Precision marketing management method and device based on offline dealer

Abstract

The invention discloses a precision marketing management method and device based on offline dealers, and relates to the technical field of channel management and data analysis. According to the offline dealer-based precision marketing management method disclosed by the invention, dealer inventory data, regional economic data, competitive product data and terminal sales data in a plurality of preset marketing regions are collected and analyzed, a regional market feature matrix is constructed, and a pre-trained double-cycle prediction model is combined, so that the precision marketing management efficiency is improved. And outputting a corresponding long-term sales prediction value and a short-term replenishment decision parameter set according to the short-term sales prediction value, thereby generating a differentiated marketing strategy set, and performing AB test to determine an optimal marketing strategy in each preset marketing area. Thus, the method can realize accurate mastering of market characteristics of each region, and through accurate analysis of a large amount of data and reasonable prediction of marketing results, the pertinence and effectiveness of marketing strategies are improved, so that the marketing effect of white spirit products is improved.

Description

Technical Field

[0001] This application relates to the technical fields of channel management and data analysis, and particularly to a precise marketing management method and device based on offline dealers. Background Art

[0002] Marketing management is a key link to improve channel efficiency and market competitiveness. In the prior art, marketing management plans for liquor products are generally formulated based on manual experience in a unified manner. Since it is impossible to accurately analyze a large amount of data and reasonably predict marketing results, the formulation of marketing strategies often has great subjectivity and blindness, resulting in unsatisfactory marketing effects.

[0003] The above content is only used to assist in understanding the technical solution of this application, and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main purpose of this application is to provide a precise marketing management method and device based on offline dealers, aiming to improve the marketing effect of liquor products.

[0005] To achieve the above purpose, this application proposes a precise marketing management method based on offline dealers, which is applied to liquor products. The method includes:

[0006] Collect dealer inventory data, regional economic data, competitor data, and terminal sales data in multiple preset marketing regions;

[0007] Determine the corresponding economic level index according to the economic data of each preset marketing region, determine the corresponding competitor penetration rate according to the competitor data of each preset marketing region, and determine the corresponding drinking culture intensity according to the terminal sales data of each preset marketing region;

[0008] Construct a regional market feature matrix based on the economic level index, competitor penetration rate, and drinking culture intensity of each preset marketing region;

[0009] Input the dealer inventory data of multiple preset marketing regions and the regional market feature matrix into a pre-trained double-cycle prediction model, and output the corresponding long-term sales prediction value and short-term replenishment decision parameter set;

[0010] Generate a differentiated marketing strategy set according to the long-term sales prediction value and short-term replenishment decision parameter set corresponding to multiple preset marketing regions, and conduct an AB test based on the differentiated marketing strategy set;

[0011] Determine the optimal marketing strategy in each preset marketing region according to the results of the AB test.

[0012] In one embodiment, the regional economic data includes regional GDP, per capita disposable income, total retail sales of consumer goods, and the proportion of liquor consumption expenditure. The step of determining the corresponding economic level index according to the economic data of each preset marketing region includes:

[0013] Normalize the regional GDP, per capita disposable income, total retail sales of consumer goods, and the proportion of liquor consumption expenditure of each preset marketing region;

[0014] Perform weighted summation on the normalized regional GDP, per capita disposable income, total retail sales of consumer goods, and the proportion of liquor consumption expenditure according to the preset weight distribution to obtain the economic level index of each preset marketing region.

[0015] In one embodiment, the competitor data includes the total number of sales channel types and the sales volume in each sales channel. The step of determining the corresponding competitor penetration rate according to the competitor data of each preset marketing region includes determining the competitor penetration rate of each preset marketing region according to the following formula:

[0016] ;

[0017] Wherein, represents the competitor penetration rate, represents the total number of sales channel types, represents the sales volume of the competitor in the preset marketing region sales channel sales volume, represents the sales volume of the liquor product in the preset marketing region sales channel sales volume.

[0018] In one embodiment, it is characterized in that the terminal sales data includes the high-frequency purchase rate, the proportion of gift box sales volume, and the peak festival sales volume. The step of determining the corresponding drinking culture intensity according to the terminal sales data of each preset marketing region includes determining the drinking culture intensity of each preset marketing region according to the following formula:

[0019] ;

[0020] Where represents the drinking culture intensity; represents the high-frequency purchase rate of the preset marketing region ; represents the highest high-frequency purchase rate among all preset marketing regions, represents the weight corresponding to the high-frequency purchase rate; represents the proportion of gift box sales volume of the preset marketing region ; Represents the highest proportion of gift box sales volume in all preset marketing regions; Represents the weight corresponding to the proportion of gift box sales volume; Represents the preset marketing region The peak festival sales volume; Represents the highest peak festival sales volume in all preset marketing regions; Represents the weight corresponding to the peak festival sales volume; where .

[0021] In one embodiment, the method further includes:

[0022] Collect historical dealer inventory data, historical regional economic data, historical competitor data, and historical terminal sales data, as well as the corresponding historical sales volume data;

[0023] Split the historical sales volume data to obtain long-term sales volume historical data and short-term sales volume historical data;

[0024] Construct a long-term sales volume prediction model based on the long-term sales volume historical data, and construct a short-term replenishment decision model based on the short-term sales volume historical data;

[0025] Fuse the long-term sales volume prediction model and the short-term replenishment decision model to obtain a dual-cycle prediction model;

[0026] Train the dual-cycle prediction model until the preset prediction accuracy requirement is met to obtain a pre-trained dual-cycle prediction model.

[0027] In one embodiment, the step of generating a differentiated marketing strategy set according to the long-term sales volume prediction values and short-term replenishment decision parameter sets corresponding to multiple preset marketing regions and performing an AB test based on the differentiated marketing strategy set includes:

[0028] Obtain a preset marketing strategy algorithm library, where the marketing strategy algorithm library includes a variety of marketing strategy generation algorithms;

[0029] Match the features of the long-term sales volume prediction values and short-term replenishment decision parameter sets with the various marketing strategy generation algorithms in the marketing strategy algorithm library to obtain the corresponding marketing strategy generation algorithms;

[0030] Generate a differentiated marketing strategy set based on the selected marketing strategy generation algorithms and perform an AB test based on the differentiated marketing strategy set.

