A personalized product selection method and system for auto parts supply chain
By building a car accessories sku candidate pool and evaluating characteristic values based on historical data, a personalized product selection list is generated, which solves the problem that medium and large automobile enterprises cannot select targeted products, and achieves the accuracy of product selection and the improvement of market response speed. , meets the market demand in different regions.
Patent Information
- Application Number
- CN202111519166.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-10
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2041-12-10
AI Technical Summary
The forward warehouses distributed across the country by medium and large automobile companies cannot select appropriate auto parts categories in a targeted manner, and cannot respond quickly to market changes, resulting in difficulty in evaluating the rationality of product selection plans, unable to meet the market demand in different regions, and the cost benefits are the greatest. It is difficult to achieve.
By constructing a candidate pool for automotive accessories sku, evaluating the characteristic values of each sku based on historical data, dividing sales scope, weighted calculations, generating a personalized product selection list, automating the product selection process, and reducing manual intervention.
It realizes the accuracy and personalization of product selection, improves market response speed, reduces the cost of manual product selection, ensures a closed business loop, and meets market demand in different regions.
Smart Images

Figure CN114912968B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of auto parts selection, and in particular to a personalized product selection method and system for an auto parts supply chain. Background Art
[0002] There are many kinds of spare parts in China's auto aftermarket. In 2020, the national car ownership reached 281 million, and the average number of parts per car reached about 10,000, and the overall scale of spare parts is huge. Moreover, the current situation of "international cars" in China is that there are many types of models, and there are large differences in accessories between different models. There is no domestic auto aftermarket company that can supply all categories of auto parts. In addition, auto aftermarket companies provide accessories to local repair shops through forward warehouses distributed across the country. The distribution of car models in different regions varies greatly, and market demands are different. Therefore, how each forward warehouse of the auto aftermarket company can select the appropriate business category to meet market demand as much as possible and maximize cost-benefit is a very challenging problem for any auto aftermarket company.
[0003] Generating product selection plans for the automotive aftermarket forward warehouses based on industry experience can meet market demand to a certain extent, but due to manual limitations, there are the following difficult-to-solve problems: it is impossible to make targeted product selection plans based on the current situation of the region where the forward warehouse is located. For small-scale enterprises, different product selection strategies can be made by increasing the number of employees, but for medium and large automotive aftermarket companies with forward warehouses across the country, it is impossible to manually make targeted product selection plans for each forward warehouse; the sales cycle of automotive aftermarket spare parts is long, and the sales frequency of a single sku is low. Manual labor cannot respond quickly to market changes for each sku and update product selection plans in a timely manner; manual product selection also has no good evaluation plan for the rationality of the product selection plan, and a business closed loop cannot be achieved. Summary of the invention
[0004] In view of the shortcomings of the above problems, the present invention provides a method and system for personalized product selection in an auto parts supply chain, the method comprising:
[0005] Build a candidate pool of auto parts sku and determine the categories of auto parts;
[0006] Evaluate each SKU in the candidate pool according to historical data to obtain an evaluation value of each SKU;
[0007] All the SKUs in each of the categories are grouped and sorted from large to small according to the evaluation values to generate a product selection list;
[0008] Determine the number of SKUs under each of the categories, select them from the product selection list in order of arrangement, and finally generate a personalized product selection list.
[0009] Preferably, building the auto parts sku candidate pool includes:
[0010] Construct a basic candidate pool of auto parts, which includes all SKUs sold by other forward warehouses, SKUs sold by Kongsou, and all SKUs sold by Tmall car maintenance stores;
[0011] The basic candidate pool of automobile parts is filtered to obtain the final candidate pool of automobile parts sku.
[0012] Preferably, filtering the basic candidate pool of automobile parts to obtain the final candidate pool of automobile parts sku includes:
[0013] Filter out the SKUs of brands that do not meet business requirements;
[0014] Filter the SKUs of the bottom 20% of the vehicle ownership ranking;
[0015] Filter SKUs with abnormal purchasing status;
[0016] Filter the skus that are added to the blacklist;
[0017] Filter the SKUs whose empty searches on the sales platform are less than the pre-set threshold.
