A cross-channel online default sale early warning method and device
By integrating sales data from the liquor industry with the BI public business direct connection database, calculating risk coefficients for individual cases and customers, and generating multi-dimensional visual dashboards, the problem of low efficiency, delayed response, and data fragmentation in cross-channel default sales monitoring in the liquor industry has been solved, achieving automated real-time early warning and precise governance.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- LUZHOU LAOJIAO CO LTD
- Filing Date
- 2026-04-27
- Publication Date
- 2026-07-14
Smart Images
Figure CN122390836A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of e-commerce supervision technology, specifically to a cross-channel online default sales early warning method and device. Background Technology
[0002] The traditional sales model in the liquor industry often employs a multi-level distribution system, with products moving from manufacturers through distributors, wholesalers, and retailers before finally reaching consumers. In this process, due to the long information transmission chain, information is prone to distortion or tampering, making it difficult for manufacturers to effectively track the true flow of their products.
[0003] In practice, some distributors, in pursuit of maximizing their own profits, may engage in breach of contract sales practices. For example, they may sell products across regions to profit from regional price differences, thereby disrupting the company's existing market pricing system; or they may use e-commerce platforms to covertly sell counterfeit or substandard products. These actions not only damage brand reputation but also cause direct economic losses to the company. However, due to frequent price fluctuations caused by regional market factors (such as supply and demand, seasonal changes, and promotional activities), companies find it difficult to accurately determine whether a particular sales activity constitutes a breach of contract, further increasing the difficulty of evidence collection and enforcement.
[0004] Currently, businesses primarily rely on manual evidence collection and post-event complaint handling to monitor online sales violations. Common methods include assigning personnel to purchase suspicious goods and tracing the origin of goods through logistics codes. These traditional methods suffer from the following technical limitations: First, the efficiency of evidence collection is low. The manual procurement and traceability process is time-consuming and labor-intensive, the cost of a single evidence collection is high, and the monitoring scope that can be covered is very limited due to the constraints of human and material resources.
[0005] Second, the response is delayed. Traditional methods can only address breaches of sales agreements after they occur, failing to provide early warnings before or in the initial stages of such breaches. This allows low-price, chaotic sales channels to persist for extended periods, continuously impacting the market price system.
[0006] Third, data is severely fragmented. Enterprises possess product warehousing data, sales data, inventory transfer records, and consumer discount records, which are scattered across different business systems. There is a lack of effective integration and analysis methods, making it difficult to quantify and assess the risk level of defaulted sales.
[0007] With the development of technologies such as the Internet of Things and big data, some companies have begun to try embedding QR codes or RFID tags on product packaging to track the product's distribution. Meanwhile, big data analytics is also being used to identify abnormal sales behavior. However, existing solutions still have the following shortcomings: First, the correlation between tracking data and actual sales behavior is not strong enough, making it difficult to distinguish between normal inventory transfers, promotional activities, and genuine breaches of contract; second, the lack of quantitative risk assessment models prevents the automatic, real-time determination of breach of contract risk levels and the triggering of early warnings, making it difficult for companies to take targeted remedial measures in a timely manner. Summary of the Invention
[0008] This invention aims to solve the problems of low efficiency, slow response and data fragmentation in the current monitoring of cross-channel breach of contract in liquor sales, which relies on manual evidence collection. It proposes a method and device for early warning of cross-channel online breach of contract sales.
[0009] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: In a first aspect, the present invention provides a cross-channel online default sales early warning method, the method comprising: Based on the BI public business direct connection database, the regional warehouse entry order data of each distributor is obtained. The regional warehouse entry order data includes the product warehouse entry data of each distributor, the box number and the bottle number associated with the box number. Each bottle number is associated with corresponding sales data, including sales time, sales region, sales price, consumer discount reward records and transfer records. Extract all sales regions corresponding to the same case number based on the sales data of each bottle number, count the number of provinces and cities covered by the sales region, calculate the single case risk coefficient corresponding to the case number based on the number of provinces and cities, and calculate the customer risk coefficient of the distributor based on the single case risk coefficient and the proportion of risky cases associated with the distributor. A risk threshold for the single-box risk coefficient is preset. When the single-box risk coefficient reaches the risk threshold, an early warning is triggered. Based on the customer risk coefficient, a violation risk level for each distributor is established, and a coordinated governance procedure is executed based on the violation risk level. When establishing a violation risk level, the system obtains the transfer records and consumer discount reward records corresponding to the box number. If one or both of the transfer records and consumer discount reward records exist, a note is added to the violation risk level; if no transfer records or consumer discount reward records exist, the violation risk level remains unchanged. When triggering a warning reminder, sort the warning information according to the violation risk level, and generate a multi-dimensional visualization dashboard according to the sales regions associated with the bottle numbers; based on the data associated with the multi-dimensional visualization dashboard, obtain the total number of sales cases, the number of non-compliant cases, and the remaining dealer warehousing data within the current time period, calculate the proportion of the number of non-compliant cases to the total number of sales cases to obtain the regional default ratio, determine the expected number of non-compliant cases flowing in according to the regional default ratio and the remaining dealer warehousing data, and perform marking processing in the multi-dimensional visualization dashboard according to the quantity size of the expected number of non-compliant cases flowing in.
[0010] Furthermore, the calculation method of the single-case risk coefficient includes: judging whether the number of provinces radiated by the same case number is greater than 1. If the number of radiated provinces is greater than 1, the value of the single-case risk coefficient is: the number of provinces + the number of cities / 2; if the number of radiated provinces is not greater than 1, the single-case risk coefficient is 0. The calculation method of the customer risk coefficient includes: customer risk coefficient = total risk coefficient × risk case ratio, where the risk case ratio = the number of risk cases / the number of opened cases, and the total risk coefficient is the sum of the single-case risk coefficients corresponding to all case numbers of the dealer.
[0011] Furthermore, the regional warehousing order data includes the dealer warehousing data of five-code products, and the five-code products are associated with each bottle number by the case number; the consumer preferential reward records include the preferential promotion prices and winning records of the corresponding bottle numbers in the promotion activities.
[0012] Furthermore, the multi-dimensional visualization dashboard includes an overview of the risks of the provinces and cities under assessment and a quadrant chart for analyzing the risks of dealers and terminals; The overview of the risks of the provinces and cities under assessment is used to display the heat distribution of the risks in each region in the form of a map; the quadrant chart for analyzing the risks of dealers and terminals calibrates the positions of high-risk entities with the risk case ratio and the comprehensive risk as the coordinate axes.
