Business monitoring method and system based on digital marketing data
Through business monitoring methods based on digital marketing data, real-time monitoring and analysis of marketing data, and establishing monitoring indicators and early warning mechanisms, the problem of unpredictable product hot sales in the existing technology is solved, and accurate monitoring and prediction of product sales status is achieved , help businesses reduce economic losses.
Patent Information
- Application Number
- CN202311372228.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-10-20
- Publication Date
- 2024-12-13
AI Technical Summary
The existing technology cannot effectively predict the hot sales of Internet products, resulting in an increase in space occupied by unsold goods in warehouses and insufficient inventory of hot-selling goods, causing economic losses to merchants, and failing to monitor and handle the abnormal sales status of products.
Using a business monitoring method based on digital marketing data, through real-time monitoring and analysis of marketing data, establish monitoring indicators, conduct product marketing forecasts and grade classification, record and warning in real time, including monitoring of marketing rate, after-sales rate and dynamic change indicators, and use nerves The network makes data predictions.
We have achieved multi-faceted monitoring of product sales status, ensure sales quantity and quality, and take timely measures to maximize merchant interests and reduce economic losses.
Smart Images
Figure CN119130533A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to a method and system for monitoring business based on digital marketing data. Background Art
[0002] With the development of science and technology, the popularity of Internet applications has increased, various application software has continued to develop, and Internet shopping and Internet product sales and marketing platforms have also emerged. Users can choose the products or services they need based on Internet product sales and marketing. With the expansion of marketing, the management and control of commodity marketing has gradually become cumbersome. When there are a large number of commodities, it is impossible to effectively predict the information of popular commodities in the current season, and it is impossible to purchase some hot-selling commodities in advance. In addition, some goods stored in the warehouse cannot be effectively processed after being unsalable, resulting in an increase in ineffective occupied space in the warehouse, causing problems in the management and control between the warehouse and the sales platform, such as unsalable goods stored in the warehouse cannot be processed, and the inventory of hot-selling commodities is insufficient, which leads to economic losses for merchants.
[0003] The existing technology only monitors and handles the sales of products based on maximizing sales, but does not monitor the abnormal sales status and abnormal after-sales status of products. If a product has high sales and a high after-sales processing rate, it will also lead to abnormal sales of the product and cause great economic losses to the merchants. Summary of the invention
[0004] In order to solve the technical problems existing in the background technology, the present invention proposes a business monitoring method and system based on digital marketing data.
[0005] The present invention proposes a digital marketing data service monitoring method, which specifically includes the following steps:
[0006] S1: Real-time monitoring and acquisition of marketing data of users on the marketing platform, and generation of a product database based on the acquired marketing data;
[0007] S2: Establish monitoring indicators based on current marketing data, and monitor marketing data based on the monitoring indicators;
[0008] S3: Make marketing forecasts for products based on their historical marketing data;
[0009] S4: Divide the marketing level of products according to their actual marketing data, record the marketing level of products in real time, and conduct real-time monitoring and early warning based on the recording results.
[0010] Preferably, the marketing data includes product information data, product transaction data, historical search data, after-sales data, and purchased or favorited data.
[0011] Preferably, in S2, a marketing rate indicator, an after-sales rate indicator and a dynamic change indicator are respectively established to monitor the marketing data in real time.
[0012] The marketing rate indicator sets a marketing upper limit τ for the number of products marketed per hour max and marketing lower limit τ min , and determine the efficiency level interval of the marketing quantity of products per hour:
[0013]
[0014] Among them, n i is the marketing rate determination result of the product per hour; τ i (i=0,…,N-1) is a marketing quantity time series with a length of N;
[0015] The marketing efficiency ratio η is as follows:
[0016]
[0017] Among them, sgn(·) is the sign function;
[0018] The after-sales rate indicator calculates the cumulative after-sales quantity G within the time N:
[0019]
[0020] Among them, NS i is the product after-sales volume at time i;
[0021] The dynamic change index is based on the historical marketing data sequence U = {u1, u2, ..., u P} and the current marketing data sequence V = {v1, v2, ..., v Q}Calculate the dynamically changing distance D(U, V):
[0022]
[0023] Where S is the length of the curved path C, d(u p , v q ) is the curved path point (u p , v q )’s cumulative distance.
