orders placed for the target product within the marketing time period [t, T], records of marketing video views within the marketing time period [t, T], and the life cycle of the target product.

[0012] S102: Orders placed for the target product within the marketing time period [t, T] and each customer's browsing history of the marketing video within the marketing time period [t, T] are sorted to obtain each customer's valid browsing frames. The valid browsing frames of each customer are sorted in chronological order to generate a valid browsing frame trajectory for each customer. This allows for screening of target customers, thereby narrowing the scope of analysis of the sales process data.

[0013] S103: Numbering the customers with valid browsing frame traces, the numbering result is: i=1, 2, ..., m, where m represents the total number of customers with valid browsing frame traces;

[0014] In the effective browsing frame trajectory of customer i, the number of order interfaces for the target product h i To obtain, when h i When ≥2, in the effective browsing frame trajectory of customer i, the interval time between the order times displayed in the order interface of two adjacent target products is obtained to obtain the interval time set. The elements stored in the interval time set are processed according to the average value calculation method to obtain U i ;

[0015] If U i ≤u+R, then the retention index of customer i is f i =1, if U i >u+R, then the retention index of customer i is f i =1 / (U i -u+R), where R represents the life cycle fluctuation value of the target product, and u represents the life cycle of the target product; according to H iT =1-exp[(1-h i )*f i ] Predict the retention rate of the target product for customer i at time T, from i = 1 to i = m for all H iT Perform summation to obtain H′ iT , for H′ iT The customer retention rate U of the target product at time T is calculated by calculating the ratio between the target product and m. T , where exp() represents an exponential function with base e and e=2.73;

[0016] Based on the overlap of each customer's effective browsing frame trajectory, and the overlap of each customer's effective browsing frame trajectory with the target browsing frame trajectory, the real-time marketing content reach rate of the marketing video and the marketing content value density of each frame of the marketing content in the marketing video are analyzed.

[0017] Analyzing the marketing data of the target product under the current marketing strategy based on the customer's effective browsing frame trajectory is conducive to intuitive analysis of the sales process of the target product and intuitive feedback from customers on their acceptance of the current marketing strategy.

[0018] Furthermore, the specific method of analyzing the real-time marketing content reach rate of the marketing video and the marketing content value density of each frame of the marketing content in the marketing video in S103 is as follows:

[0019] Number each frame of marketing content according to the order in which it is displayed in the marketing video. The numbering result is: j = 1, 2, ..., n, where n represents the total number.

[0020] Compare the effective browsing frame trajectories of each customer, and based on the comparison results, calculate the number of views S of the marketing content of the jth frame in the marketing video at time T. jT Obtain; compare the effective browsing frame trajectory of each customer with the target browsing frame trajectory of the marketing video, and based on the comparison results, calculate the total number of views K of the target browsing frame in the marketing video at time T T To obtain;

[0021] K T Calculate the ratio between the target product and m to get the marketing content reach rate C of the target product at time T. T ;

[0022] To S jT The ratio between the value of W and m is calculated to obtain the browsing density W of the marketing content of the jth frame in the marketing video of the target product at time T. jT In the effective browsing frame trajectory of customer i, the minimum number of marketing content frames G between the order interface of each target product and the marketing content of the jth frame in the marketing video is ij To determine, if G ij = 0, then the value index V of the marketing content of the jth frame in the marketing video at time T ijT =1, if G ij ≠0, then the value index V of the marketing content of the jth frame in the marketing video at time T ijT =1 / G ij , from i = 1 to i = m for all V ijT Perform summation to obtain V jT , for W jT With V jT / m to calculate the product of the marketing content value density M of the marketing content of the jth frame in the marketing video of the target product at time T. jT .

[0023] Furthermore, the S20 includes:

[0024] S201: According to W T =a*U T +b*C T +c*∑{M jT *r j *[1-exp(-j)]} calculates the optimization index of the current marketing strategy of the target product at time T, where ∑ represents the summation symbol, the subscript of ∑ is j=1, the superscript is n, and r j Represents the weight coefficient corresponding to the marketing content of the jth frame in the marketing video;

[0025] When 0.6<W T ≤1, it is necessary to optimize the current marketing strategy of the target product. T When ≤0.6, there is no need to optimize the current marketing strategy of the target product;

[0026] S202: When 0.6<W T ≤1, set M jT *r j *[1-exp(-j)] is compared with the set threshold. If M jT *r j *[1-exp(-j)]≤set threshold, then it is considered that the marketing content of the jth frame in the marketing video needs to be optimized. If M jT *r j *[1-exp(-j)]>set threshold, then it is considered that the marketing content of the jth frame in the marketing video does not need to be optimized.

