Intelligent visit plan generation method and device with dynamic adjustment of weight coefficient
Through the intelligent visit plan generation method that dynamically adjusts the weight coefficient, the problem of lack of universality and dynamic tuning of visit plan generation methods in the existing technology is solved, and a more accurate and efficient visit plan generation is achieved, which is suitable for multiple industries and segmentation Scene.
Patent Information
- Application Number
- CN202411119310.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-08-15
AI Technical Summary
The method of visit plan generation in the prior art lacks the universality of multiple industries and segmented scenarios. The algorithm is only calculated based on a single or a few factors, lacks dynamic tuning capabilities, and cannot adapt to changes in the actual visit plan implementation.
Using an intelligent visit plan generation method that dynamically adjusts the weight coefficient, collects user configuration and customer information, calculates the visit priority weight value, forms a secondary weight composition mode based on multiple weight factors, and dynamically adjusts it according to historical execution. Weight coefficient.
The generated visit plan is more accurate, conforms to actual application scenarios, improves customer visit efficiency, is suitable for multiple industries and segmented scenarios, and is scalable and flexible.
Smart Images

Figure CN118644222B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of computer big data processing, and in particular to a method and device for generating an intelligent visit plan with dynamically adjusted weight coefficients. Background Art
[0002] In the prior art, the technology for generating customer visit plans to improve efficiency includes: Chinese patent with publication number CN115081956A, which discloses an intelligent customer visit method and system based on building two-dimensional map, and the method includes the following steps: S1. When the account manager needs to visit the customer, first obtain the list of customers who need to be visited in the current task cycle; S2. The customer information is passed to the scoring calculation engine, and the scoring calculation engine calculates the customer's score according to the account manager position factor, customer position factor, customer level factor and customer industry factor, sorts the customers, and recommends them to the account manager in turn; S3. The system will add the visit data of the account manager to the calculation pool, and the weight calculation engine will also store the newly added visit data in the calculation pool; S4. After the account manager completes the visit, the visit data will be transmitted back for the intelligent customer visit system to correct the calculation factor and recommend the next best customer for the account manager to visit, which can effectively improve the visit efficiency of the account manager and reduce travel consumption.
[0003] A Chinese patent with publication number CN117557059A discloses a business allocation method based on information fusion: including collecting customer order history information and customer business information to establish customer portraits; collecting marketing personnel's data information and corresponding marketing performance information to establish marketing personnel portraits; establishing a matching model between customer portraits and marketing personnel portraits; and based on the matching model, generating a visit sequence for each marketing personnel and each customer to be processed, so as to achieve the purpose of assigning visiting tasks to marketing personnel.
[0004] It can be seen that the characteristics of the prior art are mainly:
[0005] 1. Sales personnel manually create temporary visit plans, which are applicable to all industries and all customer types and have extremely high flexibility. However, the effectiveness of plan creation depends entirely on the experience of sales personnel.
[0006] 2. Cycle planning generates visit plans regularly, taking only the visit cycle into consideration.
[0007] 3. The visit route only considers two factors: customer visit cycle and customer distance to generate a plan. Although the further upgraded solution, intelligent routing, supports the generation of plans based on multiple factors, it only applies to a single industry and scenario and is not universal.
[0008] The defects of the visit plan generation method under the existing technology:
[0009] 1. It is only applicable to a single industry or scenario and is not applicable to multiple industries and segmented scenarios.
[0010] 2. The algorithm is calculated independently based on only one or a few factors such as customer level, customer positioning, customer visit cycle, customer opinions, etc., without comprehensively considering multiple factors together.
[0011] 3. The algorithm is globally fixed and lacks scalability, and cannot be dynamically adjusted according to the actual execution of the enterprise's visit plan.
[0012] The present invention can effectively improve the corresponding deficiencies. Summary of the invention
[0013] The present invention proposes the following technical solutions to address one or more technical deficiencies in the above-mentioned prior art.