[0018] Preferably, evaluating each SKU in the candidate pool according to historical data, and obtaining the evaluation value of each SKU includes:
[0019] According to the sales area, it is divided into large area, small area, city and forward warehouse;
[0020] Calculate the feature values of the same sku in different ranges respectively;
[0021] The feature values of the same SKU in different ranges are normalized and weighted to obtain the evaluation value of each SKU.
[0022] Preferably, respectively calculating the feature values of the same sku in different ranges includes:
[0023] Perform weighted summation according to the evaluation indicators to obtain the feature values of the same SKU in different ranges;
[0024] The evaluation indicators include the number of empty searches within the scope, the sales volume of Tmall car maintenance stores, and the sales volume of forward warehouses.
[0025] Preferably, the calculation formula for the evaluation value of each SKU is:
[0026] Y=0.4*(0.4A 1 +0.4A 2 +0.4A 3 +A4 +A 5 )+0.3*(0.4B 1 +0.4B 2 +0.4B 3 +B 4 +B 5 )+0.2*
[0027] (0.4C 1 +0.4C 2 +0.4C 3 +C 4 +C 5 )+0.1*(0.4D 1 +0.4D 2 +0.4D 3 +D 4 +D 5 );
[0028] Where: Y is the evaluation value; A 1 A is the number of empty search PVs in the forward warehouse range; 2 A is the empty search frequency within the forward warehouse range; 3 A is the sales volume of Tmall car maintenance stores within the forward warehouse area; 4 A is the sales frequency of the forward warehouse within the forward warehouse range; 5 B is the sales volume of the forward warehouse within the forward warehouse range; 1 is the number of PVs of air search in the city; B 2 is the frequency of empty searches within the city; B 3 B is the sales volume of Tmall auto maintenance stores in the city; 4 is the sales frequency of the forward warehouse in the city; B 5 is the sales volume of the forward warehouse in the city; C 1 is the number of empty search PVs in the cell area; C 2 is the number of empty search frequencies within the cell range; C 3 is the sales volume of Tmall car maintenance stores in the community; C 4 is the sales frequency of the forward warehouse in the community; C 5 is the sales volume of the forward warehouse in the community; D 1 D is the number of empty search PVs in the region; 2 is the number of empty search frequencies in the large area; D 3 The sales volume of Tmall auto maintenance stores in the region; D 4 is the sales frequency of the forward warehouse in the region; D 5 The sales volume of forward warehouses in the larger region.
[0029] Preferably, all the SKUs in each of the categories are grouped and sorted from large to small according to the evaluation values, and generating a selection list includes:
[0030] For the categories in which there is a wildcard relationship between SKUs and they are of the same brand and price range, a threshold is set and the quantity within the threshold is selected, and the rest are deleted.
[0031] The present invention also provides a system for the auto parts supply chain personalized product selection method, comprising:
[0032] A construction module is used to construct a candidate pool of auto parts SKUs and determine the categories of the auto parts;
[0033] An evaluation module, used to evaluate each SKU in the candidate pool according to historical data to obtain an evaluation value of each SKU;
[0034] A sorting module, used to group and sort all the SKUs in each of the categories according to the evaluation values from large to small, and generate a product selection list;
[0035] The generation module is used to determine the number of skus under each of the categories, select them from the product selection list in order of arrangement, and finally generate a personalized product selection list.
[0036] Compared with the prior art, the present invention has the following beneficial effects:
[0037] The present invention divides the sales scope into different ranges, calculates the characteristic values of the same sku in different ranges, and weights them to obtain the evaluation value of each sku. The products are sorted according to the size of the evaluation values to make the product selection more accurate and no manual work is required. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] Figure 1 The present invention provides a flow chart of a personalized product selection method for the auto parts supply chain. DETAILED DESCRIPTION
[0039] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0040] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.
[0041] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0042] In the present invention, empty search is defined as a sku searched on the sales platform, and the sku is out of stock.