[0027] The selection of marketing content to be optimized eliminates the influence of the playback order of each frame of marketing content on the optimization analysis result, which is conducive to improving the analysis accuracy of each frame of marketing content to be optimized.

[0028] Furthermore, the S30 includes:

[0029] S301: Matrix the real-time marketing content value index of each frame of marketing content in the marketing video to obtain the matrix U T , Wherein, p=1,2,…,q represents the number corresponding to the number of times sales process data is collected, p represents the total number of times sales process data is collected, and d represents the collection interval of sales process data;

[0030] According to the matrix U T, the correlation coefficient γ between the marketing content of the xth frame in the marketing video and the marketing content of the jth frame in the marketing video xj Perform calculations, x=1,2,…,n and x≠j;

[0031] S302: Randomly select a frame of marketing content to be optimized, and record the frame number corresponding to the selected marketing content as g, where g = 1, 2, ..., n. If γ gj > 0.8 and the marketing content of the jth frame in the marketing video is the marketing content to be optimized, then the marketing content of the jth frame in the marketing video is used as the associated frame of the marketing content of the gth frame in the marketing video. If γ gj ≤0.8 and the marketing content of the jth frame in the marketing video is the marketing content to be optimized, the marketing content of the jth frame in the marketing video will not be used as the associated frame of the marketing content of the gth frame in the marketing video.

[0032] The associated frames obtained through matrix analysis are conducive to the classification and optimization of the marketing content to be optimized, thereby shortening the optimization and adjustment time of the marketing content to be optimized.

[0033] Furthermore, the S40 includes:

[0034] When the marketing content in the g-th frame of a marketing video has an associated frame:

[0035] Classify the marketing content of the g-th frame in the marketing video and the marketing content of each associated frame of the g-th frame in the marketing video into one category of marketing content;

[0036] When the marketing content in the g-th frame of a marketing video does not have an associated frame:

[0037] The marketing content of the gth frame in the marketing video is regarded as a type of marketing content;

[0038] Randomly select a category of marketing content and a frame to be browsed, and calculate the text similarity between the selected category of marketing content and the marketing content corresponding to the selected frame to be browsed;

[0039] If the calculated text similarity is higher than the target threshold and the selected frame displays marketing content that demonstrates the brand image or the target product's functions, the marketing content corresponding to the selected frame is optimized based on the dimensional shortcomings of the selected marketing content. The specific optimization method is as follows:

[0040] The marketing data of each frame of the selected marketing content is obtained during playback and represented in a five-dimensional graph. Based on the dispersion coefficient and mean of each type of marketing data on the corresponding coordinate axis, the dimensional weaknesses of the selected marketing content are determined. Dimensional weaknesses refer to marketing data types with a dispersion coefficient greater than 0.7 and a mean less than 0.6.

[0041] If the weak point of the dimension is customer retention rate, provide customers with after-sales service based on their evaluation of the target product, and add the provided after-sales solution to the selected frame to be browsed;

[0042] If the shortcoming is the marketing content reach, increase the frequency of marketing videos;

[0043] If the weak point in the dimension is the value density of marketing content, the playback order of the selected browsing frames is optimized so that the optimized selected browsing frames are played before the marketing content with a value index of 1;

[0044] If the calculated text similarity is higher than the target threshold and the selected frame to be browsed does not display marketing content that shows the brand image or the function demonstration of the target product, the selected frame to be browsed will be deleted from the marketing video;

[0045] If the calculated text similarity is lower than the target threshold, the selected frame to be browsed is retained in the marketing video.

[0046] By measuring the text similarity between marketing contents, we can selectively identify the dimensional shortcomings of marketing contents and achieve specialized optimization of marketing contents. The optimization process is based on the customer's viewing experience, ensuring that the optimized marketing strategy can adapt to the ever-changing market environment and maximize the marketing effect of the target products.

[0047] During the optimization process, it is necessary to ensure the smoothness of the marketing content in each frame of the marketing video.

[0048] A product marketing strategy optimization system based on multi-source data analysis, the system comprising a marketing data analysis module, a module for selecting marketing content to be optimized, a module for finding associated frames, and a module for generating an optimization solution;

[0049] The marketing data analysis module is used to analyze the effective browsing frame trajectory of the customer and obtain the real-time marketing data of the target product under the current marketing strategy;

[0050] The module for selecting marketing content to be optimized is used to select marketing content to be optimized in the current marketing strategy of the target product;

[0051] The associated frame search module is used to determine the associated frames of each frame of marketing content to be optimized;

[0052] The solution generation module is used to determine the optimization solution for the current marketing strategy of the target product.