[0014] A method for generating an intelligent visit plan with dynamically adjusted weight coefficients, the method comprising:
[0015] The collection step is to obtain a customer list customer_list to be visited according to the visit plan generation instruction sent by the user, and the customer list customer_list includes at least one customer to be visited, and obtain the user's personal visit configuration person_config{vst,vet,ihv,vsl}, wherein vst represents the daily visit start time, vet represents the daily visit end time, ihv represents whether to automatically create a visit plan on holidays, and vsl represents the location information of the departure point of the visit;
[0016] A calculation step, calculating a visit priority weight value of each customer to be visited in the list customer_list based on the user's departure location information, the type of the customer to be visited, the location information of the customer to be visited and the visit weight factor vector, and arranging the customers to be visited in the list customer_list in descending order based on the visit priority weight value;
[0017] A generation step, generating a list of visit plans to be executed based on vst, vet, departure location information, average visit time of customers to be visited and location information of customers to be visited;
[0018] The adjustment step is to adjust each weight vector in the visit weight factor vector based on the visit plan executed by the user in a historical period.
[0019] Furthermore, in the calculation step, the corresponding visit weight factor vector visit_factor_config is obtained based on the type of the customer to be visited. ;
[0020] Calculate the service weight value wvb, the formula is as follows:
[0021] ;
[0022] The distance weight value wvd is calculated based on the user's departure location information and the location information of the customer to be visited, and the formula is as follows:
[0023]
[0024] Based on wvb and wvd, the visit priority weight value of the customer to be visited is calculated. The formula is as follows:
[0025]
[0026] in, Indicates the time interval coefficient from the last visit to the present, with a value range of [0,1]; Indicates the time coefficient of the last order transaction, the value range is [0,1]; Indicates the coefficient of the last order transaction amount, with a value range of [0,1]; Indicates the customer level coefficient, with a value range of [0,1]; Represents the customer opinion coefficient, with a value range of [0,1]; Indicates the business weight value wvb coefficient, the value range is [0,1]; It represents the positioning distance weight wvd coefficient, with a value range of [0,1]; lvt represents the time interval from the last visit to the present, lot represents the time interval from the last order transaction to the present, loa represents the total amount of the last order transaction, cl represents the customer level, pol represents the customer opinion information level, and d represents the distance between the user and the customer to be visited, which is calculated based on the positioning information of the user's departure place and the location information of the customer to be visited.
[0027] Furthermore, in the collection step, the collected data is cleaned and stored in a unified data structure.
[0028] Furthermore, in the adjustment step, the user queries the list of visit plans to be executed, and the employee chooses whether to execute it according to the actual situation. If so, the business data generated during the visit will be collected and cleaned by the information collection system as incremental data and updated and saved, and the total number of visit plans successfully executed in a historical period svpt will be accumulated. If the user cancels the visit plan, the cancellation reason will be selected, and the cancellation reason will be mapped to the corresponding coefficient in visitit_factor_config, and the total number of unexecuted visit plans in a historical period will be accumulated. ;
[0029] The dynamic tuning task is performed according to n days as a historical period. The tuning formula is:
[0030]
[0031] Among them, sr represents the success rate of visit plan execution in a historical period, ,
[0032] Indicates the value of a single coefficient in a historical period. The calculated visit cancellation rate, ;
[0033] Represents the total number of visit plans in a historical period, i=1, 2, 3, 4, 5, 6 or 7.
[0034] Furthermore, the visit priority weight value wvc is modified according to the customer's custom expansion factor, and the formula is:
[0035]
[0036] Indicates the corrected visit priority weight value, Indicates the weight coefficient of the custom expansion factor, Indicates a custom expansion factor.
[0037] The present invention also proposes an intelligent visit plan generation device for dynamically adjusting weight coefficients, the device comprising:
[0038] The collection unit generates a visit plan instruction sent by the user, obtains a customer list customer_list to be visited, the customer list customer_list includes at least one customer to be visited, and obtains the user's personal visit configuration person_config{vst,vet,ihv,vsl}, wherein vst represents the daily visit start time, vet represents the daily visit end time, ihv represents whether to automatically create a visit plan on holidays, and vsl represents the location information of the departure place of the visit;
[0039] a calculation unit, which calculates a visit priority weight value of each customer to be visited in the list customer_list based on the departure location information of the user, the type of the customer to be visited, the location information of the customer to be visited and the visit weight factor vector, and arranges the customers to be visited in the list customer_list in descending order based on the visit priority weight value;
[0040] A generating unit generates a list of visiting plans to be executed based on vst, vet, departure location information, average visiting time of customers to be visited and location information of customers to be visited;
[0041] An adjustment unit adjusts each weight vector in the visit weight factor vector based on a visit plan executed by the user in a historical period.