[0043] The following is combined with Figure 1 The present invention is described in further detail:
[0044] like Figure 1 As shown, the present invention provides a method and system for personalized product selection in an auto parts supply chain, the method comprising:
[0045] Build a candidate pool of auto parts sku and determine the categories of auto parts;
[0046] Specifically, building an auto parts sku candidate pool includes:
[0047] Build a basic candidate pool for auto parts, which includes all SKUs sold by other forward warehouses, SKUs sold by Kongsou, and all SKUs sold by Tmall car maintenance stores;
[0048] The basic candidate pool of auto parts is filtered to obtain the final auto parts sku candidate pool.
[0049] Furthermore, the basic candidate pool of automobile parts is filtered to obtain a final candidate pool of automobile parts including:
[0050] Filter out the SKUs of brands that do not meet business requirements;
[0051] Filter the SKUs of the bottom 20% of the vehicle ownership ranking;
[0052] Filter SKUs with abnormal purchasing status;
[0053] Filter the skus that are added to the blacklist;
[0054] Filter the SKUs whose empty searches on the sales platform are less than the pre-set threshold.
[0055] Evaluate each SKU in the candidate pool based on historical data to obtain the evaluation value of each SKU;
[0056] Specifically, the sales area is divided into large regions, small regions, cities and forward warehouses from large to small. Large regions are composed of one or more provinces, and small regions are composed of one or more cities.
[0057] Calculate the feature values of the same sku in different ranges respectively;
[0058] Normalize and weight the feature values of the same SKU in different ranges to obtain the evaluation value of each SKU.
[0059] Furthermore, the characteristic values of the same sku in different ranges are calculated separately, including:
[0060] Perform weighted summation according to the evaluation indicators to obtain the feature values of the same SKU in different ranges;
[0061] Evaluation indicators include the number of empty searches within the scope, sales of Tmall car maintenance stores, and sales of forward warehouses.
[0062] The calculation formula for the evaluation value of each sku is:
[0063] Y=0.4*(0.4A 1 +0.4A 2 +0.4A 3 +A 4 +A 5 )+0.3*(0.4B 1 +0.4B 2 +0.4B 3 +B 4 +B 5 )+0.2*
[0064] (0.4C 1 +0.4C 2 +0.4C 3 +C 4 +C 5 )+0.1*(0.4D 1 +0.4D 2 +0.4D 3 +D 4 +D 5 );
[0065] Where: Y is the evaluation value; A 1 A is the number of empty search PVs in the forward warehouse range; 2 A is the empty search frequency within the forward warehouse range; 3 A is the sales volume of Tmall car maintenance stores within the forward warehouse area; 4 A is the sales frequency of the forward warehouse within the forward warehouse range; 5 B is the sales volume of the forward warehouse within the forward warehouse range; 1 is the number of PVs of air search in the city; B 2 is the frequency of empty searches within the city; B 3 B is the sales volume of Tmall auto maintenance stores in the city; 4 is the sales frequency of the forward warehouse in the city; B 5 is the sales volume of the forward warehouse in the city; C 1 is the number of empty search PVs in the cell area; C 2 is the number of empty search frequencies within the cell range; C 3 is the sales volume of Tmall car maintenance stores in the community; C 4 is the sales frequency of the forward warehouse in the community; C 5 is the sales volume of the forward warehouse in the community; D 1 D is the number of empty search PVs in the region; 2 is the number of empty search frequencies in the large area; D 3 The sales volume of Tmall auto maintenance stores in the region; D 4 is the sales frequency of the forward warehouse in the region; D 5 The sales volume of forward warehouses in the larger region.
[0066] All SKUs in each category are grouped and sorted from large to small according to the evaluation value to generate a product selection list;
[0067] Specifically, for categories with wildcard relationships between SKUs and the same brand and price range, a threshold is set and the quantity within the threshold is selected, and the rest are deleted.
[0068] Determine the number of skus under each category, select them from the product selection list in order of arrangement, and finally generate a personalized product selection list.