[0053] Furthermore, the marketing data analysis module includes a collection unit, an effective browsing trajectory generation unit, a retention rate calculation unit, a marketing content reach rate calculation unit, and a marketing content value density calculation unit;

[0054] The collection unit collects orders placed for the target product within the marketing period, records of marketing videos viewed within the marketing period, and the life cycle of the target product;

[0055] The effective browsing trajectory generation unit sorts out the orders placed for the target product during the marketing period and the browsing records of the marketing videos by each customer during the marketing period to obtain the effective browsing frame trajectory of each customer;

[0056] The retention rate calculation unit obtains the interval between the order times displayed in two adjacent target product ordering interfaces in the effective browsing frame trajectory of each customer when the set conditions are met, and constructs a calculation model based on the obtained results to calculate the real-time customer retention rate of the target product;

[0057] The marketing content reach rate calculation unit calculates the real-time marketing content reach rate of the marketing video based on the overlap of each customer's effective browsing frame trajectory;

[0058] The marketing content value density calculation unit calculates the marketing content value density of each frame of marketing content in the marketing video based on the overlap between the effective browsing frame trajectory of each customer and the target browsing frame trajectory.

[0059] Furthermore, the marketing content selection module to be optimized includes an optimization index calculation unit and an optimization judgment unit;

[0060] The optimized index calculation unit calculates the real-time optimized index of the current marketing strategy of the target product according to the constructed mathematical model;

[0061] The optimization judgment unit judges whether it is necessary to optimize the marketing content of each frame in the marketing video according to the set judgment conditions.

[0062] Furthermore, the associated frame search module includes a matrix processing unit, an associated coefficient calculation unit and a key frame search unit;

[0063] The matrix processing unit matrices the real-time marketing content value index of each frame of marketing content in the marketing video to obtain a target matrix;

[0064] The correlation coefficient calculation unit calculates the correlation coefficient between two random frames of marketing content in the marketing video according to the target matrix;

[0065] The associated frame searching unit searches for associated frames of each frame of the marketing content to be optimized based on the calculation result transmitted by the associated coefficient calculating unit.

[0066] Compared with the prior art, the present invention has the following beneficial effects:

[0067] 1. The present invention generates effective browsing frame trajectories of customers through the sales process data of the target product and obtains the marketing data of the target product. The marketing data can reflect the sales situation of the target product from multiple dimensions. Based on the marketing data, the optimization status of the current marketing strategy of the target product is determined, and a frame-by-frame analysis of the marketing strategy is achieved, thereby improving the optimization effect of the marketing strategy.

[0068] 2. The present invention matrixes the real-time marketing content value index of each frame of marketing content in the marketing video, and calculates the correlation coefficient between two random frames of marketing content in the marketing video. The calculation process eliminates the influence of the display order of each frame of marketing content on the calculation result. Based on the calculation result, the associated frames of each frame of marketing content to be optimized are searched, thereby realizing the classified optimization of various types of marketing content and improving the optimization adjustment rate of marketing strategies.

[0069] 3. The marketing strategy obtained by the present invention based on the optimization scheme can be highly accepted by customers, rather than similar optimization based on existing popular marketing strategies or simple marketing content screening. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 This is a workflow diagram of a product marketing strategy optimization method based on multi-source data analysis according to the present invention. DETAILED DESCRIPTION

[0071] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative work shall fall within the scope of protection of the present invention.

[0072] like Figure 1 As shown, the present invention provides a system and method for optimizing product marketing strategies based on multi-source data analysis, and a method for optimizing product marketing strategies based on multi-source data analysis, the method comprising:

[0073] S10: Sorting out the sales process data of the target product during the marketing period to obtain the effective browsing frame trajectory of each customer. Analyzing the effective browsing frame trajectory of each customer to obtain the real-time marketing data of the target product under the current marketing strategy. The marketing data includes customer retention rate, marketing content reach rate, and marketing content value density.

[0074] The S10 includes:

[0075] S101: Generate a marketing video based on the current marketing strategy of the target product. The marketing video includes advertising videos, live broadcast videos, etc., and distribute the marketing video on various marketing channels. Obtain the marketing time period [t, T] of the target product under the current marketing strategy. Collect the target product's orders placed within the marketing time period [t, T], the marketing video's viewing records within the marketing time period [t, T], and the target product's life cycle. The sales process data includes the target product's orders placed within the marketing time period and the marketing video's viewing records within the marketing time period.