[0031] In one embodiment, before the step of determining the optimal marketing strategy in each preset marketing region according to the results of the AB test, the method further includes:

[0032] Identify clusters of similar regions among multiple preset marketing regions;

[0033] Select and execute differentiated marketing strategies from the differentiated marketing strategy set according to different preset marketing regions within the same type of regional cluster to obtain the sub-optimal marketing strategies;

[0034] Obtain a regional strategy set for a preset marketing region within the same type of regional cluster based on the sub-optimal marketing strategies;

[0035] Divide the distributors within the preset marketing region into group A and group B, and apply the marketing strategies in the regional strategy set to the distributors in group A and group B respectively to obtain the results of the A / B test.

[0036] In one embodiment, the step of determining the optimal marketing strategy for each preset marketing region according to the results of the A / B test includes:

[0037] Collect the sales data of the distributors in group A and group B after applying different marketing strategies;

[0038] Determine the sales growth rate, market share change, and customer satisfaction of the distributors in group A and group B after applying different marketing strategies for the collected sales data;

[0039] Determine the optimal marketing strategy for each preset marketing region according to the sales growth rate, market share change, and customer satisfaction.

[0040] In one embodiment, the method further includes:

[0041] Obtain the business violation data and abnormal operation data of the preset marketing region;

[0042] Perform pattern recognition on the business violation data and abnormal operation data to construct a distributor risk assessment model;

[0043] Generate a channel warning signal corresponding to the preset marketing region according to the risk level coefficient output by the distributor risk assessment model;

[0044] When the channel warning signal reaches the preset risk threshold, trigger the distributor qualification review mechanism and the emergency replenishment strategy.

[0045] In addition, to achieve the above object, the present application also proposes a precise marketing management device based on offline distributors, and the device includes: a memory, a processor, and a precise marketing management program based on offline distributors stored on the memory and executable on the processor, and the precise marketing management program based on offline distributors is configured to implement the steps of the precise marketing management method based on offline distributors.

[0046] The precise marketing management method based on offline distributors proposed in this application is applied to liquor products. The method includes collecting the inventory data, regional economic data, competitor data, and terminal sales data of distributors in multiple preset marketing regions; then determining the corresponding economic level index according to the economic data of each preset marketing region, determining the corresponding competitor penetration rate according to the competitor data of each preset marketing region, and determining the corresponding drinking culture intensity according to the terminal sales data of each preset marketing region; then constructing a regional market characteristic matrix based on the economic level index, competitor penetration rate, and drinking culture intensity of each preset marketing region; and inputting the inventory data of distributors in multiple preset marketing regions and the regional market characteristic matrix into a pre-trained double-cycle prediction model to output the corresponding long-term sales prediction value and short-term replenishment decision parameter set; then generating a differentiated marketing strategy set according to the long-term sales prediction value and short-term replenishment decision parameter set corresponding to multiple preset marketing regions to conduct an AB test based on the differentiated marketing strategy set; finally, determining the optimal marketing strategy in each preset marketing region according to the results of the AB test. That is, by collecting and analyzing the inventory data, regional economic data, competitor data, and terminal sales data of distributors in multiple preset marketing regions, constructing a regional market characteristic matrix, and combining a pre-trained double-cycle prediction model, the corresponding long-term sales prediction value and short-term replenishment decision parameter set are output, thereby generating a differentiated marketing strategy set and conducting an AB test to determine the optimal marketing strategy in each preset marketing region. In this way, the above method can accurately grasp the characteristics of each regional market, improve the pertinence and effectiveness of marketing strategies by accurately analyzing a large amount of data and reasonably predicting marketing results, and thus improve the marketing effect of liquor products. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings herein are incorporated into and constitute a part of this specification, showing embodiments consistent with the present application and, together with the specification, are used to explain the principles of the present application.

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0049] Figure 1 It is a flowchart provided for an embodiment of the precise marketing management method based on offline distributors of the present application;

[0050] Figure 2 It is a flowchart provided for another embodiment of the precise marketing management method based on offline distributors of the present application;

[0051] Figure 3A flowchart provided for another embodiment of the precise marketing management method based on offline distributors in the present application;

[0052] Figure 4 A flowchart provided for yet another embodiment of the precise marketing management method based on offline distributors in the present application;

[0053] Figure 5 A flowchart provided for still another embodiment of the precise marketing management method based on offline distributors in the present application;

[0054] Figure 6 A structural diagram provided for an embodiment of the precise marketing management device based on offline distributors in the present application.

[0055] Explanation of the reference numerals in the accompanying drawings:

[0056] 10. Memory; 20. Processor.

[0057] The realization of the object, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. Detailed implementation manners

[0058] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0059] For a better understanding of the technical solutions of the present application, the following will be described in detail with reference to the accompanying drawings of the specification and the specific implementation manners.

[0060] The main solution of the embodiment of the present application is: collecting the inventory data, regional economic data, competitor data and terminal sales data of distributors in multiple preset marketing regions; then determining the corresponding economic level index according to the economic data of each preset marketing region, determining the corresponding competitor penetration rate according to the competitor data of each preset marketing region, and determining the corresponding drinking culture intensity according to the terminal sales data of each preset marketing region; then constructing a regional market feature matrix based on the economic level index, competitor penetration rate and drinking culture intensity of each preset marketing region; and inputting the inventory data of distributors in multiple preset marketing regions and the regional market feature matrix into a pre-trained double-cycle prediction model to output corresponding long-term sales prediction values and short-term replenishment decision parameter sets; then generating a differentiated marketing strategy set according to the long-term sales prediction values and short-term replenishment decision parameter sets corresponding to multiple preset marketing regions for AB testing based on the differentiated marketing strategy set; and finally determining the optimal marketing strategy in each preset marketing region according to the results of the AB testing.

[0061] In this embodiment, for the convenience of description, the following will be described with the precise marketing management device based on offline distributors as the execution subject.

[0062] Marketing management is a key link in improving channel efficiency and market competitiveness. In the existing technology, the marketing management plan for liquor products is generally formulated based on manual experience. Since it is unable to achieve precise analysis of a large amount of data and reasonable prediction of marketing results, the formulation of marketing strategies often has great subjectivity and blindness, resulting in unsatisfactory marketing effects.

[0063] This application provides a solution. By collecting and analyzing the inventory data of dealers, regional economic data, competitor data, and terminal sales data in multiple preset marketing regions, a regional market feature matrix is constructed, and combined with a pre-trained double-cycle prediction model, the corresponding long-term sales volume prediction value and short-term replenishment decision parameter set are output, so as to generate a differentiated marketing strategy set and conduct an AB test to determine the optimal marketing strategy in each preset marketing region. In this way, the above method can accurately grasp the characteristics of each regional market, and improve the pertinence and effectiveness of marketing strategies through precise analysis of a large amount of data and reasonable prediction of marketing results, thereby enhancing the marketing effect of liquor products.