[0024] Preferably, in the marketing rate indicator, when the marketing quantity of products per hour is lower than the marketing lower limit, it is a low marketing efficiency level; when the marketing quantity of products per hour is between the marketing lower limit and the marketing upper limit, it is a medium marketing efficiency level; when the marketing quantity of products per hour is higher than the marketing upper limit, it is a high marketing efficiency level.
[0025] Preferably, the marketing rate indicator, after-sales rate indicator and dynamic change indicator monitor and warn the marketing data according to the marketing rate indicator threshold, after-sales rate indicator threshold and dynamic change indicator threshold respectively.
[0026] Preferably, in S3, feature extraction is performed on the historical marketing data based on a feature extraction algorithm:
[0027]
[0028] Among them, R is the real-time marketing data of the product; E is the scale of marketing data; A is the number of products; a is the correlation between products; B is the circulation of products, and b is the attribute parameter of marketing data;
[0029] The collected historical marketing data is divided into several categories:
[0030]
[0031] Among them, W is the marketing data partition result; U is the number of partition templates; x is the horizontal distribution of data in space, and y is the vertical distribution of data in space; δ xy is the degree of data discretization;
[0032] Extract characteristic marketing data based on data clustering analysis:
[0033] X=(d s -f k ) 2
[0034] Among them, X is the extracted characteristic marketing data, d is the marketing data mining dimension, f is the marketing data iterative processing coefficient, s is the similarity between marketing data, and k is the difference between marketing data.
[0035] Preferably, in S3, characteristic data is obtained based on the clustering result and characteristic marketing data set X=(X1, X2, ..., X M ) is input into the neural network, and the marketing prediction of the product is performed based on the neural network. The input F(X) of the neural network is:
[0036]
[0037] Among them, ω j represents the weight of the j-th feature marketing data, b represents the bias term, g() represents the activation function,
[0038] The activation function g(X) is as follows:
[0039]
[0040] The weighted sum of the input values of the jth neuron in the lth layer is expressed as:
[0041]
[0042] Among them, M represents neuron data, l represents the number of layers, represents the output weight between the mth neuron in the l-1th layer and the jth neuron in the lth layer, represents the output value of the mth neuron in the l-1th layer, represents the bias term of the jth neuron in the l-1th layer;
[0043]
[0044] in, represents the output value of the mth neuron in the lth layer, Respectively indicate that when the input value is and is the input of the neural network.
[0045] Preferably, in S3, the marketing level of the product is divided according to the historical marketing data of the product:
[0046] When the monthly transaction data of a product is lower than 10% of the average transaction data, the product is classified as a slow-moving product;
[0047] When the monthly transaction data of a product is between 10% and 50% of the average transaction data, the product is classified as a low-end hot-selling product;
[0048] When the monthly transaction data of a product is between 50% and 150% of the average transaction data, the product is classified as a medium-selling product;
[0049] When the monthly transaction data of a product is higher than 200% of the average transaction data, the product is classified as a hot-selling product.
[0050] Preferably, in S4, changes in the marketing level of the product are monitored in real time, and when the marketing level of the product changes by more than two levels, a real-time warning is issued.
[0051] A business monitoring system based on digital marketing data is proposed, including:
[0052] Marketing monitoring module: used to monitor and obtain users' marketing data on the marketing platform in real time, and to generate a product database based on the obtained marketing data;
[0053] Indicator monitoring module: used to establish monitoring indicators based on current marketing data, and monitor marketing data based on the monitoring indicators;
[0054] Marketing forecasting module: used to make marketing forecasts for products based on their historical marketing data;
[0055] Monitoring and early warning module: used to classify the marketing level of products according to their actual marketing data, record the marketing level of products in real time, and conduct real-time monitoring and early warning based on the recording results.
[0056] The present invention proposes a digital marketing data service monitoring method and system, which has the following features:
[0057] Beneficial effects:
[0058] The present invention conducts multi-faceted monitoring of product marketing data by establishing multiple monitoring indicators, performs after-sales inspection of products while monitoring product transaction data, ensures product quality while ensuring product sales quantity, and also conducts product marketing forecasts based on the product's historical marketing data, so that merchants can take timely response measures to maximize their interests. BRIEF DESCRIPTION OF THE DRAWINGS
[0059] Figure 1 A method flow chart of a digital marketing data service monitoring method proposed by the present invention;
[0060] Figure 2 This is a system block diagram of a digital marketing data business monitoring system proposed by the present invention.