[0076] S102: Orders placed for the target product within the marketing time period [t, T] and each customer's browsing history of the marketing video within the marketing time period [t, T] are sorted to obtain each customer's valid browsing frames. The valid browsing frames include the marketing content interface browsed by the customer in the marketing video, the target product sales interface accessed by the customer through the marketing video while browsing the marketing video, and the target product order interface. The valid browsing frames of each customer are sorted in chronological order to generate a valid browsing frame trajectory for each customer.

[0077] S103: Numbering the customers with valid browsing frame traces, the numbering result is: i=1, 2, ..., m, where m represents the total number of customers with valid browsing frame traces;

[0078] In the effective browsing frame trajectory of customer i, the number of order interfaces for the target product h i To obtain, when h i When ≥2, in the effective browsing frame trajectory of customer i, the interval time between the order times displayed in the order interface of two adjacent target products is obtained to obtain the interval time set. The elements stored in the interval time set are processed according to the average value calculation method to obtain U i ;

[0079] If U i ≤u+R, then the retention index of customer i is f i =1, if U i >u+R, then the retention index of customer i is f i =1 / (U i -u+R), where R represents the life cycle fluctuation value of the target product, and u represents the life cycle of the target product; according to H iT =1-exp[(1-h i )*f i ] Predict the retention rate of the target product for customer i at time T, from i = 1 to i = m for all H iT Perform summation to obtain H′ iT , for H′ iTThe customer retention rate U of the target product at time T is calculated by calculating the ratio between the target product and m. T , where exp() represents an exponential function with base e and e=2.73;

[0080] Based on the overlap of each customer's effective browsing frame trajectory, as well as the overlap between each customer's effective browsing frame trajectory and the target browsing frame trajectory, we analyze the real-time marketing content reach rate of the marketing video and the marketing content value density of each frame of the marketing content in the marketing video. The specific method is as follows:

[0081] Number each frame of marketing content according to the order in which it is displayed in the marketing video. The numbering result is: j = 1, 2, ..., n, where n represents the total number.

[0082] Compare the effective browsing frame trajectories of each customer, and based on the comparison results, calculate the number of views S of the marketing content of the jth frame in the marketing video at time T. jT Obtain; compare the effective browsing frame trajectory of each customer with the target browsing frame trajectory of the marketing video, and based on the comparison results, calculate the total number of views K of the target browsing frame in the marketing video at time T T The target browsing frame refers to the marketing content interface that displays the brand image and target product function demonstration in the marketing video;

[0083] K T Calculate the ratio between the target product and m to get the marketing content reach rate C of the target product at time T. T ;

[0084] To S jT The ratio between the value of W and m is calculated to obtain the browsing density W of the marketing content of the jth frame in the marketing video of the target product at time T. jT In the effective browsing frame trajectory of customer i, the minimum number of marketing content frames G between the order interface of each target product and the marketing content of the jth frame in the marketing video is ij To determine, if G ij = 0, then the value index V of the marketing content of the jth frame in the marketing video at time T ijT =1, if G ij ≠0, then the value index V of the marketing content of the jth frame in the marketing video at time T ijT =1 / G ij , from i = 1 to i = m for all V ijT Perform summation to obtain V jT , for W jT With V jT / m to calculate the product of the marketing content value density M of the marketing content of the jth frame in the marketing video of the target product at time T. jT ;

[0085] S20: Analyze the optimization index of the current marketing strategy of the target product based on the marketing data, and select the marketing content to be optimized in the current marketing strategy of the target product according to the analysis results;

[0086] The S20 includes:

[0087] S201: According to W T =a*U T +b*C T +c*∑{M jT *r j *[1-exp(-j)]} calculates the optimization index of the current marketing strategy of the target product at time T, where ∑ represents the summation symbol, the subscript of ∑ is j=1, the superscript is n, and r j Represents the weight coefficient corresponding to the marketing content of the jth frame in the marketing video;

[0088] When 0.6<W T ≤1, it is necessary to optimize the current marketing strategy of the target product. T When ≤0.6, there is no need to optimize the current marketing strategy of the target product;

[0089] S202: When 0.6<W T ≤1, M jT *r j *[1-exp(-j)] is compared with the set threshold. If M jT *r j *[1-exp(-j)]≤set threshold, then it is considered that the marketing content of the jth frame in the marketing video needs to be optimized. If M jT *r j *[1-exp(-j)]>set threshold, then it is considered that the marketing content of the jth frame in the marketing video does not need to be optimized;

[0090] S30: Matrixing the real-time marketing content value index of each frame of marketing content in the marketing video, analyzing the correlation between the marketing content of each frame, and determining the associated frames of the marketing content to be optimized for each frame based on the analysis results;

[0091] The S30 includes:

[0092] S301: Matrix the real-time marketing content value index of each frame of marketing content in the marketing video to obtain the matrix U T , Where p = 1, 2, ..., q represents the number corresponding to the number of times the sales process data is collected, p represents the total number of times the sales process data is collected, d represents the collection interval of the sales process data, and time Tq*d is after time t, where M jT*[1-exp(-j)] represents the marketing content value index of the j-th frame of the marketing video at time T;

[0093] According to the matrix U T , the correlation coefficient γ between the marketing content of the xth frame in the marketing video and the marketing content of the jth frame in the marketing video xj To calculate, the specific calculation method of the correlation coefficient is: for the matrix U T First, calculate the mean of each column, then subtract the mean from each column to obtain the centered data, then calculate the standard deviation of each column, and then divide each centered column by the standard deviation of the corresponding column to obtain the standardized matrix Z. Calculate the transpose of matrix Z, multiply matrix Z by the transpose of matrix Z, and then divide by q to obtain the correlation coefficient matrix of matrix Z. Based on the correlation coefficient matrix, obtain the correlation coefficient between two random frames of marketing content in the marketing video, x = 1, 2, ..., n and x ≠ j;

[0094] S302: Randomly select a frame of marketing content to be optimized, and record the frame number corresponding to the selected marketing content as g, where g = 1, 2, ..., n. If γ gj > 0.8 and the marketing content of the jth frame in the marketing video is the marketing content to be optimized, then the marketing content of the jth frame in the marketing video is used as the associated frame of the marketing content of the gth frame in the marketing video. If γ gj If the value is ≤0.8 and the marketing content of the jth frame in the marketing video is the marketing content to be optimized, the marketing content of the jth frame in the marketing video will not be used as the associated frame of the marketing content of the gth frame in the marketing video;

[0095] S40: Determine the optimization plan for the current marketing strategy of the target product;

[0096] The S40 includes:

[0097] When the marketing content in the g-th frame of a marketing video has an associated frame:

[0098] Classify the marketing content of the g-th frame in the marketing video and the marketing content of each associated frame of the g-th frame in the marketing video into one category of marketing content;

[0099] When the marketing content in the g-th frame of a marketing video does not have an associated frame:

[0100] The marketing content of the gth frame in the marketing video is regarded as a type of marketing content;

[0101] Randomly select a category of marketing content and a to-be-browsed frame, and calculate the text similarity between the selected category of marketing content and the marketing content corresponding to the selected to-be-browsed frame. The text similarity calculation method belongs to the prior art. The to-be-browsed frame refers to the marketing content interface in the marketing video that has not been viewed by the customer.

[0102] If the calculated text similarity is higher than the target threshold and the selected frame displays marketing content that demonstrates the brand image or the target product's functions, the marketing content corresponding to the selected frame is optimized based on the dimensional shortcomings of the selected marketing content. The specific optimization method is as follows:

[0103] The marketing data of each frame of the selected marketing content is obtained during playback, and the obtained marketing data is represented in a five-dimensional graph. Based on the dispersion coefficient and mean of each type of marketing data on the corresponding coordinate axes, the calculation methods of the dispersion coefficient and mean are all existing technologies. The dimensional shortcomings of the selected marketing content are determined. The dimensional shortcomings refer to marketing data types with a dispersion coefficient greater than 0.7 and a mean less than 0.6.

[0104] The coordinate axes of the five-dimensional graph include x, y, z, w, and v. In five-dimensional space, an object can be described by five coordinate values, corresponding to x, y, z, w, and v. For example, customer retention rate is represented on the x-axis, marketing content reach is represented on the z-axis, and marketing content value density is represented on the v-axis.

[0105] If the weak point of the dimension is customer retention rate, provide customers with after-sales service based on their evaluation of the target product, and add the provided after-sales solution to the selected frame to be browsed;

[0106] If the shortcoming is the marketing content reach, increase the frequency of marketing videos;

[0107] If the weak point in the dimension is the value density of marketing content, the playback order of the selected browsing frames is optimized so that the optimized selected browsing frames are played before the marketing content with a value index of 1;

[0108] If the calculated text similarity is higher than the target threshold and the selected frame to be browsed does not display marketing content that shows the brand image or the function demonstration of the target product, the selected frame to be browsed will be deleted from the marketing video;

[0109] If the calculated text similarity is lower than the target threshold, the selected frame to be browsed is retained in the marketing video.