[0013] Furthermore, the implementation of the linkage governance program based on the violation risk level specifically includes: If the violation risk level of the dealer is high risk, restrict the dealer's subsequent replenishment authority and push a violation warning letter; If the violation risk level of the dealer is medium and low risk, include the dealer in the key monitoring list and track its subsequent sales data in real time.
[0014] Furthermore, the single-case risk coefficient includes two groups, namely the first group of single-case risk coefficients and the second group of single-case risk coefficients; The first group of single-case risk coefficients: After removing the transfer records and consumer preferential reward records, the pure single-case risk coefficient calculated according to the number of provinces and cities spanned by the sales regions corresponding to the same case number. The second group of single-box risk coefficients: While retaining the aforementioned transfer records and consumer discount reward records, the comprehensive single-box risk coefficients are calculated based on the number of provinces and cities spanned by the sales region corresponding to the same box number.
[0015] Further, after obtaining two sets of single-box risk coefficients, a verification judgment is performed: Retrieve the regional sales benchmark data from the previous time period, and compare the total sales volume, average price fluctuation, and regional distribution of the regional sales benchmark data with the sales data for the current time period; If the sales data for the current time period is better than the sales benchmark data for the region, it is determined that the discount record has a positive effect on sales. The first group of single-box risk coefficients and the second group of single-box risk coefficients are simultaneously marked in the corresponding violation risk level to distinguish between basic risk and risk affected by discounts. If the sales data for the current time period is the same as or lower than the sales benchmark data for the region, the violation risk level based on the single-box risk coefficient of the second group will be directly output, and the specific reasons why the discount did not promote sales will be recorded in the remarks.
[0016] Furthermore, determine the expected number of containers flowing into the area that violate regulations, specifically including: Data is split according to the administrative regions under assessment. The total number of boxes sold and the number of boxes in violation in each region during the current time period are counted. The proportion of boxes in violation to the total number of boxes sold is calculated to obtain the regional default rate. Based on the number of provinces and cities affected by the historical violations of the boxes in this region, the association information of the box numbers in the remaining distributor's warehouse data is matched to determine the potential risk of cross-regional sales. Using the regional default rate as a weight, and combining the potential cross-regional sales risk, the number of boxes that may be added after the remaining distributor inventory data flows into the market is estimated, which is taken as the expected number of boxes flowing into the market in violation of regulations.
[0017] Furthermore, based on the expected number of non-compliant containers, the data is marked on the multi-dimensional visualization dashboard, specifically including: Determine whether the expected number of non-compliant containers has reached a preset container number threshold; If the threshold number of boxes is reached, it will be marked with the first color in the overall risk overview of the assessed provinces and cities, and the dealer position of the corresponding region will be highlighted in the dealer and terminal risk analysis quadrant chart. At the same time, obtain price fluctuation data for the region, including recent average market price, number of low-price links, and frequency of promotional activities; If the price fluctuation data in the region meets the following conditions: the number of low-priced links exceeds the preset low-priced link threshold and the expected number of inflow illegal boxes exceeds the box number threshold, then a price warning enhancement note will be added to the multi-dimensional visualization dashboard. The percentage of defaults in the region and the expected number of boxes flowing into the region that violate regulations are dynamically updated at fixed intervals, and the above-mentioned marking process is re-executed.
[0018] In a second aspect, the present invention provides a cross-channel online default sales early warning device for implementing the cross-channel online default sales early warning method as described in the first aspect, the device comprising: The data acquisition module is used to obtain regional warehouse entry order data of each distributor based on the BI public business direct connection database. The regional warehouse entry order data includes product warehouse entry data, box number and bottle number associated with box number of each distributor. Each bottle number is associated with corresponding sales data, including sales time, sales region, sales price, consumer discount reward record and transfer record. The risk assessment module is used to extract all sales regions corresponding to the same box number based on the sales data of each bottle number, count the number of provinces and cities covered by the sales region, calculate the single box risk coefficient corresponding to the box number based on the number of provinces and cities, and calculate the customer risk coefficient of the distributor based on the single box risk coefficient and the proportion of risk boxes associated with the distributor. The control module is used to preset the risk threshold of the single-box risk coefficient. When the single-box risk coefficient reaches the risk threshold, an early warning is triggered. Based on the customer risk coefficient, a violation risk level is established for each distributor, and a coordinated governance procedure is executed based on the violation risk level. When establishing the violation risk level, the module obtains the transfer records and consumer discount reward records corresponding to the box number. If one or both of these records exist, a note is added to the violation risk level. If neither records exist, the violation risk level remains unchanged. When an early warning is triggered, the module sorts the warning information according to the violation risk level and generates a multi-dimensional visualization dashboard based on the sales region associated with the bottle number. Based on the data associated with the multi-dimensional visualization dashboard, the module obtains the total number of boxes sold, the number of boxes with violations, and the remaining distributor inventory data for the current time period. The module calculates the ratio of the number of boxes with violations to the total number of boxes sold to obtain the regional default ratio. Based on the regional default ratio and the remaining distributor inventory data, the module determines the expected number of boxes with violations flowing in and marks them on the multi-dimensional visualization dashboard according to the expected number of boxes with violations flowing in.
[0019] The beneficial effects of this invention are as follows: The cross-channel online default sales early warning method and device provided by this invention automatically integrates regional warehousing order data, box and bottle association information, sales data, transfer records, and consumer discount reward records of various distributors through the BI public business direct connection database, realizing the unified collection and association of multi-source data; on this basis, the sales behavior is quantitatively evaluated using the single box risk coefficient and customer risk coefficient, and the risk threshold is preset to automatically trigger early warning reminders, transforming the traditional manual post-event evidence collection into the system's automatic real-time early warning, significantly improving monitoring efficiency and response speed; at the same time, by establishing a violation risk level, generating a multi-dimensional visual dashboard, and calculating and marking the expected number of boxes flowing into violation, enterprises can intuitively grasp the regional and channel risk distribution, achieve accurate positioning and pre-emptive management of high-risk distributors, and thus effectively solve the problems of data fragmentation, risk ambiguity, and lagging management in the prior art. Attached Figure Description
[0020] Figure 1 A flowchart illustrating the cross-channel online default sales early warning method provided in this embodiment; Figure 2 A schematic diagram of the cross-channel online default sales early warning device provided in this embodiment. Detailed Implementation
[0021] To enable those skilled in the art to better understand the present invention, the technical solutions in this embodiment will be clearly and completely described below with reference to the accompanying drawings.