[0064] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of implementing the above functions, a precise marketing management device based on offline dealers, etc. Hereinafter, a precise marketing management device based on offline dealers will be taken as an example to illustrate this embodiment and the following embodiments.

[0065] Based on this, the embodiment of this application provides a precise marketing management method based on offline dealers, referring to Figure 1 , Figure 1 is a schematic flowchart of an embodiment of the precise marketing management method based on offline dealers of this application.

[0066] In this embodiment, the precise marketing management method based on offline dealers includes steps S100 to S600, where:

[0067] Step S100, collect the inventory data of dealers, regional economic data, competitor data, and terminal sales data in multiple preset marketing regions.

[0068] In this embodiment, multiple preset marketing regions can be pre-divided according to factors such as geographical location, economic level, and cultural background, and each preset marketing region is regarded as an independent market unit. It can be understood that in the management of offline distributors, due to the large differences in the economic level, competing product situations, and consumer drinking cultures in different regions, the traditional unified management strategy is difficult to meet the market demands of different regions. Therefore, in this solution, each preset marketing region is regarded as an independent market unit to separately collect data for different preset marketing regions according to the actual situations of each region, so as to facilitate the formulation of precise marketing strategies in the follow-up.

[0069] In this embodiment, the distributor inventory data can include the inventory quantity of distributors within the preset marketing region and the historical replenishment quantity of the distributors. Specifically, the inventory quantity of distributors can be obtained in real time through the distributor management system or the inventory management system, while the historical replenishment quantity of the distributors can be obtained through the analysis of historical data. The regional economic data can include the regional GDP, per capita disposable income, total retail sales of social consumer goods, and the proportion of liquor consumption expenditure. Among them, the regional GDP and per capita disposable income can be obtained by accessing the public data of relevant statistical bureaus, and the total retail sales of social consumer goods and the proportion of liquor consumption expenditure can be obtained through market research or third-party data service agencies. The competing product data can include the total number of sales channel types and the sales amount in each sales channel, and the competing product data can be obtained through the market research report of competing products or the competing product analysis system. The terminal sales data can include the high-frequency purchase rate, the proportion of gift box sales volume, and the peak sales volume during festivals. The above data can be obtained through the analysis of sales records within the preset marketing region. By collecting the distributor inventory data, regional economic data, competing product data, and terminal sales data in multiple preset marketing regions, it is convenient for subsequent data analysis to comprehensively understand the market demands and consumer behaviors within the preset marketing region, thereby providing strong support for the formulation of subsequent marketing strategies.

[0070] Step S200: Determine the corresponding economic level index according to the economic data of each preset marketing region, determine the corresponding competing product penetration rate according to the competing product data of each preset marketing region, and determine the corresponding drinking culture intensity according to the terminal sales data of each preset marketing region.

[0071] In this embodiment, the economic level index can be obtained by comprehensively evaluating various indicators in the regional economic data. For example, indicators such as regional GDP, per capita disposable income, total retail sales of social consumer goods, and the proportion of liquor consumption expenditure can be standardized, and then the economic level index can be obtained through methods such as weighted average or principal component analysis. The penetration rate of competing products can be obtained by calculating the proportion of the sales volume of competing products in the total sales volume within the preset marketing area. The intensity of drinking culture can be obtained by analyzing indicators such as the high-frequency purchase rate, the proportion of gift box sales volume, and the peak sales volume during festivals. For example, a preset marketing area with a high high-frequency purchase rate, a large proportion of gift box sales volume, and an obvious peak sales volume during festivals can be considered to have a high intensity of drinking culture.

[0072] In a feasible implementation manner, the step of determining the corresponding economic level index according to the economic data of each preset marketing area includes:

[0073] Normalize the regional GDP, per capita disposable income, total retail sales of social consumer goods, and the proportion of liquor consumption expenditure of each preset marketing area.

[0074] In this embodiment, normalization processing refers to converting indicators of different magnitudes or dimensions into dimensionless standardized values for subsequent data analysis and comparison. Specifically, the min-max normalization method can be used to convert the values of each indicator into the range of 0 to 1. The conversion formula is: Xnorm=(X - Xmin) / (Xmax - Xmin), where X is the original data, Xmin is the minimum value of the indicator, Xmax is the maximum value of the indicator, and Xnorm is the converted standardized value. In this way, the standardized values of the economic level index of each preset marketing area can be obtained, providing a basis for subsequent analysis and comparison.

[0075] Perform weighted summation on the normalized regional GDP, per capita disposable income, total retail sales of social consumer goods, and the proportion of liquor consumption expenditure according to the preset weight distribution to obtain the economic level index of each preset marketing area.

[0076] In this embodiment, the preset weight distribution can be set according to the actual situation. For example, the weight distribution can be performed according to the influence degree of each indicator on the formulation of marketing strategies. Specifically, methods such as the expert scoring method and the analytic hierarchy process can be used for weight distribution to ensure that the weight distribution of each indicator is reasonable and accurate. After determining the preset weight distribution, the normalized regional GDP, per capita disposable income, total retail sales of social consumer goods, and the proportion of liquor consumption expenditure can be weighted and summed according to the preset weight distribution to obtain the economic level index of each preset marketing area, so as to reflect the economic situation and consumption potential of each marketing area and provide an important reference for subsequent formulation of marketing strategies.

[0077] In a feasible implementation, the step of determining the corresponding competitor penetration rate according to the competitor data of each preset marketing area includes determining the competitor penetration rate of each preset marketing area according to the following formula:

[0078] ;

[0079] where represents the competitor penetration rate, represents the total number of sales channel types, represents the sales volume of the competitor in the preset marketing area sales channel ; represents the sales volume of the baijiu product in the preset marketing area sales channel ;

[0080] In this embodiment, the competitor penetration rate is the proportion of the sales volume of the competitor in a certain preset marketing area to the total sales volume in that area. Specifically, by counting the sales volume of the competitor in each sales channel and calculating the proportion of the competitor's sales volume to the total sales volume (the sum of the competitor's sales volume and the sales volume of the baijiu product in this application), the competitor penetration rate is obtained, which reflects the market share and competitiveness of the competitor in the preset marketing area. By analyzing the competitor penetration rate, the market performance and sales strategy of the competitor can be understood, so as to provide reference for formulating its own marketing strategy.