[0069] Example 1: Selection of products from old stores
[0070] In this embodiment, a brand within the community to which the store belongs is selected, and 95 categories for product selection are defined.
[0071] Building the SKU candidate pool for old auto parts stores includes:
[0072] On the basis of the original old store, add SKUs that are currently not available but sold by other forward warehouses and Tmall car maintenance stores, as well as SKUs searched on the sales platform, and the number of times the SKU is out of stock is greater than the pre-set threshold;
[0073] According to the set regional brands of large areas, small communities, and cities, as well as the third-, fourth-, and fifth-level category brands, brands that do not meet the rules will be filtered out and deleted;
[0074] Filter and delete the SKUs of the models that rank in the bottom 20% of the model ownership in the community;
[0075] Filter and delete those with abnormal purchasing status, i.e. those that cannot be purchased normally;
[0076] Filter and delete the SKUs in the blacklist created by all store owners in the community.
[0077] Evaluate each SKU in the candidate pool based on historical data to obtain the evaluation value of each SKU;
[0078] Sort all SKUs in the category by evaluation value from large to small to generate a product selection list;
[0079] Specifically, for categories with wildcard relationships between SKUs and the same brand and price range, a threshold is set and the quantity within the threshold is selected, and the rest are deleted.
[0080] Determine the number of skus under each category, select them from the product selection list in order of arrangement, and finally generate a personalized product selection list.
[0081] In this embodiment, the number of skus under each category is determined by proportionally expanding the number of skus under each category on the original list, and the number of skus under each category is subject to certain restrictions. In this embodiment, the maximum expansion is 1.5 times, and the overall number of skus is not higher than 10,000.
[0082] Example 2: New Store Product Selection
[0083] In this embodiment, brands in the community to which the store belongs are selected, and the initial stocking categories of the selected products are limited to 26. The initial stocking list of the new store needs to meet the minimum sku richness of the forward warehouse, and it is necessary to continuously iterate and explore customer needs.
[0084] Building a new store auto parts sku candidate pool includes:
[0085] Determine whether there is a forward warehouse in the city where the new store is located. If so, count the sku sales in the city; if not, count the sku sales in the community;
[0086] According to the set regional brands of large areas, small communities, and cities, as well as the third-, fourth-, and fifth-level category brands, brands that do not meet the rules will be filtered out and deleted;
[0087] Filter and delete the SKUs of the models that rank in the bottom 20% of the model ownership in the community;
[0088] Filter and delete those with abnormal purchasing status, i.e. those that cannot be purchased normally;
[0089] Filter and delete the SKUs in the blacklist created by all store owners in the community.
[0090] The number of SKU types is limited under the overall and category conditions to obtain the final personalized product list for the new store. The overall number of SKU types is less than 5,000, and the number of single categories is less than 1,000. In the case of an unknown market, limiting the number of SKU types can better control costs and continuously improve the product list through gradual iteration.
[0091] The present invention also provides a system for a personalized product selection method for an auto parts supply chain, comprising:
[0092] A construction module is used to construct a candidate pool of auto parts SKUs and determine the categories of the auto parts;
[0093] An evaluation module, used to evaluate each SKU in the candidate pool according to historical data to obtain an evaluation value of each SKU;
[0094] A sorting module, used to group and sort all the SKUs in each of the categories according to the evaluation values from large to small, and generate a product selection list;
[0095] The generation module is used to determine the number of skus under each of the categories, select them from the product selection list in order of arrangement, and finally generate a personalized product selection list.