[0022] Traditional methods for monitoring sales defaults primarily rely on manual evidence collection and post-event complaint handling, identifying defaults through methods such as manually purchasing suspicious goods and tracing logistics codes. However, this approach has significant technical limitations: First, manual purchasing and traceability processes are time-consuming and labor-intensive, with high costs per instance and limited coverage, resulting in low monitoring efficiency. Second, it can only address defaults after they have occurred, failing to provide pre-emptive warnings, allowing low-price and chaotic links to persist and severely delaying responses. Third, multi-source data, including product warehousing, sales, inventory transfers, and promotions, are scattered across different systems, lacking effective integration and hindering the quantitative assessment of default risks.
[0023] Based on this, the technical solution of the present invention is proposed. In the present invention, firstly, regional warehousing order data of each distributor is automatically obtained from the BI public business direct connection database. This data includes the box number and its associated bottle number. Each bottle number is associated with sales data such as sales time, sales region, sales price, consumer discount reward records, and transfer records. Then, based on the sales regions of all bottle numbers under the same box number, the number of provinces and cities covered are counted, and the single-box risk coefficient of each box number is calculated. Furthermore, the customer risk coefficient of each distributor is calculated by combining the proportion of risky boxes. Then, a risk threshold for the single-box risk coefficient is preset. When the single-box risk coefficient of a certain box number reaches the threshold, an early warning reminder is automatically triggered, and the customer risk is determined accordingly. The system establishes a risk level for each distributor's violations, enabling the implementation of differentiated collaborative governance procedures. During the risk level establishment process, it also checks for records of goods transfers or consumer incentives; if any are found, notes are added to the risk level to reflect the actual business scenario. Finally, when an alert is triggered, the alert information is sorted according to the risk level, and a multi-dimensional visual dashboard is generated based on the sales region. Simultaneously, using the total number of boxes sold, the number of boxes with violations, and the remaining distributor inventory data within the current time period, the system calculates the regional default rate and estimates the expected number of boxes with violations. These are then marked on the dashboard based on their quantity, thus achieving automatic, quantitative, and proactive early warning of cross-channel sales violations.
[0024] The technical solutions in this embodiment will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.
[0025] Figure 1 A flowchart illustrating a cross-channel online default sales warning method is provided. Please refer to [link / reference]. Figure 1 The method includes the following steps: Step 1: Obtain the regional warehouse entry data of each distributor based on the BI public business direct connection database. The regional warehouse entry data includes the product warehouse entry data of each distributor, the box number and the bottle number associated with the box number. Each bottle number is associated with corresponding sales data, including sales time, sales region, sales price, consumer discount reward records and transfer records.
[0026] Specifically, firstly, the system retrieves regional inventory receipt data from each distributor using the BI public business direct connection database. This BI public business direct connection database is a central database within the enterprise used to integrate sales data from various regions, and it can synchronize inventory receipt information from each distributor in real time.
[0027] The regional warehousing data specifically includes: product warehousing data for each distributor, box numbers, and bottle numbers associated with those box numbers. The product warehousing data records the batches and quantities of products received by the distributor at a specific time and in a specific assessment province / city; the box number is a unique identifier assigned to each box; and each box number is associated with multiple bottle numbers, each bottle number being a unique identifier. This association between box and bottle numbers enables refined management from whole boxes to individual bottles. In this embodiment, the product can be liquor, employing a "five-code product" system, where a unique box number and bottle number are printed on both the box and each bottle, and these are linked to the BI public business direct connection database.
[0028] Each bottle number is associated with corresponding sales data. Sales data includes sales time, sales region, sales price, consumer discount and reward records, and restocking records. Specifically, sales time indicates the exact date the bottle was sold; sales region indicates the administrative area (e.g., province, city) where the bottle was sold; sales price indicates the actual transaction price of the bottle; consumer discount and reward records include promotional prices during promotional activities (e.g., discounts, spending thresholds) and prize-winning records (e.g., prize-winning information from scanning QR codes); and restocking records represent the flow information reported by distributors for regional restocking, including the sender, receiver, quantity, and corresponding box number.
[0029] The above data is collected in real time through the BI public business direct connection database, ensuring the timeliness and completeness of the data. For example, when a bottle of liquor is scanned and sold, its sales time, sales region, and sales price are automatically recorded, and associated with the case number to which the bottle belongs and the corresponding distributor information. At the same time, if the sale involves consumer discounts (such as promotional discounts) or involves a transfer of goods, the corresponding records will also be written to the BI public business direct connection database.
[0030] This step unifies and aggregates product warehousing data, sales data, discount records, and transfer records that were originally scattered across different business systems. It also establishes a full-chain data association using "box number - bottle number" as the link, providing a complete and accurate data foundation for subsequent risk assessment.
[0031] Step 2: Extract all sales regions corresponding to the same box number according to the sales data of each bottle number, count the number of provinces and cities covered by the sales region, calculate the single box risk coefficient corresponding to the box number based on the number of provinces and cities; and calculate the customer risk coefficient of the distributor based on the single box risk coefficient and risk box ratio associated with the distributor.
[0032] Specifically, after completing the data collection in step 1, the risk assessment stage begins. This step aims to calculate the single-box risk coefficient corresponding to each box number based on the sales data associated with the bottle number, and further calculate the customer risk coefficient for each distributor.
[0033] First, calculate the risk factor for a single container: In practical applications, based on the sales data of each bottle number, all sales regions corresponding to the same carton number are extracted. Since each carton number is associated with multiple bottle numbers, and each bottle number generates a sales region (usually the province or city where the distributor or retailer is located) when sold, multiple bottle numbers under the same carton number may correspond to multiple different sales regions. Based on these sales regions, the number of provinces and cities covered by that carton number is calculated.
[0034] It should be noted that the number of provinces covered refers to the number of different provinces covered by the sales regions of all bottles under the same case number, while the number of cities covered refers to the number of different cities covered. For example, if a case number contains 10 bottles of liquor, of which 4 are sold in Chengdu, Sichuan Province, 3 in Mianyang, Sichuan Province, 2 in Chongqing, and 1 in Guiyang, Guizhou Province, then the number of provinces covered is 3 (Sichuan, Chongqing, Guizhou), and the number of cities covered is 4 (Chengdu, Mianyang, Chongqing, Guiyang).
[0035] Based on the statistically obtained number of provinces and cities, the single-box risk coefficient is calculated according to the following rules: Determine whether the number of provinces covered by the same box number is greater than 1; if the number of provinces covered is greater than 1, the single-box risk coefficient is: number of provinces + number of cities / 2; if the number of provinces covered is not greater than 1 (i.e., sales are only within one province), the single-box risk coefficient is 0.