[0110] A product marketing strategy optimization system based on multi-source data analysis, the system includes a marketing data analysis module, a module for selecting marketing content to be optimized, a module for finding related frames, and a module for generating optimization solutions;

[0111] The marketing data analysis module is used to analyze the customer's effective browsing frame trajectory and obtain the real-time marketing data of the target product under the current marketing strategy;

[0112] The marketing data analysis module includes a collection unit, an effective browsing trajectory generation unit, a retention rate calculation unit, a marketing content reach rate calculation unit, and a marketing content value density calculation unit;

[0113] The collection unit collects orders placed for the target product during the marketing period, records of marketing videos viewed during the marketing period, and the life cycle of the target product;

[0114] The effective browsing trajectory generation unit sorts out the orders placed for the target product during the marketing period, as well as the browsing records of the marketing videos by each customer during the marketing period, to obtain the effective browsing frame trajectory of each customer;

[0115] When the set conditions are met, the retention rate calculation unit obtains the interval between the order times displayed in the order interface of two adjacent target products in the effective browsing frame trajectory of each customer, and builds a calculation model based on the obtained results to calculate the real-time customer retention rate of the target product;

[0116] The marketing content reach rate calculation unit calculates the real-time marketing content reach rate of the marketing video based on the overlap of each customer's effective browsing frame trajectory;

[0117] The marketing content value density calculation unit calculates the marketing content value density of each frame of marketing content in the marketing video based on the overlap between the effective browsing frame trajectory of each customer and the target browsing frame trajectory;

[0118] The module for selecting marketing content to be optimized is used to select marketing content to be optimized in the current marketing strategy of the target product;

[0119] The module for selecting marketing content to be optimized includes an optimization index calculation unit and an optimization judgment unit;

[0120] The optimized index calculation unit calculates the real-time optimized index of the current marketing strategy of the target product based on the constructed mathematical model;

[0121] The optimization judgment unit judges whether it is necessary to optimize the marketing content of each frame in the marketing video according to the set judgment conditions;

[0122] The associated frame search module is used to determine the associated frames of each frame of marketing content to be optimized;

[0123] The associated frame search module includes a matrix processing unit, an associated coefficient calculation unit and a key frame search unit;

[0124] The matrix processing unit matrices the real-time marketing content value index of each frame of marketing content in the marketing video to obtain a target matrix;

[0125] The correlation coefficient calculation unit calculates the correlation coefficient between two random frames of marketing content in the marketing video according to the target matrix;

[0126] The associated frame searching unit searches for associated frames of each frame of the marketing content to be optimized based on the calculation result transmitted by the associated coefficient calculating unit.

[0127] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art will be able to modify the technical solutions described in the aforementioned embodiments or substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A product marketing strategy optimization method based on multi-source data analysis, characterized by: The method comprises: S10: Sorting out the sales process data of the target product during the marketing period to obtain the effective browsing frame trajectory of each customer. Analyzing the effective browsing frame trajectory of each customer to obtain the real-time marketing data of the target product under the current marketing strategy. The marketing data includes customer retention rate, marketing content reach rate, and marketing content value density. S20: Analyze the optimization index of the current marketing strategy of the target product based on the marketing data, and select the marketing content to be optimized in the current marketing strategy of the target product according to the analysis results; S30: Matrixing the real-time marketing content value index of each frame of marketing content in the marketing video, analyzing the correlation between the marketing content of each frame, and determining the associated frames of the marketing content to be optimized for each frame based on the analysis results; S40: Determine an optimization plan for the current marketing strategy of the target product.

2. The method for optimizing product marketing strategies based on multi-source data analysis according to claim 1, characterized in that: The S10 includes: S101: Generate a marketing video based on the current marketing strategy of the target product and distribute the marketing video on various marketing channels. Obtain the marketing time period [t, T] of the target product under the current marketing strategy. Collect orders placed for the target product within the marketing time period [t, T], records of marketing video views within the marketing time period [t, T], and the life cycle of the target product. S102: Orders placed for the target product within the marketing time period [t, T] and each customer's browsing history of the marketing video within the marketing time period [t, T] are sorted to obtain each customer's valid browsing frames, and the valid browsing frames of each customer are sorted in chronological order to generate a valid browsing frame trajectory for each customer; S103: Numbering the customers with valid browsing frame traces, the numbering result is: i=1, 2, ..., m, where m represents the total number of customers with valid browsing frame traces; In the effective browsing frame trajectory of customer i, the number of order interfaces for the target product h i To obtain, when h i When ≥2, in the effective browsing frame trajectory of customer i, the interval time between the order times displayed in the order interface of two adjacent target products is obtained to obtain the interval time set. The elements stored in the interval time set are processed according to the average value calculation method to obtain U i ; If U i ≤u+R, then the retention index of customer i is f i =1, if U i >u+R, then the retention index of customer i is f i =1 / (U i -u+R), where R represents the life cycle fluctuation value of the target product, and u represents the life cycle of the target product; according to H iT =1-exp[(1-h i )*f i ] Predict the retention rate of the target product for customer i at time T, from i = 1 to i = m for all H iT Perform summation to obtain H′ iT , for H′ iT The customer retention rate U of the target product at time T is calculated by calculating the ratio between the target product and m. T , where exp() represents an exponential function with base e and e=2.73; Based on the overlap of each customer's effective browsing frame trajectory, and the overlap of each customer's effective browsing frame trajectory with the target browsing frame trajectory, the real-time marketing content reach rate of the marketing video and the marketing content value density of each frame of the marketing content in the marketing video are analyzed.