[0061] The meaning of the marks in the figure:
[0062] 100. Marketing monitoring module; 200. Index monitoring module; 300. Marketing forecasting module; 400. Monitoring and early warning module. DETAILED DESCRIPTION
[0063] Reference Figure 1 This embodiment provides a method for monitoring digital marketing data services, including:
[0064] S1: Real-time monitoring and acquisition of marketing data of users on the marketing platform, and generation of a product database based on the acquired marketing data;
[0065] Among them, marketing data includes product information data, such as product name, product type, etc.; product transaction data, such as product sales volume data; historical search data; after-sales data such as return data; and purchased or favorited data, including data on users adding the product to the shopping cart or adding it to favorites.
[0066] S2: Establish monitoring indicators based on current marketing data, and monitor marketing data based on the monitoring indicators;
[0067] Establish marketing rate indicators, after-sales rate indicators and dynamic change indicators respectively to monitor marketing data in real time.
[0068] The marketing rate indicator sets a marketing upper limit τ for the number of marketing products per hourmax and marketing lower limit τ min , and determine the efficiency level interval of the marketing quantity of products per hour:
[0069]
[0070] Among them, n i is the marketing rate determination result of the product per hour; τ i (i=0,…,N-1) is a marketing quantity time series with a length of N;
[0071] Among them, n i =-1 corresponds to the marketing quantity of products per hour being lower than the marketing lower limit, which is the marketing low efficiency level, n i =0 corresponds to the marketing quantity of products per hour being between the marketing lower limit and the marketing upper limit, which is the medium marketing efficiency level. i =1 corresponds to the marketing quantity of products per hour being higher than the marketing upper limit, which is the high marketing efficiency level.
[0072] The marketing efficiency ratio η is as follows:
[0073]
[0074] Among them, sgn(·) is the sign function;
[0075] Monitor the marketing efficiency ratios when the marketing efficiency is low and when the marketing efficiency is high, so as to make a quick judgment on the current marketing status.
[0076] After-sales rate indicator calculation time N cumulative after-sales quantity G:
[0077]
[0078] Among them, NS i is the product after-sales volume at time i;
[0079] Dynamic change indicators are based on the historical marketing data sequence U = {u1, u2, ..., u P} and the current marketing data sequence V = {v1, v2, ..., v Q}Calculate the dynamically changing distance D(U, V):
[0080]
[0081] Where S is the length of the curved path C, d(u p , v q ) is the curved path point (u p , v q )’s cumulative distance.
[0082] The marketing rate indicator, the after-sales rate indicator and the dynamic change indicator monitor and warn the marketing data according to the marketing rate indicator threshold, the after-sales rate indicator threshold and the dynamic change indicator threshold respectively.
[0083] By preliminarily monitoring the product's marketing efficiency and after-sales rate indicators, we can grasp the product's marketing status in real time, and measure whether the current marketing rules have changed through dynamic change indicators; by monitoring and warning abnormal marketing data (such as low marketing efficiency level below the low marketing rate indicator threshold, high marketing efficiency level above the high marketing rate indicator threshold, product after-sales volume or abnormal dynamic change indicators), merchants can handle them in a timely manner.
[0084] S3: Make marketing forecasts for products based on their historical marketing data;
[0085] Extract features from historical marketing data based on feature extraction algorithms:
[0086]
[0087] Among them, R is the real-time marketing data of the product; E is the scale of marketing data; A is the number of products; a is the correlation between products; B is the circulation of products, and b is the attribute parameter of marketing data;
[0088] The collected historical marketing data is divided into several categories:
[0089]
[0090] Among them, W is the marketing data partition result; U is the number of partition templates; x is the horizontal distribution of data in space, and y is the vertical distribution of data in space; δ xy is the degree of data discretization;
[0091] Extract characteristic marketing data based on data clustering analysis:
[0092] X=(d s -f k ) 2
[0093] Among them, X is the extracted characteristic marketing data, d is the marketing data mining dimension, f is the marketing data iterative processing coefficient, s is the similarity between marketing data, and k is the difference between marketing data.
[0094] Feature marketing data extracted based on mining technology and clustering processing can grasp the correlation between marketing data and improve the ability to represent marketing data.