[0081] In a feasible implementation, the step of determining the corresponding drinking culture intensity according to the terminal sales data of each preset marketing area includes determining the drinking culture intensity of each preset marketing area according to the following formula:

[0082] ;

[0083] where represents the drinking culture intensity; represents the high-frequency purchase rate of the preset marketing area ; represents the highest high-frequency purchase rate among all preset marketing areas, represents the weight corresponding to the high-frequency purchase rate; represents the proportion of the sales volume of gift boxes in the preset marketing area ; represents the highest proportion of the sales volume of gift boxes among all preset marketing areas; represents the weight corresponding to the proportion of the sales volume of gift boxes; represents the peak festival sales volume of the preset marketing area ; represents the highest peak festival sales volume among all preset marketing areas; Indicates the weight corresponding to the peak sales volume during festivals; among which .

[0084] In this embodiment, the drinking culture intensity is a quantitative assessment of the drinking habits and consumption behaviors of consumers in a preset marketing area. Specifically, the high-frequency purchase rate reflects the purchase frequency of consumers for baijiu products. The higher the high-frequency purchase rate, the stronger the demand of consumers in this area for baijiu products. The proportion of gift box sales volume reflects the degree of demand of consumers for baijiu products as gifts. The larger the proportion of gift box sales volume, the more consumers in this area pay attention to the gift attribute of baijiu products. The peak sales volume during festivals reflects the purchase volume of consumers for baijiu products during specific festivals. The higher the peak sales volume during festivals, the more concentrated the demand of consumers in this area for baijiu products during festivals. The weights corresponding to the high-frequency purchase rate, the proportion of gift box sales volume, and the peak sales volume during festivals can be set according to the actual situation to ensure that the weight distribution of each indicator is reasonable and accurate. After determining the weight distribution of each indicator, the high-frequency purchase rate, the proportion of gift box sales volume, and the peak sales volume during festivals can be weighted and summed according to the corresponding weights to obtain the drinking culture intensity of each preset marketing area, so as to reflect the drinking habits and consumption behavior characteristics of consumers in each marketing area and provide a reference for formulating subsequent marketing strategies.

[0085] In a feasible implementation manner, the method can also obtain the terminal bottle-opening data, terminal inventory data, and regional population data of the distributor and integrate them into the construction process of the regional market feature matrix. Specifically, the terminal bottle-opening data is used to collect the actual bottle-opening frequency of consumers in real time through Internet of Things devices, which is used to correct the real consumption activity of the high-frequency purchase rate indicator; the regional population data includes the median age of the permanent population, the household size, and the proportion of migrant workers, which is used to construct a population structure impact factor to optimize the regional feature adaptability of the long-term sales prediction model. The terminal inventory data is obtained in real time through the API interface of the distributor management system, specifically including the inventory turnover rate and the proportion of near-expiry products, which is used to dynamically correct the generation logic of the short-term replenishment decision parameter set. In the data integration stage, the terminal bottle-opening data and the high-frequency purchase rate are cross-validated, and a consumption behavior confidence indicator is constructed by calculating the time series correlation coefficient of the bottle-opening frequency and the purchase frequency; at the same time, a regression analysis is performed on the household size in the regional population data and the proportion of gift box sales volume to establish a packaging specification adaptation coefficient, which is used to guide the SKU combination strategy for different preset marketing areas.

[0086] When constructing the regional market characteristic matrix, the three core indicators of economic level index, competitive product penetration rate, and drinking culture intensity are coupled with consumer behavior confidence index and population structure influencing factors in multiple dimensions. Specifically, the five-dimensional feature space is reduced in dimension through tensor decomposition technology to form a three-dimensional feature vector group with strong interpretability, in which the first principal component reflects the market consumption potential and is composed of a linear combination of the economic level index and the population structure influencing factors; the second principal component represents the market competition pattern and is composed of nonlinear transformation values ​​of competitive product penetration rate and drinking culture intensity; the third principal component indicates the characteristics of consumer behavior and integrates the dynamic interaction effect of consumer behavior confidence index and packaging specification adaptation coefficient. The characteristic matrix is ​​updated in time series through a sliding window mechanism to ensure that the marketing strategy can respond to the dynamic evolution of regional market characteristics.

[0087] For the preset marketing areas where the inventory turnover rate is lower than the industry average and the difference with the industry average is lower than the preset turnover rate difference, the device automatically triggers the channel health warning module, and adjusts the smoothing coefficient of the short-term replenishment decision parameter in the dual-cycle forecasting model to reduce the replenishment recommendation value in the high inventory risk area. At the same time, the terminal promotion strategy generation engine is linked to design a gradient discount plan for expiring products. When the proportion of migrant workers in the regional population data exceeds the threshold of 35%, the population mobility compensation algorithm is activated, and the seasonal migration factor is introduced in the long-term sales forecast. The superimposed effect of the migrant workers' return cycle on holiday sales is predicted through the ARIMA model.

[0088] Step S300, constructing a regional market characteristic matrix based on the economic level index, competitive product penetration rate and drinking culture intensity of each preset marketing area.

[0089] In this embodiment, the regional market characteristic matrix is ​​a data matrix containing multiple dimensions, in which each row represents a preset marketing area, and each column represents a different market characteristic index, such as economic level index, competitive product penetration rate, and drinking culture intensity. By constructing the regional market characteristic matrix, the performance and differences of each preset marketing area in different market characteristics can be intuitively displayed, thereby providing data support for the subsequent marketing strategy formulation.

[0090] Step S400, input the dealer inventory data of multiple preset marketing areas and the regional market feature matrix into the pre-trained dual-period forecasting model, and output the corresponding long-term sales forecast value and short-term replenishment decision parameter set.

[0091] In this embodiment, the pre-trained dual-cycle prediction model is a prediction model constructed based on big data and machine learning algorithms. It can accurately predict future sales trends and replenishment demands by analyzing and learning historical data. Specifically, the dual-cycle prediction model can include two parts: a long-term prediction model and a short-term prediction model. The long-term prediction model is used to predict sales trends over a relatively long period, while the short-term prediction model is used to predict replenishment demands over a relatively short period. By inputting the dealer inventory data and regional market feature matrices of multiple preset marketing regions into the pre-trained dual-cycle prediction model, long-term sales prediction values and short-term replenishment decision parameter sets for each preset marketing region in the future period can be obtained, thus providing data support for subsequent marketing strategy formulation.