3. The method for optimizing product marketing strategies based on multi-source data analysis according to claim 2, characterized in that: The specific method of analyzing the real-time marketing content reach rate of the marketing video and the marketing content value density of each frame of the marketing content in the marketing video in S103 is: Number each frame of marketing content according to the order in which it is displayed in the marketing video. The numbering result is: j = 1, 2, ..., n, where n represents the total number. Compare the effective browsing frame trajectories of each customer, and based on the comparison results, calculate the number of views S of the marketing content of the jth frame in the marketing video at time T. jT Obtain; compare the effective browsing frame trajectory of each customer with the target browsing frame trajectory of the marketing video, and based on the comparison results, calculate the total number of views K of the target browsing frame in the marketing video at time T T To obtain; K T Calculate the ratio between the target product and m to get the marketing content reach rate C of the target product at time T. T ; To S jT The ratio between the value of W and m is calculated to obtain the browsing density W of the marketing content of the jth frame in the marketing video of the target product at time T. jT In the effective browsing frame trajectory of customer i, the minimum number of marketing content frames G between the order interface of each target product and the marketing content of the jth frame in the marketing video is ij To determine, if G ij = 0, then the value index V of the marketing content of the jth frame in the marketing video at time T ijT =1, if G ij ≠0, then the value index V of the marketing content of the jth frame in the marketing video at time T ijT =1 / G ij , from i = 1 to i = m for all V ijT Perform summation to obtain V jT , for W jT With V jT / m to calculate the product of the marketing content value density M of the marketing content of the jth frame in the marketing video of the target product at time T. jT .

4. The method for optimizing product marketing strategies based on multi-source data analysis according to claim 3, characterized in that: The S20 includes: S201: According to W T =a*U T +b*C T +c*∑{M jT *r j *[1-exp(-j)]} calculates the optimization index of the current marketing strategy of the target product at time T, where ∑ represents the summation symbol, the subscript of ∑ is j=1, the superscript is n, and r j Represents the weight coefficient corresponding to the marketing content of the jth frame in the marketing video; When 0.6<W T ≤1, it is necessary to optimize the current marketing strategy of the target product. T When ≤0.6, there is no need to optimize the current marketing strategy of the target product; S202: When 0.6<W T ≤1, M jT *r j *[1-exp(-j)] is compared with the set threshold. If M jT *r j *[1-exp(-j)]≤set threshold, then it is considered that the marketing content of the jth frame in the marketing video needs to be optimized. If M jT *r j *[1-exp(-j)]>set threshold, then it is considered that the marketing content of the jth frame in the marketing video does not need to be optimized.

5. The method for optimizing product marketing strategies based on multi-source data analysis according to claim 4, characterized in that: The S30 includes: S301: Matrix the real-time marketing content value index of each frame of marketing content in the marketing video to obtain the matrix U T , Wherein, p=1,2,…,q represents the number corresponding to the number of times sales process data is collected, p represents the total number of times sales process data is collected, and d represents the collection interval of sales process data; According to the matrix U T , the correlation coefficient γ between the marketing content of the xth frame in the marketing video and the marketing content of the jth frame in the marketing video xj Perform calculations, x=1,2,…,n and x≠j; S302: Randomly select a frame of marketing content to be optimized, and record the frame number corresponding to the selected marketing content as g, where g = 1, 2, ..., n. If γ gj > 0.8 and the marketing content of the jth frame in the marketing video is the marketing content to be optimized, then the marketing content of the jth frame in the marketing video is used as the associated frame of the marketing content of the gth frame in the marketing video. If γ gj ≤0.8 and the marketing content of the jth frame in the marketing video is the marketing content to be optimized, the marketing content of the jth frame in the marketing video will not be used as the associated frame of the marketing content of the gth frame in the marketing video.