[0095] Based on the clustering results, characteristic data is obtained and the characteristic marketing data set X = (X1, X2, ..., X M) is input into the neural network, and the marketing prediction of the product is performed based on the neural network. The input F(X) of the neural network is:
[0096]
[0097] Among them, ω j represents the weight of the j-th feature marketing data, b represents the bias term, g() represents the activation function,
[0098] The activation function g(X) is as follows:
[0099]
[0100] The weighted sum of the input values of the jth neuron in the lth layer is expressed as:
[0101]
[0102] Among them, M represents neuron data, l represents the number of layers, represents the output weight between the mth neuron in the l-1th layer and the jth neuron in the lth layer, represents the output value of the mth neuron in the l-1th layer, represents the bias term of the jth neuron in the l-1th layer;
[0103]
[0104] in, represents the output value of the mth neuron in the lth layer, Respectively indicate that when the input value is and is the input of the neural network.
[0105] By learning the product's marketing data based on the deep learning network, it is possible to more accurately predict the product's current and future marketing status, provide a reference for the product's marketing status, and more accurately judge the product's current marketing health status. It can also reserve products based on the prediction results to maximize the merchant's interests.
[0106] The marketing level of products can be divided according to their historical marketing data:
[0107] When the monthly transaction data of a product is lower than 10% of the average transaction data, the product is classified as a slow-moving product;
[0108] When the monthly transaction data of a product is between 10% and 50% of the average transaction data, the product is classified as a low-end hot-selling product;
[0109] When the monthly transaction data of a product is between 50% and 150% of the average transaction data, the product is classified as a medium-selling product;
[0110] When the monthly transaction data of a product is higher than 200% of the average transaction data, the product is classified as a hot-selling product.
[0111] S4: Divide the marketing level of products according to their actual marketing data, record the marketing level of products in real time, and conduct real-time monitoring and early warning based on the recording results.
[0112] Divide the marketing level of products according to their actual marketing data, record the marketing level of products in real time, and monitor the changes in their marketing levels in real time. When the marketing level of a product changes by more than two levels, for example, the monthly transaction data of the product in January is 1,300 pieces, the monthly transaction data of the product in February is 1,134 pieces, and the monthly transaction data of the product in March is 362 pieces, and the monthly transaction data of the product drops from a hot-selling product to a lower-selling product, the product will be monitored abnormally and warned to remind relevant personnel to handle it.
[0113] Reference Figure 2 , providing a business monitoring system based on digital marketing data, including:
[0114] The marketing monitoring module 100 monitors and obtains the marketing data of users on the marketing platform in real time, and also generates a product database based on the obtained marketing data;
[0115] The indicator monitoring module 200 establishes monitoring indicators according to current marketing data, and monitors the marketing data according to the monitoring indicators;
[0116] The marketing prediction module 300 performs marketing prediction of the product based on the historical marketing data of the product;
[0117] The monitoring and early warning module 400 classifies the marketing level of the product according to the actual marketing data of the product, records the marketing level of the product in real time, and performs real-time monitoring and early warning according to the recording results.
[0118] Embodiment 2
[0119] The present invention is illustrated by specific data. Taking the sales of a certain type of key chain pendant as an example, the data of the key chain pendant sold in the previous year is obtained. The monthly sales in the twelve months are 200 pieces, 195 pieces, 310 pieces, 432 pieces, 475 pieces, 602 pieces, 995 pieces, 332 pieces, 410 pieces, 267 pieces, 235 pieces and 198 pieces respectively; the number of searches for the term "key chain pendant" reaches 200 times each month; the number of purchases or favorites reaches 100 times each month;
[0120] In the prior art, the hot-selling level of a product is classified only by detecting the sales volume of the product within a preset time period, as follows:
[0121] When the sales volume of a product exceeds one thousand pieces within a preset time period, the product information data will be marked as a high-end hot-selling product;
[0122] When the sales volume of a product is between 500 and 1,000 pieces, the product information data is marked as a medium-selling product;
[0123] When the sales volume of a product is between 100 and 500 pieces, the product information data is marked as a low-end hot-selling product;
[0124] When the sales volume of a commodity is between ten and one hundred pieces, the commodity information data is marked as a commodity to be unsold; when the sales volume of a commodity is less than ten pieces, the commodity information data is marked as a commodity to be unsold.
[0125] Through the marketing strategies in the prior art, the number of such key chain pendants sold each month that were returned was 55, 18, 99, 13, 49, 196, 76, 82, 56, 78 and 67 respectively; the merchants did not handle the after-sales situation of the products, resulting in the after-sales volume of the products remaining high, increasing the merchants' after-sales costs, such as after-sales freight, labor handling costs, etc., and ultimately the sales volume of the products may decline irreversibly due to product problems.