[0036] In the above calculation formula, when a box of products is sold to multiple provinces, it indicates cross-regional sales and carries the risk of default; the more provinces and cities involved, the higher the risk level. The number of cities is weighted at half, reflecting the contribution of the degree of city dispersion to the risk.
[0037] For example, if a box of products serves 2 provinces and 3 cities, then the risk coefficient per box is 2 + 3 / 2 = 3.5. Another box of products is sold only within one province; regardless of the number of cities, the risk coefficient per box is 0.
[0038] Then, a customer risk coefficient is calculated to comprehensively assess the overall default risk tendency of the dealer: In practical application, firstly, determine which of the distributor's box numbers belong to the "risk boxes". A risk box is defined as a box number with a single-box risk coefficient greater than 0. The number of risk boxes is the total number of these box numbers. The number of opened boxes refers to the total number of box numbers for which the distributor has been opened and sold (i.e., has a sales record). Then, calculate the risk box percentage using the formula: Risk Box Percentage = Number of Risk Boxes / Number of Opened Boxes. Next, calculate the distributor's total risk coefficient, which is equal to the sum of the single-box risk coefficients for all box numbers for that distributor. Finally, calculate the customer risk coefficient using the following formula: Customer Risk Coefficient = Total Risk Coefficient × Risk Box Percentage.
[0039] In the above calculation formula, the total risk coefficient reflects the absolute intensity of the dealer's default behavior (the severity of cross-regional sales), and the risk box percentage reflects the prevalence of the default behavior (what percentage of boxes involve cross-regional sales). Multiplying the two yields a comprehensive customer risk coefficient; the higher the value, the greater the dealer's default risk.
[0040] For example, a distributor sells a total of 100 cases, of which 20 cases are sold across provinces (the risk coefficients for each case are 3, 4, 5, etc., totaling 80). The total risk coefficient is 80, the proportion of risky cases is 20 / 100 = 0.2, and the customer risk coefficient is 80 × 0.2 = 16. If another distributor sells 50 cases, of which 10 cases are sold across provinces (the total risk coefficient for each case is 30), then the customer risk coefficient is 30 × (10 / 50) = 6, which is lower than the former.
[0041] Through the above calculations, the original sales region information is transformed into quantified single-box risk coefficients and customer risk coefficients, providing numerical basis for subsequent threshold judgment, level classification, and early warning triggering.
[0042] Step 3: Preset the risk threshold of the single-box risk coefficient. When the single-box risk coefficient reaches the risk threshold, trigger an early warning reminder. Establish the violation risk level of each distributor based on the customer risk coefficient, and execute the linkage governance procedure based on the violation risk level.
[0043] Specifically, after calculating the single-box risk coefficient for each box number and the customer risk coefficient for each distributor in step 2, the process proceeds to the early warning judgment and handling stage. This step includes three sub-steps: Step 301: Preset risk threshold and trigger early warning alert: Specifically, a risk threshold for the risk coefficient of a single container is preset. In this embodiment, the risk threshold is set to 2. Since the risk coefficient of a single container is only greater than 0 when the number of provinces affected is greater than 1 (when the number of provinces affected is 2, the risk coefficient of a single container is at least 2 + 1 / 2 = 2.5), setting the threshold to 2 is equivalent to determining whether the risk coefficient of a single container is greater than 0 and whether it crosses at least two provinces. Enterprises can also adjust the specific value of the threshold according to actual control needs, for example, setting the threshold to 3 to filter minor cross-regional behaviors.
[0044] In practical applications, the risk coefficient of each container number is monitored in real time. Once the risk coefficient of a container number reaches a preset risk threshold, an early warning is immediately triggered. Early warnings can be sent via system pop-ups, push notifications, email notifications, or other means to inform channel management personnel of sales activities that pose a risk of breach of contract.
[0045] Step 302: Establish the dealer's violation risk level: Upon triggering the alert, a violation risk level is established for each distributor based on the customer risk coefficient calculated in step 2. The violation risk level can be segmented: for example, distributors with a customer risk coefficient greater than or equal to 10 are marked as "high risk," those with a coefficient between 5 and 10 are marked as "medium risk," and those with a coefficient less than 5 are marked as "low risk." Specific segmentation thresholds can be adjusted by the company based on historical data and control measures.
[0046] It should be noted that the establishment of the violation risk level and the triggering of the early warning can be carried out simultaneously or sequentially. The early warning focuses on identifying the breach of contract for a specific box number, while the violation risk level focuses on the overall risk profile of the distributor.
[0047] Step 303: Execute the coordinated governance procedure: Specifically, this embodiment implements differentiated collaborative governance procedures based on the determined level of violation risk of the dealer. This embodiment provides two typical governance measures: If a distributor's violation risk level is high, their subsequent replenishment privileges will be restricted. Specifically, when the distributor submits a new product warehousing application to the company, the application will be automatically blocked or frozen until the risk is eliminated. Simultaneously, a violation warning letter will be sent to the distributor and their superiors, listing evidence such as the number of the defaulting boxes, the affected provinces, and the risk level.
[0048] If a distributor's violation risk level is classified as low to medium (including medium and low risk), replenishment will not be immediately restricted. Instead, the distributor will be added to a key monitoring list. During the monitoring period, the distributor's subsequent sales data will be tracked in real time, including new sales times, sales regions, and sales prices. If new violations are discovered during the monitoring period that cause the customer's risk level to rise to high risk, high-risk control measures will be escalated.
[0049] The aforementioned joint governance procedures can be configured according to the company's actual management strategy, such as increasing fines or canceling dealer qualifications, and other more stringent measures.
[0050] This step achieves a complete closed loop from risk values to early warning signals and then to governance actions, transforming traditional passive post-event handling into automatic, timely, and tiered pre-emptive intervention, effectively curbing the spread of default sales behavior.
[0051] In this embodiment, when establishing the violation risk level, the transfer records and consumer discount reward records corresponding to the box number are obtained. If one or both of the transfer records and consumer discount reward records exist, a note is added to the violation risk level; if no transfer records and consumer discount reward records exist, the violation risk level remains unchanged.
[0052] Specifically, in step 3, when establishing the dealer violation risk level, in addition to classifying the level based on the customer risk coefficient, it is also necessary to consider the impact of inventory transfer records and consumer discount reward records on risk assessment. This is because in actual business scenarios, some cross-regional sales activities may stem from compliant inventory transfer reporting between dealers, or from consumers participating in promotional activities resulting in purchases from other locations. These situations should not be simply judged as breaches of contract. Therefore, the following note processing procedure should be followed when establishing the violation risk level.