6. The method for optimizing product marketing strategies based on multi-source data analysis according to claim 5, characterized in that: The S40 includes: When the marketing content in the g-th frame of a marketing video has an associated frame: Classify the marketing content of the g-th frame in the marketing video and the marketing content of each associated frame of the g-th frame in the marketing video into one category of marketing content; When the marketing content in the g-th frame of a marketing video does not have an associated frame: The marketing content of the gth frame in the marketing video is regarded as a type of marketing content; Randomly select a category of marketing content and a frame to be browsed, and calculate the text similarity between the selected category of marketing content and the marketing content corresponding to the selected frame to be browsed; If the calculated text similarity is higher than the target threshold and the selected frame displays marketing content that demonstrates the brand image or the target product's functions, the marketing content corresponding to the selected frame is optimized based on the dimensional shortcomings of the selected marketing content. The specific optimization method is as follows: The marketing data of each frame of the selected marketing content is obtained during playback and represented in a five-dimensional graph. Based on the dispersion coefficient and mean of each type of marketing data on the corresponding coordinate axis, the dimensional weaknesses of the selected marketing content are determined. Dimensional weaknesses refer to marketing data types with a dispersion coefficient greater than 0.7 and a mean less than 0.

6. If the weak point of the dimension is customer retention rate, provide customers with after-sales service based on their evaluation of the target product, and add the provided after-sales solution to the selected frame to be browsed; If the shortcoming is the marketing content reach, increase the frequency of marketing videos; If the weak point in the dimension is the value density of marketing content, the playback order of the selected browsing frames is optimized so that the optimized selected browsing frames are played before the marketing content with a value index of 1; If the calculated text similarity is higher than the target threshold and the selected frame to be browsed does not display marketing content that shows the brand image or the function demonstration of the target product, the selected frame to be browsed will be deleted from the marketing video; If the calculated text similarity is lower than the target threshold, the selected frame to be browsed is retained in the marketing video.

7. A product marketing strategy optimization system based on multi-source data analysis, applied to the product marketing strategy optimization method based on multi-source data analysis according to any one of claims 1 to 6, characterized in that: The system includes a marketing data analysis module, a marketing content selection module to be optimized, a related frame search module and an optimization solution generation module; The marketing data analysis module is used to analyze the effective browsing frame trajectory of the customer and obtain the real-time marketing data of the target product under the current marketing strategy; The module for selecting marketing content to be optimized is used to select marketing content to be optimized in the current marketing strategy of the target product; The associated frame search module is used to determine the associated frames of each frame of marketing content to be optimized; The solution generation module is used to determine the optimization solution for the current marketing strategy of the target product.

8. The product marketing strategy optimization system based on multi-source data analysis according to claim 7 is characterized by: The marketing data analysis module includes a collection unit, an effective browsing trajectory generation unit, a retention rate calculation unit, a marketing content reach rate calculation unit, and a marketing content value density calculation unit; The collection unit collects orders placed for the target product within the marketing period, records of marketing videos viewed within the marketing period, and the life cycle of the target product; The effective browsing trajectory generation unit sorts out the orders placed for the target product during the marketing period and the browsing records of the marketing videos by each customer during the marketing period to obtain the effective browsing frame trajectory of each customer; The retention rate calculation unit obtains the interval between the order times displayed in two adjacent target product ordering interfaces in the effective browsing frame trajectory of each customer when the set conditions are met, and constructs a calculation model based on the obtained results to calculate the real-time customer retention rate of the target product; The marketing content reach rate calculation unit calculates the real-time marketing content reach rate of the marketing video based on the overlap of each customer's effective browsing frame trajectory; The marketing content value density calculation unit calculates the marketing content value density of each frame of marketing content in the marketing video based on the overlap between the effective browsing frame trajectory of each customer and the target browsing frame trajectory.

9. The product marketing strategy optimization system based on multi-source data analysis according to claim 8, characterized in that: The marketing content selection module to be optimized includes an optimization index calculation unit and an optimization judgment unit; The optimized index calculation unit calculates the real-time optimized index of the current marketing strategy of the target product according to the constructed mathematical model; The optimization judgment unit judges whether it is necessary to optimize the marketing content of each frame in the marketing video according to the set judgment conditions.

10. The product marketing strategy optimization system based on multi-source data analysis according to claim 9, characterized in that: The associated frame search module includes a matrix processing unit, an associated coefficient calculation unit and a key frame search unit; The matrix processing unit matrices the real-time marketing content value index of each frame of marketing content in the marketing video to obtain a target matrix; The correlation coefficient calculation unit calculates the correlation coefficient between two random frames of marketing content in the marketing video according to the target matrix; The associated frame searching unit searches for associated frames of each frame of the marketing content to be optimized based on the calculation result transmitted by the associated coefficient calculating unit.

Images