[0042] Furthermore, in the calculation unit, a corresponding visit weight factor vector visit_factor_config is obtained based on the type of the customer to be visited. ;
[0043] Calculate the service weight value wvb, the formula is as follows:
[0044] ;
[0045] The distance weight value wvd is calculated based on the user's departure location information and the location information of the customer to be visited, and the formula is as follows:
[0046]
[0047] Based on wvb and wvd, the visit priority weight value of the customer to be visited is calculated. The formula is as follows:
[0048]
[0049] in, Indicates the time interval coefficient from the last visit to the present, with a value range of [0,1]; Indicates the time coefficient of the last order transaction, the value range is [0,1]; Indicates the coefficient of the last order transaction amount, with a value range of [0,1]; Indicates the customer level coefficient, with a value range of [0,1]; Represents the customer opinion coefficient, with a value range of [0,1]; Indicates the business weight value wvb coefficient, the value range is [0,1]; It represents the positioning distance weight wvd coefficient, with a value range of [0,1]; lvt represents the time interval from the last visit to the present, lot represents the time interval from the last order transaction to the present, loa represents the total amount of the last order transaction, cl represents the customer level, pol represents the customer opinion information level, d represents the distance between the user and the customer to be visited, which is calculated based on the positioning information of the user's departure place and the location information of the customer to be visited.
[0050] Furthermore, in the acquisition unit, the acquired data is cleaned and stored in a unified data structure.
[0051] Furthermore, in the adjustment unit, the user queries the list of visit plans to be executed, and the employee chooses whether to execute it according to the actual situation. If so, the business data generated during the visit will be collected and cleaned by the information collection system as incremental data and updated and saved, and the total number of visit plans successfully executed in a historical period svpt will be accumulated. If the user cancels the visit plan, the cancellation reason will be selected, and the cancellation reason will be mapped to the corresponding coefficient in visitit_factor_config, and the total number of unexecuted visit plans in a historical period will be accumulated. ;
[0052] The dynamic tuning task is performed according to n days as a historical period. The tuning formula is:
[0053]
[0054] Among them, sr represents the success rate of visit plan execution in a historical period, ,
[0055] Indicates the value of a single coefficient in a historical period. The calculated visit cancellation rate, ;
[0056] Represents the total number of visit plans in a historical period, i=1, 2, 3, 4, 5, 6 or 7.
[0057] Furthermore, the visit priority weight value wvc is modified according to the customer's custom expansion factor, and the formula is:
[0058]
[0059] Indicates the corrected visit priority weight value, Indicates the weight coefficient of the custom expansion factor, Indicates a custom expansion factor.
[0060] The technical effect of the present invention is: the present invention provides a method, device and storage medium for generating an intelligent visit plan with dynamically adjusted weight coefficients, the method comprising: a collection step S101, according to a visit plan generation instruction sent by a user, obtaining a list of customers to be visited customer_list, the list customer_list including at least one customer to be visited, obtaining the user's personal visit configuration person_config{vst,vet,ihv,vsl}, wherein vst represents the daily visit start time, vet represents the daily visit end time, ihv represents whether a visit plan is automatically created on holidays, and vsl represents the location information of the departure point of the visit; a calculation step S102, S102, based on the departure location information of the user, the type of the customer to be visited, the location information of the customer to be visited and the visit weight factor vector, calculate the visit priority weight value of each customer to be visited in the list customer_list, and arrange the customers to be visited in the list customer_list in descending order based on the visit priority weight value; generation step S103, based on vst, vet, the departure location information, the average visit time of the customer to be visited and the location information of the customer to be visited; adjustment step S104, adjust each weight vector in the visit weight factor vector based on the visit plan executed by the user in a historical period. The intelligent visit plan generation method of the present invention adopts multiple weight factors to participate in the calculation, and obtains corresponding weights based on the type of customers to be visited, forming a two-level weight composition mode, which is convenient for flexible adjustment of weights, and the weight coefficients are dynamically optimized, so that the generated customer visiting method is more accurate and in line with the actual application scenario. In the present invention, the corresponding coefficient configuration vector is selected according to the customer type, thereby forming a two-level weight configuration mode, so that the weight vector is more in line with the actual situation of the customer, and the generated visit plan is more in line with objective needs. On the basis of the above, the present invention proposes a specific formula for calculating the visit priority weight value of the customer to be visited. After actual use, the generated visit plan meets the actual needs and can improve the customer's visit efficiency. The present invention can also be based on the number of customers' devices. Make corrections so that the corrected The value is more in line with the customer's actual situation to generate an accurate visit plan. BRIEF DESCRIPTION OF THE DRAWINGS
[0061] Other features, objects and advantages of the present application will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings.