[0053] First, for each box number that triggered the alert, retrieve the corresponding stock transfer record and consumer discount reward record. The stock transfer record is submitted by the distributor according to the company's prescribed procedures, and includes the sending distributor, the receiving distributor, the box number, the transfer time, and the approval status. The consumer discount reward record comes from the sales data associated with the bottle number, including promotional prices (such as discounts and spending thresholds) and prize-winning records (such as QR code prizes and cashback rewards).
[0054] Then, determine whether there are any records of product transfers or consumer discount / rewards: If there are records of goods transfers (i.e., the case number has been reported by the distributor as a compliant transfer), or records of consumer incentive rewards (i.e., the sales associated with the bottle number include promotional offers or prizes), or both, a note will be added to the current distributor's corresponding violation risk level. The note can be marked as "includes goods transfer factors," "includes incentive factors," or "includes both goods transfer and incentive factors," depending on which type of record actually exists. The purpose of adding the note is to remind the company that this violation risk level includes elements of normal business operations, which should be treated differently in subsequent governance to avoid misjudgment.
[0055] If there are neither records of goods transfers nor records of consumer discounts or rewards, the original violation risk level will remain unchanged, and no further notes will be added. In this case, the cross-regional sales of this box number lacks a compliant explanation and is considered a breach of contract.
[0056] The added risk level for violations will be displayed with a special identifier in the subsequent multi-dimensional visualization dashboard. For example, in the dealer risk list, a "stock transfer" or "discount" label will be displayed next to the level with the added notes. When the company views the details of a high-risk dealer, the reason for the note and the corresponding record details will be displayed at the same time, making it easier for managers to quickly identify whether there is any interference from compliant stock transfers or promotional activities.
[0057] For example, Distributor A has a case of product numbered BX001, of which 5 bottles are sold in Sichuan Province and 3 bottles in Chongqing. The risk coefficient for this case is calculated as 2 + 2 / 2 = 3 (number of provinces covered: 2; number of cities covered: 2), reaching the risk threshold and triggering an alert. When establishing the violation risk level for Distributor A, the corresponding transfer record for this case number is retrieved, showing that Distributor A has reported to the company that some products have been transferred to Distributor B in Chongqing, and the transfer record approval status is "approved." Simultaneously, it is also discovered that one bottle of wine sold in Chongqing has a consumer discount reward record (a 10 RMB red envelope won). Because both transfer records and consumer discount reward records exist, a note is added to Distributor A's violation risk level: "Includes transfer and discount factors." When managers review the alert list, seeing this note allows them to determine that the alert is not a complete breach of contract, thus avoiding overly drastic control measures against Distributor A.
[0058] Conversely, if distributor C sells a box of products across provinces, but no stock transfer records or consumer discount / reward records can be found, its violation risk level will remain unchanged, and no notes will be made. This distributor will be deemed to be in breach of contract, triggering governance procedures such as restrictions on replenishment.
[0059] Through the above-mentioned remarks processing mechanism, while maintaining the strictness of risk assessment, it also takes into account the actual scenarios of goods transfer and promotion in operation, effectively reducing the misjudgment rate and improving the flexibility and scientific nature of channel management.
[0060] In this embodiment, when an early warning is triggered, the warning information is sorted according to the violation risk level, and a multi-dimensional visualization dashboard is generated according to the sales region associated with the bottle number. Based on the data associated with the multi-dimensional visualization dashboard, the total number of boxes sold, the number of boxes with violations, and the remaining distributor inventory data within the current time period are obtained. The proportion of the number of boxes with violations to the total number of boxes sold is calculated to obtain the regional violation ratio. Based on the regional violation ratio and the remaining distributor inventory data, the expected number of boxes with violations is determined, and the expected number of boxes with violations is marked on the multi-dimensional visualization dashboard according to the size of the expected number of boxes with violations.
[0061] Specifically, after triggering the alert, the alert output step is executed. This step includes the following three sub-steps: Step 311: Sort by violation risk level and generate a multi-dimensional visualization dashboard: Based on the violation risk levels established in step 3, all warning messages that triggered alerts were sorted. The sorting rule was as follows: high-risk warning messages were listed first, followed by medium-risk messages, and low-risk messages were listed last; within the same level, warning messages could be further sorted from high to low based on the customer's risk coefficient.
[0062] After sorting, a multi-dimensional visualization dashboard is generated based on the sales region associated with each bottle number. The sales region refers to the administrative region (province, city) recorded when each bottle number was sold. The dashboard uses the sales region as the spatial dimension, aggregating and displaying alert information by region.
[0063] Specifically, the multi-dimensional visualization dashboard includes an overview of risks in the assessed provinces and cities, as well as a quadrant chart for risk analysis of distributors and terminals. The overview of risks in the assessed provinces and cities is used to display the risk distribution heatmap of each region in map form. For example, different shades of color represent the number of risk boxes or the total customer risk coefficient in each province and city, with darker colors indicating higher risks. The quadrant chart for risk analysis of distributors and terminals uses the proportion of risk boxes and overall risk as coordinate axes to mark the location of high-risk entities. Specifically, the proportion of risk boxes can be used as the horizontal axis and overall risk as the vertical axis to mark the location of each distributor and terminal retail point in a two-dimensional coordinate system. The calculation method for the proportion of risk boxes is the same as in step 2 (number of risk boxes / number of opened boxes), and the overall risk can be the customer risk coefficient or the total risk coefficient. The position located in the upper right corner of the quadrant chart indicates a high proportion and high risk, requiring special attention.
[0064] Step 312: Calculate the regional default rate and determine the expected number of non-compliant containers flowing in, specifically including: Data is split according to the administrative regions under assessment. The total number of boxes sold and the number of boxes in violation in each region during the current time period are counted. The proportion of boxes in violation to the total number of boxes sold is calculated to obtain the regional default rate. Based on the number of provinces and cities affected by the historical violations of the boxes in this region, the association information of the box numbers in the remaining distributor's warehouse data is matched to determine the potential risk of cross-regional sales. Using the regional default rate as a weight, and combining the potential cross-regional sales risk, the number of boxes that may be added after the remaining distributor inventory data flows into the market is estimated, which is taken as the expected number of boxes flowing into the market in violation of regulations.