[0096] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A personalized product selection method for auto parts supply chain, It is characterized in that include: Build a candidate pool of auto parts sku and determine the categories of auto parts; Evaluate each SKU in the candidate pool according to historical data to obtain an evaluation value of each SKU; All the SKUs in each of the categories are grouped and sorted from large to small according to the evaluation values to generate a product selection list; Determine the number of SKUs under each of the categories, select them from the product selection list in order of arrangement, and finally generate a personalized product selection list; Building an auto parts sku candidate pool includes: Construct a basic candidate pool of auto parts, which includes all SKUs sold by the forward warehouse, the SKUs searched by the empty warehouse, and all SKUs sold by Tmall car maintenance stores. The SKUs searched by the empty warehouse are out of stock. Filtering the basic candidate pool of automobile parts to obtain a final candidate pool of automobile parts sku; Evaluating each SKU in the candidate pool according to historical data to obtain an evaluation value of each SKU includes: According to the sales area, it is divided into large area, small area, city and forward warehouse; Calculate the feature values of the same sku in different ranges respectively; Normalize and weight the feature values of the same SKU in different ranges to obtain the evaluation value of each SKU; Calculating the feature values of the same sku in different ranges includes: Perform weighted summation according to the evaluation indicators to obtain the feature values of the same SKU in different ranges; The evaluation indicators include the number of empty searches within the scope, the sales volume of Tmall car maintenance stores, and the sales volume of forward warehouses; The calculation formula for the evaluation value of each of the sku is: Y=0.4*(0.4A 1 +0.4A 2 +0.4A 3 +A 4 +A 5 )+0.3*(0.4B 1 +0.4B 2 +0.4B 3 +B 4 +B 5 )+0.2* (0.4C 1 +0.4C 2 +0.4C 3 +C 4 +C 5 )+0.1*(0.4D 1 +0.4D 2 +0.4D 3 +D 4 +D 5 ); Where: Y is the evaluation value; A 1 A is the number of empty search PVs in the forward warehouse range; 2 A is the empty search frequency within the forward warehouse range; 3 A is the sales volume of Tmall car maintenance stores within the forward warehouse area; 4 A is the sales frequency of the forward warehouse within the forward warehouse range; 5 B is the sales volume of the forward warehouse within the forward warehouse range; 1 is the number of PVs of air search in the city; B 2 is the frequency of empty searches within the city; B 3 B is the sales volume of Tmall auto maintenance stores in the city; 4 is the sales frequency of the forward warehouse in the city; B 5 is the sales volume of the forward warehouse in the city; C 1 is the number of empty search PVs in the cell area; C 2 is the number of empty search frequencies within the cell range; C 3 is the sales volume of Tmall car maintenance stores in the community; C 4 is the sales frequency of the forward warehouse in the community; C 5 is the sales volume of the forward warehouse in the community; D 1 D is the number of empty search PVs in the region; 2 is the number of empty search frequencies in the large area; D 3 The sales volume of Tmall auto maintenance stores in the region; D 4 is the sales frequency of the forward warehouse in the region; D 5 The sales volume of forward warehouses in the larger region.
2. The personalized product selection method for the auto parts supply chain according to claim 1, It is characterized in that The final auto parts sku candidate pool obtained by filtering the auto parts basic candidate pool includes: Filter out the SKUs of brands that do not meet business requirements; Filter the SKUs of the bottom 20% of the vehicle ownership ranking; Filter SKUs with abnormal purchasing status; Filter the skus that are added to the blacklist; Filter the SKUs whose empty searches on the sales platform are less than the pre-set threshold.
3. The personalized product selection method for the auto parts supply chain according to claim 2, It is characterized in that All the SKUs in each of the categories are grouped and sorted from large to small according to the evaluation values, and a selection list is generated, including: For the categories in which there is a wildcard relationship between SKUs and they are of the same brand and price range, a threshold is set and the quantity within the threshold is selected, and the rest are deleted.
4. A system for providing a personalized product selection method for an auto parts supply chain as claimed in any one of claims 1 to 3, It is characterized in that include: A construction module is used to construct a candidate pool of auto parts SKUs and determine the categories of the auto parts; An evaluation module, used to evaluate each SKU in the candidate pool according to historical data to obtain an evaluation value of each SKU; A sorting module, used to group and sort all the SKUs in each of the categories according to the evaluation values from large to small, and generate a product selection list; The generation module is used to determine the number of skus under each of the categories, select them from the product selection list in order of arrangement, and finally generate a personalized product selection list.
Citation Information
Patent Citations
Method and a system for promoting product selection
CN109801102A
Commodity selection method and system and computer readable storage medium
CN112613953A