[0126] And generate corresponding product marketing strategies based on the product's popularity level classification results; however, for daily necessities with low or high demand, the method of judging their popularity level by sales volume cannot accurately express the product's marketing status. The same marketing strategy cannot maximize the merchant's resource utilization. Moreover, only focusing on the transaction volume of goods while ignoring the quality of products will gradually affect product sales, causing irreparable losses to merchants.
[0127] The present invention monitors the sales situation of this type of key chain pendant according to the marketing rate index, after-sales rate index and dynamic change index of this type of key chain pendant.
[0128] For example, if the daily marketing upper limit is set at 10 and the marketing lower limit is set at 5, the sales of such key chain pendants that exceed the marketing upper limit and are lower than the marketing lower limit will be accumulated respectively. When the marketing threshold is exceeded, such as less than 3 or more than 13, a reminder can be sent to the merchant, and the merchant can prepare stocks in time.
[0129] Merchants also set a daily return limit of 2 based on the daily after-sales processing volume of this type of key chain pendants, such as the number of returns. When the return limit is exceeded, an early warning will be sent to the merchant, who will promptly detect the reason for the return and make timely adjustments based on the reason for the return, thereby increasing the order conversion rate and reducing return costs.
[0130] The merchant monitors product sales in real time according to the present invention, and the number of returns in the first month is 55 pieces. The merchant makes timely improvements based on the reason for returns of such key chain pendants sold in the first month: the key chain pendants are too heavy and not easy to carry. The number of returns in the second month is 45 pieces. The merchant again extracts the reason for returns from users, which is too large, and makes improvements again. The number of returns in the third month is 25 pieces, and the reason for returns is basically multiple or wrong bidding. The number of returns in the following months is less than 30 pieces. The timely improvements reduce the after-sales rate and the after-sales cost.
[0131] Merchants can also issue warnings based on dynamic change indicators, that is, based on whether the current sales patterns of this type of key chain pendants have changed, such as abnormal product transaction volume or abnormal return volume, so that they can take timely measures to reduce unnecessary losses.
[0132] By calculating the relationship between the monthly transaction data and the average transaction data through the present invention, the popularity level of the current product can be reflected more accurately.
[0133] First calculate the average transaction data of the product within 12 months:
[0134]
[0135] January:
[0136] February:
[0137] March:
[0138] April:
[0139] May:
[0140] June:
[0141] July:
[0142] August:
[0143] September:
[0144] October:
[0145] November:
[0146] December:
[0147] When the monthly transaction data of a product is lower than 10% of the average transaction data, the product is classified as a slow-moving product;
[0148] When the monthly transaction data of a product is between 10% and 50% of the average transaction data, the product is classified as a low-end hot-selling product; the corresponding month is January;
[0149] When the monthly transaction data of a product is between 50% and 150% of the average transaction data, the product is classified as a medium-selling product; the corresponding months are February, March, April, May, August, September, October, November and December;
[0150] When the monthly transaction data of a product is higher than 200% of the average transaction data, the product is classified as a hot-selling product; the corresponding months are June and July.
[0151] According to the above-mentioned product classification method, it is possible to make a relatively accurate classification of hot-selling levels according to product types, judge the current sales status of products, provide merchants with relatively accurate and comprehensive marketing status monitoring, and monitor product quality while monitoring marketing volume for merchants, thereby improving merchants' marketing benefits.
[0152] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A method for monitoring business based on digital marketing data, characterized in that: The specific steps include: S1: Real-time monitoring and acquisition of marketing data of users on the marketing platform, and generation of a product database based on the acquired marketing data; S2: Establish monitoring indicators based on current marketing data, and monitor marketing data based on the monitoring indicators; S3: Make marketing forecasts for products based on their historical marketing data; S4: Divide the marketing level of products according to their actual marketing data, record the marketing level of products in real time, and conduct real-time monitoring and early warning based on the recording results.
2. The digital marketing data-based business monitoring method according to claim 1 is characterized in that: The marketing data includes product information data, product transaction data, historical search data, after-sales data, and added purchase or favorite data.