[0065] Specifically, the data is first split according to the administrative region being assessed. The total number of boxes sold and the number of non-compliant boxes in each region within the current time period are then calculated, determining the proportion of non-compliant boxes to the total number of boxes sold, thus obtaining the regional default rate for that region. Next, based on the number of provinces and cities historically affected by non-compliant boxes in that region, the association information of box numbers in the remaining distributor's inventory data is matched. For example, if historically non-compliant boxes were mostly sold across two or more provinces, the remaining inventory products are deemed to have a high potential risk of cross-regional sales. Finally, using the regional default rate as the base weight, a comprehensive calculation is performed based on the aforementioned potential cross-regional sales risk. The regional default rate is multiplied by the number of boxes remaining in the distributor's inventory, and the product result is adjusted according to historical radiation characteristics (e.g., increasing the weight of regions with higher risks). This yields the expected number of non-compliant boxes that may be added after the remaining distributor's inventory data enters the market, which is the anticipated number of non-compliant boxes entering the market. In this way, by combining the current default rate with historical cross-regional behavior characteristics, a forward-looking quantitative assessment of the future default risk of unsold products is achieved.
[0066] Step 313: Mark and process the boxes according to the expected number of non-compliant boxes: Specifically, based on the calculated number of expected non-compliant containers, a marking process is performed on a multi-dimensional visualization dashboard. The marking rules are as follows: A preset threshold for the number of boxes is established (e.g., 50 boxes). The system determines whether the expected number of boxes flowing into the region to violate regulations reaches this threshold. If it does, the region is marked in the overall risk overview of the assessed province / city with the first color (e.g., red), and the corresponding distributor in that region is highlighted in the distributor and terminal risk analysis quadrant chart (e.g., by increasing the icon size or adding a border). Simultaneously, price fluctuation data for that region is obtained, including the recent average market price, the number of low-price links, and the frequency of promotional activities. If the number of low-price links in that region exceeds the preset threshold (e.g., 10 links), and the expected number of boxes flowing into the region to violate regulations exceeds the threshold, a "Price Warning Enhancement" note is added to the multi-dimensional visualization dashboard to remind the company that the region may face the risk of a price system collapse. The regional default rate and the expected number of boxes flowing into the region to violate regulations are dynamically updated at fixed intervals (e.g., 24 hours), and the above marking process is re-executed to ensure the timeliness of the warning information.
[0067] For example, consider the monitoring during a liquor company's 618 promotional period. After an alert was triggered, a visual dashboard was generated, sorted by violation risk level. In the overall risk overview of assessed provinces and cities, Guangdong Province was displayed in dark red, with a regional default rate of 20%, 8,000 remaining cases in storage, and an expected inflow of 1,600 cases of non-compliant products, exceeding the preset threshold of 100 cases. Guangdong Province was marked in red on the map, and the icons of the top 5 distributors in Guangdong Province were enlarged and highlighted in the distributor quadrant chart. Simultaneously, 25 recent low-price links in Guangdong Province were detected, exceeding the threshold of 15, prompting the addition of a "Price Warning Enhanced" note. Upon seeing the dashboard, company management immediately launched a special inspection of distributors in Guangdong Province, successfully intercepting approximately 1,500 cases of expected non-compliant products from entering the market.
[0068] Through the aforementioned early warning output steps, complex risk data is transformed into intuitive and actionable visual information. By marking the expected number of boxes flowing into non-compliant areas, a leap from "post-event statistics" to "pre-event prediction" is achieved, providing forward-looking decision support for enterprise channel management.
[0069] In this embodiment, the single-container risk coefficient includes two sets, namely the first set of single-container risk coefficients and the second set of single-container risk coefficients; The first group of single-box risk coefficients: After removing the aforementioned transfer records and consumer discount reward records, the pure single-box risk coefficients are calculated based on the number of provinces and cities spanned by the sales region corresponding to the same box number; The second group of single-box risk coefficients: While retaining the aforementioned transfer records and consumer discount reward records, the comprehensive single-box risk coefficients are calculated based on the number of provinces and cities spanned by the sales region corresponding to the same box number.
[0070] After obtaining the risk coefficients of two sets of single containers, a verification judgment is performed: Retrieve the regional sales benchmark data from the previous time period, and compare the total sales volume, average price fluctuation, and regional distribution of the regional sales benchmark data with the sales data for the current time period; If the sales data for the current time period is better than the sales benchmark data for the region, it is determined that the discount record has a positive effect on sales. The first group of single-box risk coefficients and the second group of single-box risk coefficients are simultaneously marked in the corresponding violation risk level to distinguish between basic risk and risk affected by discounts. If the sales data for the current time period is the same as or lower than the sales benchmark data for the region, the violation risk level based on the single-box risk coefficient of the second group will be directly output, and the specific reasons why the discount did not promote sales will be recorded in the remarks.
[0071] Specifically, in actual sales scenarios, inventory transfer records and consumer incentive records may interfere with the risk assessment of cross-regional sales activities. For example, compliant inventory transfers by distributors may result in products flowing to other provinces, and consumer participation in promotional activities (such as discounts or prizes) may also lead to purchases from other regions. To avoid misjudging these normal business activities as breaches of contract, this embodiment provides a set of sales verification measures in addition to the core risk assessment process. By calculating two sets of single-box risk coefficients and performing comparative judgments, the level of violation risk is finely calibrated.
[0072] In practical applications, for the same container number, two sets of single-container risk coefficients are calculated separately: The first group of single-box risk coefficients (pure single-box risk coefficients): In calculation, the influence of stock transfer records and consumer discount / reward records is actively removed. The coefficient is calculated solely based on the number of provinces and cities spanned by the sales region corresponding to the same box number, using the formula: (Number of provinces + Number of cities / 2). This coefficient reflects the pure physical sales geographic distribution, without considering any compliant stock transfers or promotional factors.
[0073] The second group of single-box risk coefficients (comprehensive single-box risk coefficients): When calculating, transfer records and consumer discount / reward records are retained. Similarly, the calculation is based on the number of provinces and cities spanned by the sales region corresponding to the same box number, using the formula: number of provinces + number of cities / 2. The first group of single-box risk coefficients reflects the actual geographical distribution of sales, including transfer and discount factors.
[0074] Both sets of coefficients are calculated based on the same number of provinces and cities. The only difference lies in whether transfer and discount records are removed during the data preprocessing stage. For example, if a product with a certain box number is actually sold across 2 provinces and 3 cities, and some of these sales are accompanied by compliant transfer reports, the pure single-box risk coefficient is still 2 + 3 / 2 = 3.5, while the comprehensive single-box risk coefficient may also be 3.5 because the transfer records are retained. However, if the transfer records are identified as compliance factors by the system, the first set of single-box risk coefficients will be used as the "no-interference" benchmark, while the second set of single-box risk coefficients will be used as the "actual observation" value.