[0092] In a feasible implementation manner, referring to Figure 2 , the method further includes steps S710 to S750, where:

[0093] Step S710, collect historical dealer inventory data, historical regional economic data, historical competitor data, historical terminal sales data, and corresponding historical sales data.

[0094] In this embodiment, in order to construct the above-mentioned pre-trained dual-cycle prediction model, a large amount of historical data needs to be collected to construct a training sample set. Specifically, historical dealer inventory data can reflect the inventory changes of dealers, historical regional economic data can reflect the economic conditions and consumption potential of each region, historical competitor data can reflect the market performance and sales strategies of competitors, and historical terminal sales data can reflect the purchasing behaviors and consumption habits of consumers. The corresponding historical sales data is the target value that the model needs to predict, that is, the actual sales volume of each preset marketing region in different time periods.

[0095] Step S720, split the historical sales data to obtain long-term sales historical data and short-term sales historical data.

[0096] In this embodiment, since the change rules and influencing factors of long-term sales and short-term sales are different, it is necessary to split the historical sales data in order to construct a long-term prediction model and a short-term prediction model respectively. Specifically, the historical sales data can be split into long-term sales historical data and short-term sales historical data according to different time scales. The long-term sales historical data is used to construct a long-term prediction model to predict sales trends over a relatively long period; while the short-term sales historical data is used to construct a short-term prediction model to predict replenishment demands over a relatively short period.

[0097] Step S730, construct a long-term sales prediction model based on the long-term sales historical data and construct a short-term replenishment decision model based on the short-term sales historical data.

[0098] In this embodiment, the long-term sales volume prediction model and the short-term replenishment decision model are two important components of the double-cycle prediction model. By analyzing and learning the historical data of long-term sales volume, the long-term sales volume prediction model can capture the long-term trends and periodic patterns of sales volume changes, so as to predict the sales volume in a relatively long period in the future. It is of great significance for enterprises to formulate long-term marketing strategies and production plans, and can help enterprises adjust production capacity and market layout in advance to cope with market changes. The short-term replenishment decision model, on the other hand, focuses more on predicting short-term sales volume fluctuations and replenishment demands. By analyzing the historical data of short-term sales volume, the short-term replenishment decision model can accurately predict the sales volume changes and replenishment demands in a relatively short period, so as to provide accurate replenishment suggestions for enterprises. This helps enterprises optimize inventory management, reduce inventory backlogs and out-of-stock phenomena, and improve inventory turnover and customer satisfaction.

[0099] In this embodiment, in the process of constructing the long-term sales volume prediction model and the short-term replenishment decision model, machine learning algorithms and big data technologies can be adopted. Through in-depth mining and analysis of historical data, the model can automatically learn the complex relationships between sales volume changes, replenishment demands and various market characteristic indicators, so as to achieve accurate prediction of future sales volume and replenishment demands.

[0100] In addition, this embodiment also takes into account the differences between different preset marketing regions. Since there are significant differences in economic levels, competitor situations, and consumer drinking cultures in different regions, when the model predicts the sales volume and replenishment demands in different regions, it will fully consider the above-mentioned difference factors to ensure the accuracy of the prediction results.

[0101] Step S740: Integrate the long-term sales volume prediction model and the short-term replenishment decision model to obtain a double-cycle prediction model.

[0102] In this embodiment, after obtaining the long-term sales volume prediction model and the short-term replenishment decision model, it is necessary to integrate these two models to construct a complete double-cycle prediction model. The integration process can be to perform weighted averaging on the prediction results of the two models or adopt other appropriate integration strategies to ensure that the double-cycle prediction model can capture both the long-term trends and short-term fluctuations of sales volume changes, so as to provide more accurate and comprehensive prediction results.

[0103] Step S750: Train the double-cycle prediction model until the preset prediction accuracy requirement is met to obtain a pre-trained double-cycle prediction model.

[0104] In this embodiment, in order to obtain a dual-period prediction model with excellent performance, the constructed training sample set can be used to train it. During the training process, the parameters of the model can be continuously adjusted and the structure of the model can be optimized to reduce the prediction error and improve the prediction accuracy. At the same time, in order to determine whether the model has been sufficiently trained, some preset convergence conditions can be set, such as the prediction error being less than a certain threshold or the number of training times reaching the preset upper limit, etc. When these convergence conditions are met, it can be considered that it meets the preset prediction accuracy requirements, and thus a pre-trained dual-period prediction model can be obtained, which can be used for subsequent sales volume prediction and replenishment decision support.

[0105] Step S500, generate a differentiated marketing strategy set according to the long-term sales volume prediction values and short-term replenishment decision parameter sets corresponding to multiple preset marketing regions, so as to conduct an AB test based on the differentiated marketing strategy set.

[0106] In this embodiment, the differentiated marketing strategy set is a set of targeted marketing strategies formulated according to the characteristics and needs of different preset marketing regions. Specifically, according to the long-term sales volume prediction values, the sales volume trends of each preset marketing region in a relatively long period in the future can be understood, so as to formulate corresponding market expansion or maintenance strategies. For example, for regions with higher sales volume prediction values, the market promotion efforts can be increased, the sales channels and product varieties can be increased to further improve the market share; while for regions with lower sales volume prediction values, a steady business strategy can be adopted, the product structure can be optimized, and the product quality and service level can be improved to stabilize the market share. At the same time, according to the short-term replenishment decision parameter set, the replenishment requirements of each preset marketing region in a relatively short period can be accurately predicted, so as to formulate reasonable inventory management and replenishment strategies to reduce inventory backlog and out-of-stock phenomena and improve the inventory turnover rate and customer satisfaction.

[0107] In this embodiment, during the process of generating the differentiated marketing strategy set, market characteristic indicators such as the economic level index, competitor penetration rate, and drinking culture intensity of each preset marketing region are also considered. Since these indicators can reflect the differences in economic conditions, market competition situations, and consumer behaviors in each region, when formulating marketing strategies, the above-mentioned differential factors need to be fully considered to ensure the pertinence and effectiveness of the marketing strategies. For example, in regions with a higher economic level, a high-end marketing strategy can be adopted to launch high-quality and high-value-added products and services; while in regions with a higher competitor penetration rate, brand building and market promotion need to be strengthened to improve brand awareness and reputation to resist the competitive pressure of competitors.

[0108] In addition, in this embodiment, the generated differentiated marketing strategy set is evaluated and optimized by using the A / B test. Specifically, the preset marketing area can be randomly divided into two groups, and different marketing strategies are used for promotion and sales respectively. Then, by analyzing and comparing the sales data of the two groups of areas, the effects, advantages and disadvantages of different marketing strategies are evaluated. According to the results of the A / B test, the differentiated marketing strategy set can be adjusted and optimized to improve the effectiveness and adaptability of the marketing strategy.