[0062] Figure 1 It is a flowchart of a method for generating an intelligent visit plan with dynamically adjusted weight coefficients according to an embodiment of the present invention.
[0063] Figure 2 It is a structural diagram of an intelligent visit plan generation device that dynamically adjusts weight coefficients according to an embodiment of the present invention. DETAILED DESCRIPTION
[0064] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It should also be noted that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings.
[0065] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0066] Figure 1 The invention shows an intelligent visit plan generation method for dynamically adjusting weight coefficients, the method comprising:
[0067] Acquisition step S101, according to the visit plan generation instruction sent by the user, obtain the customer list customer_list to be visited, the list customer_list includes at least one customer to be visited, and obtain the user's personal visit configuration person_config{vst,vet,ihv,vsl}, where vst represents the daily visit start time, vet represents the daily visit end time, ihv represents whether to automatically create a visit plan on holidays, and vsl represents the location information of the departure point of the visit;
[0068] Calculation step S102, calculating the visit priority weight value of each customer to be visited in the list customer_list based on the departure location information of the user, the type of the customer to be visited, the location information of the customer to be visited and the visit weight factor vector, and arranging the customers to be visited in the list customer_list in descending order based on the visit priority weight value;
[0069] Generating step S103, generating a list of visiting plans to be executed based on vst, vet, departure location information, average visiting time of the customers to be visited and location information of the customers to be visited;
[0070] In the adjustment step S104, each weight vector in the visit weight factor vector is adjusted based on the visit plan executed by the user in a historical period.
[0071] The intelligent visit plan generation method of the present invention adopts multiple weight factors to participate in the calculation, and obtains corresponding weights based on the type of customers to be visited, forming a two-level weight composition pattern, which is convenient for flexible adjustment of weights and dynamic optimization of weight coefficients, so that the generated customer visiting method is more accurate and in line with actual application scenarios. This is one of the important inventive concepts of the present invention.
[0072] In one embodiment, in the calculation step S102, the corresponding visit weight factor vector visitit_factor_config is obtained based on the type of the customer to be visited. ;
[0073] Calculate the service weight value wvb, the formula is as follows:
[0074] ;
[0075] The distance weight value wvd is calculated based on the user's departure location information and the location information of the customer to be visited, and the formula is as follows:
[0076]
[0077] Based on wvb and wvd, the visit priority weight value of the customer to be visited is calculated. The formula is as follows:
[0078]
[0079] in, Indicates the time interval coefficient from the last visit to the present, with a value range of [0,1]; Indicates the time coefficient of the last order transaction, the value range is [0,1]; Indicates the coefficient of the last order transaction amount, with a value range of [0,1]; Indicates the customer level coefficient, with a value range of [0,1]; Represents the customer opinion coefficient, with a value range of [0,1]; Indicates the business weight value wvb coefficient, the value range is [0,1]; It represents the positioning distance weight wvd coefficient, with a value range of [0,1]; lvt represents the time interval from the last visit to the present, lot represents the time interval from the last order transaction to the present, loa represents the total amount of the last order transaction, cl represents the customer level, pol represents the customer opinion information level, d represents the distance between the user and the customer to be visited, which is calculated based on the positioning information of the user's departure place and the location information of the customer to be visited.