[0075] After obtaining the two sets of coefficients, the regional sales benchmark data (including total sales, average price fluctuations, and regional distribution) from the previous time period is retrieved and compared with the sales data for the current time period. If the current sales data is better than the benchmark data (e.g., increased sales volume, stable average price), the promotional record is deemed to have a positive promoting effect. The first and second sets of single-box risk coefficients are simultaneously marked in the corresponding violation risk level to distinguish between basic risk and risk affected by the promotion. If the current sales data is the same as or lower than the benchmark data, the violation risk level based on the second set of single-box risk coefficients is directly output, and the specific reasons why the promotion did not promote sales are recorded in the remarks (e.g., the magnitude of sales volume change, price fluctuation range). Through the above judgment, misjudgments caused by promotional activities can be avoided.
[0076] In summary, the cross-channel online default sales early warning method provided in this embodiment integrates regional warehousing order data, box and bottle association information, sales data, transfer records, and consumer discount reward records from the BI public business direct connection database. It constructs a quantitative assessment model from single-box risk coefficient to customer risk coefficient, and automatically triggers early warning reminders by preset risk thresholds. This transforms traditional manual post-event processing into automatic real-time early warning, significantly improving monitoring efficiency and response speed. The cost of single-item evidence collection is greatly reduced, coverage is expanded nationwide, and the efficiency of risk box detection is improved compared to traditional methods. Simultaneously, by building… By establishing a risk level system for violations and generating a multi-dimensional visual dashboard that includes an overview of risks in the assessed provinces and cities, as well as quadrant diagrams for risk analysis of distributors and terminals, and adding notes based on inventory transfer and promotion records, the system enables precise identification and differentiated management of high-risk distributors, thereby improving the rate of removal of defaulted links from shelves. Furthermore, by calculating the regional default rate and estimating the expected number of boxes flowing into the market in violation, combined with sales verification measures to distinguish between promotional interference and genuine defaults, and regional price verification measures to strengthen price risk early warning, the system achieves proactive risk management from passive governance to active prevention, effectively protecting the brand's pricing system and comprehensively reducing the company's channel control costs.
[0077] Based on the above technical solution, this embodiment also provides a cross-channel online default sales early warning device, used to implement the cross-channel online default sales early warning method as described in the embodiment. Please refer to [link to relevant documentation]. Figure 2 The device includes: The data acquisition module is used to obtain regional warehouse entry order data of each distributor based on the BI public business direct connection database. The regional warehouse entry order data includes product warehouse entry data, box number and bottle number associated with box number of each distributor. Each bottle number is associated with corresponding sales data, including sales time, sales region, sales price, consumer discount reward record and transfer record. The risk assessment module is used to extract all sales regions corresponding to the same box number based on the sales data of each bottle number, count the number of provinces and cities covered by the sales region, calculate the single box risk coefficient corresponding to the box number based on the number of provinces and cities, and calculate the customer risk coefficient of the distributor based on the single box risk coefficient and the proportion of risk boxes associated with the distributor. The control module is used to preset the risk threshold of the single-box risk coefficient. When the single-box risk coefficient reaches the risk threshold, an early warning is triggered. Based on the customer risk coefficient, a violation risk level is established for each distributor, and a coordinated governance procedure is executed based on the violation risk level. When establishing the violation risk level, the module obtains the transfer records and consumer discount reward records corresponding to the box number. If one or both of these records exist, a note is added to the violation risk level. If neither records exist, the violation risk level remains unchanged. When an early warning is triggered, the module sorts the warning information according to the violation risk level and generates a multi-dimensional visualization dashboard based on the sales region associated with the bottle number. Based on the data associated with the multi-dimensional visualization dashboard, the module obtains the total number of boxes sold, the number of boxes with violations, and the remaining distributor inventory data for the current time period. The module calculates the ratio of the number of boxes with violations to the total number of boxes sold to obtain the regional default ratio. Based on the regional default ratio and the remaining distributor inventory data, the module determines the expected number of boxes with violations flowing in and marks them on the multi-dimensional visualization dashboard according to the expected number of boxes with violations flowing in.
[0078] It is understood that since the cross-channel online default sales warning device described in this embodiment is a device for implementing the cross-channel online default sales warning method described in the embodiment, the device disclosed in the embodiment is relatively simple to describe because it corresponds to the method disclosed in the embodiment. For relevant parts, please refer to the description of the method, and it will not be repeated here.
Claims
1. A cross-channel online default sales early warning method, characterized in that, The method includes: Based on the BI public business direct connection database, the regional warehouse entry order data of each distributor is obtained. The regional warehouse entry order data includes the product warehouse entry data of each distributor, the box number and the bottle number associated with the box number. Each bottle number is associated with corresponding sales data, including sales time, sales region, sales price, consumer discount reward records and transfer records. Extract all sales regions corresponding to the same case number based on the sales data of each bottle number, count the number of provinces and cities covered by the sales region, calculate the single case risk coefficient corresponding to the case number based on the number of provinces and cities, and calculate the customer risk coefficient of the distributor based on the single case risk coefficient and the proportion of risky cases associated with the distributor. A risk threshold for the single-box risk coefficient is preset. When the single-box risk coefficient reaches the risk threshold, an early warning is triggered. Based on the customer risk coefficient, a violation risk level for each distributor is established, and a coordinated governance procedure is executed based on the violation risk level. When establishing a violation risk level, the system obtains the transfer records and consumer discount reward records corresponding to the box number. If one or both of the transfer records and consumer discount reward records exist, a note is added to the violation risk level; if no transfer records or consumer discount reward records exist, the violation risk level remains unchanged. When an early warning is triggered, the warning information is sorted according to the violation risk level, and a multi-dimensional visualization dashboard is generated according to the sales region associated with the bottle number. Based on the data associated with the multi-dimensional visualization dashboard, the total number of boxes sold, the number of boxes with violations, and the remaining distributor inventory data within the current time period are obtained. The proportion of the number of boxes with violations to the total number of boxes sold is calculated to obtain the regional violation ratio. Based on the regional violation ratio and the remaining distributor inventory data, the expected number of boxes with violations is determined, and the expected number of boxes with violations is marked on the multi-dimensional visualization dashboard according to the size of the expected number of boxes with violations.