[0109] In a feasible implementation manner, referring to Figure 3 , step S500 includes steps S510 to S530, where:

[0110] Step S510, obtain a preset marketing strategy algorithm library, where the marketing strategy algorithm library includes a variety of marketing strategy generation algorithms.

[0111] In this embodiment, in order to formulate accurate marketing strategies, this embodiment constructs a marketing strategy algorithm library containing a variety of marketing strategy generation algorithms. These algorithms are based on different mathematical models and optimization theories, and can automatically generate a variety of possible marketing strategy solutions according to the input long-term sales forecast value and short-term replenishment decision parameter set. For example, marketing strategy generation algorithms based on machine learning algorithms such as linear regression, decision tree, random forest, and neural network can be used. These algorithms can learn the relationship between sales volume changes and marketing strategies from historical data, so as to provide a scientific basis for formulating future marketing strategies. In addition, the marketing strategy algorithm library can also include heuristic algorithms based on expert experience and business knowledge, so as to be able to generate more practical marketing strategy solutions in combination with the actual situation and business needs.

[0112] Step S520, match the characteristics of the long-term sales forecast value and the short-term replenishment decision parameter set with a variety of marketing strategy generation algorithms in the marketing strategy algorithm library to obtain the corresponding marketing strategy generation algorithm.

[0113] In this embodiment, after obtaining the preset marketing strategy algorithm library, it is necessary to select the most suitable algorithm according to the specific situation and requirements of each preset marketing area to generate differentiated marketing strategies. In this process, factors such as the economic level index, competitor penetration rate, drinking culture intensity, and historical sales data of each area can be comprehensively considered to ensure that the selected algorithm can accurately reflect the actual situation of each area and generate targeted marketing strategies.

[0114] Step S530, generate a differentiated marketing strategy set based on the selected marketing strategy generation algorithm, and perform an A / B test based on the differentiated marketing strategy set.

[0115] In this embodiment, when generating a differentiated marketing strategy set, the differences of each preset marketing region need to be fully considered. For example, for regions with a relatively high economic level and a relatively low penetration rate of competing products, a high-end marketing strategy can be adopted, focusing on enhancing the brand image and product quality; while for regions with a relatively low economic level and a relatively high penetration rate of competing products, strategies such as price discounts and promotional activities may be required to improve the market competitiveness of the product and the purchasing willingness of consumers. At the same time, reasonable inventory management and replenishment strategies also need to be formulated according to the short-term replenishment decision parameter set to ensure the smooth progress of product supply and sales.

[0116] Through the above steps, we can obtain a differentiated marketing strategy set for different preset marketing regions. These strategy sets not only consider the economic conditions and consumption potential of each region, but also fully consider the market performance of competing products and the behavioral characteristics of consumers, so they have high pertinence and effectiveness. When the marketing strategy set is applied to actual marketing activities, an AB test is carried out to evaluate and optimize the effect of the strategy.

[0117] In a feasible implementation manner, referring to Figure 4 , the method further includes step S810 to step S840, where:

[0118] Step S810, identifying clusters of similar regions among multiple preset marketing regions.

[0119] In this embodiment, in order to further optimize the formulation and implementation effect of the marketing strategy, this embodiment proposes a method for identifying clusters of similar regions among multiple preset marketing regions. Specifically, a cluster of similar regions refers to a set of preset marketing regions with similar market characteristics and economic conditions. By identifying clusters of similar regions, we can group regions with similar market environments and consumption demands into one category, so as to formulate and implement marketing strategies more precisely.

[0120] In the process of identifying clusters of similar regions, techniques such as cluster analysis can be used to analyze the market characteristic data and economic data of the preset marketing regions. By calculating the similarity and distance between regions, they can be divided into different clusters, and the regions within each cluster have great similarity in market characteristics and economic conditions.

[0121] Step S820, selecting and implementing differentiated marketing strategies from the differentiated marketing strategy set according to different preset marketing regions within the cluster of similar regions to obtain a class-optimal marketing strategy.

[0122] In this embodiment, after identifying the clusters of similar regions, different pre-set marketing regions within the clusters can be used to select and execute differentiated marketing strategies from the set of differentiated marketing strategies. Since the regions within the clusters of similar regions have similar market environments and consumer demands, targeted class-optimal marketing strategies can be formulated based on these common characteristics.

[0123] Step S830, obtain a regional strategy set for a pre-set marketing region within the cluster of similar regions based on the class-optimal marketing strategy.

[0124] In this embodiment, after obtaining the class-optimal marketing strategy, we can generate a regional strategy set for a certain pre-set marketing region within the cluster of similar regions based on this strategy. The regional strategy set is a set of targeted marketing strategies formulated according to the specific situation and needs of this region. When formulating the regional strategy set, factors such as the economic level, competition situation, and consumer behavior of this marketing region can be considered to ensure the effectiveness and adaptability of the strategies.

[0125] Step S840, divide the distributors within the pre-set marketing region into group A and group B, and apply the marketing strategies within the regional strategy set to the distributors in group A and group B respectively to obtain the results of the AB test.

[0126] In this embodiment, in order to evaluate and optimize the effectiveness of the regional strategy set, the distributors within the pre-set marketing region can be divided into group A and group B, and the marketing strategies within the regional strategy set are applied to the two groups of distributors respectively. By comparing the sales data and market performance of the two groups of distributors, we can evaluate the effectiveness and advantages and disadvantages of different marketing strategies, and then further adjust and optimize the regional strategy set to improve the effectiveness and adaptability of the marketing strategies.

[0127] Step S600, determine the optimal marketing strategy for each pre-set marketing region according to the results of the AB test.

[0128] In this embodiment, after completing the AB test, we need to conduct in-depth analysis and comparison of the test results. Specifically, the sales data, market share, customer satisfaction and other indicators of the distributors in group A and group B after adopting different marketing strategies can be compared to evaluate the effectiveness and advantages and disadvantages of different marketing strategies, and then determine the optimal marketing strategy for each pre-set marketing region.