3. The digital marketing data-based business monitoring method according to claim 2 is characterized in that: In S2, a marketing rate indicator, an after-sales rate indicator and a dynamic change indicator are respectively established to monitor the marketing data in real time; The marketing rate indicator sets a marketing upper limit τ for the number of products marketed per hour max and marketing lower limit τ min , and determine the efficiency level interval of the marketing quantity of products per hour: Among them, n o is the marketing rate determination result of the product per hour; τ i (i=0,…,N-1) is a marketing quantity time series with a length of N; The marketing efficiency ratio η is as follows: Among them, sgn(•) is the sign function; The after-sales rate indicator calculates the cumulative after-sales quantity G within the time N: Among them, NS i is the product after-sales volume at time i; The dynamic change index is based on the historical marketing data sequence U = {u1, u2, ..., u P } and the current marketing data sequence V = {v1,v2,…,v Q }Calculate the dynamically changing distance D(U,V): Where S is the length of the curved path C, d(u p ,v q ) is the curved path point (u p ,v q )’s cumulative distance.
4. The digital marketing data-based business monitoring method according to claim 3 is characterized in that: In the marketing rate indicator, when the marketing quantity of products per hour is lower than the marketing lower limit, it is a low marketing efficiency level; when the marketing quantity of products per hour is between the marketing lower limit and the marketing upper limit, it is a medium marketing efficiency level; when the marketing quantity of products per hour is higher than the marketing upper limit, it is a high marketing efficiency level.
5. The digital marketing data-based business monitoring method according to claim 4 is characterized in that: The marketing rate indicator, after-sales rate indicator and dynamic change indicator respectively monitor and warn the marketing data according to the marketing rate indicator threshold, the after-sales rate indicator threshold and the dynamic change indicator threshold.
6. The digital marketing data-based business monitoring method according to claim 5 is characterized in that: In S3, feature extraction is performed on historical marketing data based on a feature extraction algorithm: Among them, R is the real-time marketing data of the product; E is the scale of marketing data; A is the number of products; a is the correlation between products; B is the circulation of products, and b is the attribute parameter of marketing data; The collected historical marketing data is divided into several categories: Among them, W is the marketing data partition result; U is the number of partition templates; x is the horizontal distribution of data in space, and y is the vertical distribution of data in space; δ xy is the degree of data discretization; Extract characteristic marketing data based on data clustering analysis: X=(d s -f k ) 2 Among them, X is the extracted characteristic marketing data, d is the marketing data mining dimension, f is the marketing data iterative processing coefficient, s is the similarity between marketing data, and k is the difference between marketing data.
7. The method for monitoring business based on digital marketing data according to claim 6, characterized in that: In S3, feature data is obtained based on the clustering result, and the feature marketing data set X = (X1, X2, ..., X M ) is input into the neural network, and the marketing prediction of the product is performed based on the neural network. The input F(X) of the neural network is: Among them, ω j represents the weight of the j-th feature marketing data, b represents the bias term, g() represents the activation function, The activation function g(X) is as follows: The weighted sum of the input values of the jth neuron in the lth layer is expressed as: Among them, M represents neuron data, l represents the number of layers, represents the output weight between the mth neuron in the l-1th layer and the jth neuron in the lth layer, represents the output value of the mth neuron in the l-1th layer, represents the bias term of the jth neuron in the l-1th layer; in, represents the output value of the mth neuron in the lth layer, Respectively indicate that when the input value is and is the input of the neural network.
8. The digital marketing data-based business monitoring method according to claim 7 is characterized in that: In S3, the marketing levels of products are divided according to the historical marketing data of the products: When the monthly transaction data of a product is lower than 10% of the average transaction data, the product is classified as a slow-moving product; When the monthly transaction data of a product is between 10% and 50% of the average transaction data, the product is classified as a low-end hot-selling product; When the monthly transaction data of a product is between 50% and 150% of the average transaction data, the product is classified as a medium-selling product; When the monthly transaction data of a product is higher than 200% of the average transaction data, the product is classified as a hot-selling product.
9. The digital marketing data-based business monitoring method according to claim 8, characterized in that: In S4, the change of the marketing level of the product is monitored in real time, and when the marketing level of the product changes by more than two levels, a real-time warning is issued.
10. A digital marketing data business monitoring system, characterized in that: include: Marketing monitoring module (100): used to monitor and obtain the marketing data of users on the marketing platform in real time, and to generate a product database based on the obtained marketing data; Index monitoring module (200): used to establish monitoring indicators according to current marketing data, and monitor the marketing data according to the monitoring indicators; Marketing prediction module (300): used to make marketing predictions for products based on historical marketing data of products; Monitoring and early warning module (400): used to classify the marketing level of products according to the actual marketing data of products, record the marketing level of products in real time, and conduct real-time monitoring and early warning based on the recording results.