2. The cross-channel online default sales early warning method according to claim 1, characterized in that, The method for calculating the single-container risk coefficient includes: determining whether the number of provinces covered by the same container number is greater than 1; if the number of provinces covered is greater than 1, the single-container risk coefficient is: number of provinces + number of cities / 2; if the number of provinces covered is not greater than 1, the single-container risk coefficient is 0. The calculation method for the customer risk coefficient includes: Customer risk coefficient = Total risk coefficient × Risk box percentage, where risk box percentage = Number of risk boxes / Number of opened boxes, and the total risk coefficient is the sum of the single box risk coefficients for all box numbers corresponding to the distributor.
3. The cross-channel online default sales early warning method according to claim 1, characterized in that, The regional warehouse entry data includes distributor warehouse entry data for the five-code products, and the five-code products are associated with each bottle number using the box number; the consumer discount reward record includes the promotional price and prize winning records of the corresponding bottle number in promotional activities.
4. The cross-channel online default sales early warning method according to claim 1, characterized in that, The multi-dimensional visualization dashboard includes an overview of risks in the assessed provinces and cities, as well as quadrant diagrams for risk analysis of dealers and terminals. The risk overview of the assessed provinces and cities is used to display the risk distribution heat map of each region. The risk analysis quadrant diagram for distributors and terminals uses the risk box percentage and overall risk as coordinate axes to mark the position of high-risk entities.
5. The cross-channel online default sales early warning method according to claim 1, characterized in that, The governance procedure based on the aforementioned violation risk level specifically includes: If the dealer's violation risk level is high, then the dealer's subsequent replenishment rights will be restricted, and a violation warning letter will be sent. If the dealer's violation risk level is low to medium, the dealer will be included in the key monitoring list, and its subsequent sales data will be tracked in real time.
6. The cross-channel online default sales early warning method according to claim 1, characterized in that, The single-container risk coefficient includes two groups, namely the first group of single-container risk coefficients and the second group of single-container risk coefficients. The first group of single-box risk coefficients: After removing the aforementioned transfer records and consumer discount reward records, the pure single-box risk coefficients are calculated based on the number of provinces and cities spanned by the sales region corresponding to the same box number; The second group of single-box risk coefficients: While retaining the aforementioned transfer records and consumer discount reward records, the comprehensive single-box risk coefficients are calculated based on the number of provinces and cities spanned by the sales region corresponding to the same box number.
7. The cross-channel online default sales early warning method according to claim 6, characterized in that, After obtaining the risk coefficients of two sets of single containers, a verification judgment is performed: Retrieve the regional sales benchmark data from the previous time period, and compare the total sales volume, average price fluctuation, and regional distribution of the regional sales benchmark data with the sales data for the current time period; If the sales data for the current time period is better than the sales benchmark data for the region, it is determined that the discount record has a positive effect on sales. The first group of single-box risk coefficients and the second group of single-box risk coefficients are simultaneously marked in the corresponding violation risk level to distinguish between basic risk and risk affected by discounts. If the sales data for the current time period is the same as or lower than the sales benchmark data for the region, the violation risk level based on the single-box risk coefficient of the second group will be directly output, and the specific reasons why the discount did not promote sales will be recorded in the remarks.
8. The cross-channel online default sales early warning method according to claim 1, characterized in that, Determine the expected number of containers flowing into the illegal area, specifically including: Data is split according to the administrative regions under assessment. The total number of boxes sold and the number of boxes in violation in each region during the current time period are counted. The proportion of boxes in violation to the total number of boxes sold is calculated to obtain the regional default rate. Based on the number of provinces and cities affected by the historical violations of the boxes in this region, the association information of the box numbers in the remaining distributor's warehouse data is matched to determine the potential risk of cross-regional sales. Using the regional default rate as a weight, and combining the potential cross-regional sales risk, the number of boxes that may be added after the remaining distributor inventory data flows into the market is estimated, which is taken as the expected number of boxes flowing into the market in violation of regulations.
9. The cross-channel online default sales early warning method according to claim 8, characterized in that, Based on the expected number of non-compliant containers, the data is marked on the multi-dimensional visualization dashboard, specifically including: Determine whether the expected number of non-compliant containers has reached a preset container number threshold; If the threshold number of boxes is reached, it will be marked with the first color in the overall risk overview of the assessed provinces and cities, and the dealer position of the corresponding region will be highlighted in the dealer and terminal risk analysis quadrant chart. At the same time, obtain price fluctuation data for the region, including recent average market price, number of low-price links, and frequency of promotional activities; If the price fluctuation data in the region meets the following conditions: the number of low-priced links exceeds the preset low-priced link threshold and the expected number of inflow illegal boxes exceeds the box number threshold, then a price warning enhancement note will be added to the multi-dimensional visualization dashboard. The percentage of defaults in the region and the expected number of boxes flowing into the region that violate regulations are dynamically updated at fixed intervals, and the above-mentioned marking process is re-executed.
10. A cross-channel online default sales early warning device, used to implement the cross-channel online default sales early warning method as described in any one of claims 1 to 9, characterized in that, The device includes: The data acquisition module is used to obtain regional warehouse entry order data of each distributor based on the BI public business direct connection database. The regional warehouse entry order data includes product warehouse entry data, box number and bottle number associated with box number of each distributor. Each bottle number is associated with corresponding sales data, including sales time, sales region, sales price, consumer discount reward record and transfer record. The risk assessment module is used to extract all sales regions corresponding to the same box number based on the sales data of each bottle number, count the number of provinces and cities covered by the sales region, calculate the single box risk coefficient corresponding to the box number based on the number of provinces and cities, and calculate the customer risk coefficient of the distributor based on the single box risk coefficient and the proportion of risk boxes associated with the distributor. The control module is used to preset the risk threshold of the single-box risk coefficient. When the single-box risk coefficient reaches the risk threshold, an early warning is triggered. Based on the customer risk coefficient, a violation risk level is established for each distributor, and a coordinated governance procedure is executed based on the violation risk level. When establishing the violation risk level, the module obtains the transfer records and consumer discount reward records corresponding to the box number. If one or both of these records exist, a note is added to the violation risk level. If neither records exist, the violation risk level remains unchanged. When an early warning is triggered, the module sorts the warning information according to the violation risk level and generates a multi-dimensional visualization dashboard based on the sales region associated with the bottle number. Based on the data associated with the multi-dimensional visualization dashboard, the module obtains the total number of boxes sold, the number of boxes with violations, and the remaining distributor inventory data for the current time period. The module calculates the ratio of the number of boxes with violations to the total number of boxes sold to obtain the regional default ratio. Based on the regional default ratio and the remaining distributor inventory data, the module determines the expected number of boxes with violations flowing in and marks them on the multi-dimensional visualization dashboard according to the expected number of boxes with violations flowing in.