[0129] In a feasible implementation manner, referring to Figure 5 , step S600 includes steps S610 to S630, where:

[0130] Step S610, collect the sales data of the distributors in group A and group B after applying different marketing strategies;

[0131] In this embodiment, in order to determine the optimal marketing strategy within each preset marketing area, this embodiment collects the sales data of Group A and Group B distributors after applying different marketing strategies. The sales data includes but is not limited to key indicators such as sales volume, sales quantity, market share, and customer feedback. By analyzing the sales data, we can understand the performance of different marketing strategies in the actual market and provide data support for subsequent strategy optimization.

[0132] Step S620: Determine the sales growth rate, market share change, and customer satisfaction of Group A and Group B distributors after applying different marketing strategies for the collected sales data.

[0133] In this embodiment, the sales growth rate can reflect the direct impact of the marketing strategy on sales. By comparing the sales growth rates of Group A and Group B distributors after applying different marketing strategies, we can determine which strategy can better promote sales growth. At the same time, the market share change is also one of the important indicators for evaluating the effectiveness of the marketing strategy. It can help us understand the performance of the strategy in enhancing market competitiveness. And customer satisfaction is the key to measuring whether the marketing strategy meets the needs of consumers. By collecting and analyzing customer feedback, we can evaluate the effectiveness of different strategies in enhancing customer satisfaction, and further optimize the strategy to improve customer satisfaction and loyalty.

[0134] Step S630: Determine the optimal marketing strategy within each preset marketing area based on the sales growth rate, market share change, and customer satisfaction.

[0135] In this embodiment, after determining the key indicators such as the sales growth rate, market share change, and customer satisfaction, we can conduct a comprehensive analysis of these indicators to obtain the optimal marketing strategy within each preset marketing area. Specifically, the optimal marketing strategy should be the strategy that performs most prominently after comprehensively considering multiple dimensions such as the sales growth rate, market share increase, and customer satisfaction. This strategy can not only promote the rapid growth of sales volume, but also effectively increase the market share, while meeting the needs of consumers, improving customer satisfaction and loyalty. By applying the optimal marketing strategy to each preset marketing area, we can maximize the effectiveness of the marketing strategy and enhance the market competitiveness and brand influence of the product.

[0136] In a feasible implementation manner, the method further includes:

[0137] Obtain business violation data and abnormal operation data of the preset marketing area, perform pattern recognition on the business violation data and abnormal operation data, build a dealer risk assessment model, and then generate a channel early warning signal corresponding to the preset marketing area based on the risk level coefficient output by the dealer risk assessment model. When the channel early warning signal reaches the preset risk threshold, the dealer qualification review mechanism and emergency replenishment strategy are triggered.

[0138] In this embodiment, the business violation data includes the number of historical violations of the dealer, the type of violation, the amount involved and the rectification situation, while the abnormal operation data covers indicators such as abnormal fluctuations in inventory turnover, decreased order fulfillment rate, and a surge in terminal customer complaints. By using cluster analysis and anomaly detection algorithms to perform pattern recognition on the above data, high-risk violation patterns (such as periodic false orders, cross-regional cross-selling behavior) and abnormal operation characteristics (such as inventory backlog exceeding the threshold, abnormal increase in replenishment frequency) can be extracted. The dealer risk assessment model adopts an integrated learning method combining logistic regression and random forest to weightedly fuse the identified violation pattern feature vector with the abnormal operation index, and output a quantitative risk level coefficient (value range 0~1). When the risk level coefficient reaches the preset threshold (for example, 0.7), the system automatically generates a red channel warning signal and triggers the following mechanisms: 1) The transaction vouchers and logistics data of the dealer in the past 6 months are retrieved through the blockchain evidence module to start the qualification review process; 2) The safety inventory threshold is dynamically adjusted according to the emergency replenishment strategy, and the emergency transfer authority is opened to dealers in adjacent areas, and priority delivery instructions are issued to the intelligent replenishment robot. In this way, automated monitoring and rapid response to high-risk dealers can be achieved.

[0139] In this embodiment, a precision marketing management method based on offline dealers is applied to baijiu products. The method includes collecting dealer inventory data, regional economic data, competitor data, and terminal sales data within multiple preset marketing regions; then determining the corresponding economic level index according to the economic data of each preset marketing region, determining the corresponding competitor penetration rate according to the competitor data of each preset marketing region, and determining the corresponding drinking culture intensity according to the terminal sales data of each preset marketing region; then constructing a regional market characteristic matrix based on the economic level index, competitor penetration rate, and drinking culture intensity of each preset marketing region; and inputting the dealer inventory data of multiple preset marketing regions and the regional market characteristic matrix into a pre-trained double-cycle prediction model to output corresponding long-term sales prediction values and short-term replenishment decision parameter sets; then generating a differentiated marketing strategy set according to the long-term sales prediction values and short-term replenishment decision parameter sets corresponding to multiple preset marketing regions to conduct an AB test based on the differentiated marketing strategy set; finally, determining the optimal marketing strategy within each preset marketing region according to the results of the AB test. That is, by collecting and analyzing dealer inventory data, regional economic data, competitor data, and terminal sales data within multiple preset marketing regions, constructing a regional market characteristic matrix, and combining a pre-trained double-cycle prediction model, corresponding long-term sales prediction values and short-term replenishment decision parameter sets are output, thereby generating a differentiated marketing strategy set and conducting an AB test to determine the optimal marketing strategy within each preset marketing region. In this way, the above method can accurately grasp the characteristics of each regional market, and improve the pertinence and effectiveness of marketing strategies by accurately analyzing a large amount of data and reasonably predicting marketing results, thereby enhancing the marketing effect of baijiu products.

[0140] This application also provides a precision marketing management device based on offline dealers. Refer to Figure 6 , the device includes: a memory 10, a processor 20, and a precision marketing management program based on offline dealers stored on the memory 10 and executable on the processor 20. The precision marketing management program based on offline dealers is configured to implement the steps of the precision marketing management method based on offline dealers.

[0141] The precision marketing management device based on offline dealers provided by this application adopts the precision marketing management method based on offline dealers in the above embodiment, and can enhance the marketing effect of baijiu products. Compared with the prior art, the beneficial effects of the precision marketing management device based on offline dealers provided by this application are the same as those of the precision marketing management method based on offline dealers provided in the above embodiment, and other technical features in the precision marketing management device based on offline dealers are the same as the features disclosed in the method of the above embodiment, and will not be elaborated here.

[0142] The above are only some embodiments of the present application, and thus do not limit the patent scope of the present application. Any equivalent structural transformation made under the technical concept of the present application by using the content of the specification and drawings of the present application, or any direct / indirect application in other related technical fields is included in the patent protection scope of the present application.

Claims

1. A precise marketing management method based on offline distributors, applied to liquor products, is characterized in that The method described above includes: Collecting dealer inventory data, regional economic data, competitor data, and terminal sales data within multiple preset marketing regions; Determining the corresponding economic level index according to the economic data of each preset marketing region, determining the corresponding competitor penetration rate according to the competitor data of each preset marketing region, and determining the corresponding drinking culture intensity according to the terminal sales data of each preset marketing region; Constructing a regional market feature matrix based on the economic level index, competitor penetration rate, and drinking culture intensity of each preset marketing region; Inputting the dealer inventory data of multiple preset marketing regions and the regional market feature matrix into a pre-trained dual-cycle prediction model to output corresponding long-term sales prediction values and short-term replenishment decision parameter sets; Generating a differentiated marketing strategy set according to the long-term sales prediction values and short-term replenishment decision parameter sets corresponding to multiple preset marketing regions to conduct an AB test based on the differentiated marketing strategy set; Determining the optimal marketing strategy within each preset marketing region according to the results of the AB test.

2. The precise marketing management method based on offline distributors as claimed in claim 1, wherein The regional economic data includes regional GDP, per capita disposable income, total retail sales of social consumer goods, and the proportion of liquor consumption expenditure. The step of determining the corresponding economic level index according to the economic data of each preset marketing region includes: Normalizing the regional GDP, per capita disposable income, total retail sales of social consumer goods, and the proportion of liquor consumption expenditure of each preset marketing region; Performing weighted summation on the normalized regional GDP, per capita disposable income, total retail sales of social consumer goods, and the proportion of liquor consumption expenditure according to preset weight distribution to obtain the economic level index of each preset marketing region.

3. The precise marketing management method based on offline dealers according to claim 1, characterized in that, The competitor data includes the total number of sales channel types and the sales amount in each sales channel. The step of determining the corresponding competitor penetration rate according to the competitor data of each preset marketing region includes determining the competitor penetration rate of each preset marketing region according to the following formula: ； Among them, represents the penetration rate of competing products, represents the total number of sales channel types, represents the sales volume of competing products in the preset marketing area sales channel ; represents the sales volume of baijiu products in the preset marketing area sales channel .

4. The precise marketing management method based on offline dealers as claimed in claim 1, wherein The terminal sales data includes the high-frequency purchase rate, the proportion of gift box sales volume, and the peak festival sales volume. The step of determining the corresponding drinking culture intensity according to the terminal sales data of each preset marketing region includes determining the drinking culture intensity of each preset marketing region according to the following formula: ； wherein represents the intensity of the drinking culture; represents the high-frequency purchase rate of the preset marketing area ; represents the highest high-frequency purchase rate among all preset marketing areas, represents the weight corresponding to the high-frequency purchase rate; represents the proportion of the sales volume of gift boxes in the preset marketing area ; represents the highest proportion of the sales volume of gift boxes among all preset marketing areas; represents the weight corresponding to the proportion of the sales volume of gift boxes; represents the peak festival sales volume in the preset marketing area ; represents the highest peak festival sales volume among all preset marketing areas; represents the weight corresponding to the peak festival sales volume; wherein .

5. The precise marketing management method based on offline dealers as claimed in claim 1, wherein The method further includes: Collecting historical dealer inventory data, historical regional economic data, historical competitor data, historical terminal sales data, and corresponding historical sales data; Splitting the historical sales data to obtain long-term sales historical data and short-term sales historical data; Constructing a long-term sales prediction model based on the long-term sales historical data and a short-term replenishment decision model based on the short-term sales historical data; Fusing the long-term sales prediction model and the short-term replenishment decision model to obtain a dual-cycle prediction model; Training the dual-cycle prediction model until the preset prediction accuracy requirement is met to obtain a pre-trained dual-cycle prediction model.

6. The precise marketing management method based on offline distributors as claimed in claim 1, wherein The step of generating a differentiated marketing strategy set according to the long-term sales prediction values and short-term replenishment decision parameter sets corresponding to multiple preset marketing regions to conduct an AB test based on the differentiated marketing strategy set includes: Obtain a preset marketing strategy algorithm library, where the marketing strategy algorithm library includes multiple marketing strategy generation algorithms; Match the characteristics of the long-term sales volume prediction value and the short-term replenishment decision parameter set with the multiple marketing strategy generation algorithms in the marketing strategy algorithm library to obtain the corresponding marketing strategy generation algorithm; Generate a differentiated marketing strategy set based on the selected marketing strategy generation algorithm for AB testing based on the differentiated marketing strategy set.

7. The precise marketing management method based on offline dealers as claimed in claim 6, wherein Before the step of determining the optimal marketing strategy in each preset marketing area according to the results of the AB test, the method further includes: Identify clusters of similar areas among multiple preset marketing areas; Select and execute differentiated marketing strategies from the differentiated marketing strategy set according to different preset marketing areas within the cluster of similar areas to obtain a quasi-optimal marketing strategy; Obtain a regional strategy set for a preset marketing area in the cluster of similar areas based on the quasi-optimal marketing strategy; Divide the distributors in the preset marketing area into group A and group B, and apply the marketing strategies in the regional strategy set to the distributors in group A and group B respectively to obtain the results of the AB test.

8. The precise marketing management method based on offline dealers as claimed in claim 7, wherein The step of determining the optimal marketing strategy in each preset marketing area according to the results of the AB test includes: Collect the sales data of the distributors in group A and group B after applying different marketing strategies; Determine the sales growth rate, market share change, and customer satisfaction of the distributors in group A and group B after applying different marketing strategies for the collected sales data; Determine the optimal marketing strategy in each preset marketing area according to the sales growth rate, market share change, and customer satisfaction.

9. The precise marketing management method based on offline distributors according to claim 1, wherein The method further includes: Obtain the business violation data and abnormal operation data of the preset marketing area; Perform pattern recognition on the business violation data and abnormal operation data to construct a distributor risk assessment model; Generate a channel warning signal for the corresponding preset marketing area according to the risk level coefficient output by the distributor risk assessment model; When the channel warning signal reaches the preset risk threshold, trigger the distributor qualification review mechanism and the emergency replenishment strategy.

10. A precise marketing management device based on offline dealers, characterized in that, The device includes: a memory, a processor, and a precise marketing management program based on offline distributors stored on the memory and executable on the processor. The precise marketing management program based on offline distributors is configured to implement the steps of the precise marketing management method based on offline distributors as described in any one of claims 1 to 9.

Images