[0080] visitit_factor_config is the visit plan weight factor coefficient vector, which is used to bind the relationship and refine it into the secondary customer classification. In the same enterprise, employees with different roles may visit different customers. If an enterprise only uses one coefficient configuration, the accuracy will be greatly affected in scenarios with large differences in customers.
[0081] Taking brand manufacturers in the fast-moving consumer goods industry as an example, the company's customers are mainly classified into two types: terminal stores and channel distributors. The corresponding field staff role for store visits is the sales representative, and the corresponding role for channel distributor visits is the regional manager.
[0082] The number of terminal stores is two to three orders of magnitude larger than that of channel distributors. Moreover, terminal stores are divided into convenience stores, small supermarkets and large supermarkets according to customer levels. The visiting cycles of stores of different levels are also different, including weekly visits, biweekly visits and four-weekly visits. The corresponding last visit interval coefficient is The coefficient of time since the last order transaction Relatively large. However, there is very little opinion information from terminal stores, so the opinion coefficient Small. Sales representatives are usually responsible for visiting all terminal stores in an area, and the average number of terminal stores visited per day can reach more than 20. Route planning has a great impact on the efficiency of visits, so the positioning weight coefficient Larger, business weight coefficient Smaller.
[0083] The coefficient configuration of the terminal store is vistit_factor_config. A template configuration example is {0.7, 0.7, 0.5, 0.5, 0.1, 0.5, 1}
[0084] The dealer visit cycle is uncertain and relatively long. Usually, the dealer needs to visit when the brand needs to promote new policies or new cooperation. Therefore, the customer level coefficient Larger, customer opinion coefficient Larger, the last visit interval coefficient The coefficient of time since the last order transaction Relatively small. Dealers are usually scattered. Dealers of large companies may be distributed all over the country. Different dealers are relatively far away, so the positioning weight coefficient Smaller, overall business weight coefficient Larger.
[0085] A template configuration example of the channel distributor's coefficient configuration vistit_factor_config is {0.3, 0.3, 0.5, 0.9, 0.7, 1, 0.1}.
[0086] The present invention also maintains a data structure visit_customer_data, which contains 6 key parameter values {lvt, lot, loa, a, cl, pol}, which are specifically defined as follows:
[0087] lvt: Abbreviation of last_visit_time, which indicates the time interval from the last visit to the present, with the unit accurate to hours.
[0088] lot:last_order_time is the abbreviation of lot:last_order_time, which indicates the time interval from the last order transaction to the present, with the unit accurate to the hour.
[0089] loa: Abbreviation of last_order_amount, which indicates the total amount of the last order transaction, with the unit accurate to RMB.
[0090] a: Abbreviation of address, indicating the customer address location information, including longitude and latitude information (latitude, longitude).
[0091] cl: abbreviation of customer_level, indicating customer level, integer.
[0092] pol: abbreviation of public_opinion_level, indicating the level of customer opinion information, an integer.
[0093] Each parameter of visit_customer_data has a preset mapping relationship with the field value of customer_data_set in the data warehouse.
[0094] You can use the intelligent visit plan generator to generate a visit plan, and use the weight calculation engine to calculate the visit priority weight value. The intelligent visit plan generator sorts the customer_list according to the results returned by the weight calculation engine, combines the average visit time and the in-transit time intransit_time, creates the corresponding visit plan, and reports the total number of visit plans vpt created within the cumulative period (n days) to the weight calculation engine. Each time a visit plan is created, the remaining visit time for the day is calculated. When the time is insufficient, the remaining customer visit plans will no longer be created.
[0095] In the present invention, the corresponding coefficient configuration vector visitit_factor_config is selected according to the customer type {frist_type, second_type}, thereby forming a two-level weight configuration mode, so that the weight vector is more in line with the actual situation of the customer, and the generated visit plan is more in line with objective needs. On the basis of the above, the present invention proposes a specific formula for calculating the visit priority weight value of the customer to be visited. After actual use, the generated visit plan meets actual needs and can improve the customer's visit efficiency. This is another important invention point of the present invention.
[0096] In one embodiment, in the collection step S101, the collected data is cleaned and then stored in a unified data structure.
[0097] In one embodiment, in the adjustment step S104, the user queries the list of visit plans to be executed, and the employee chooses whether to execute according to the actual situation. If yes, the business data generated during the visit will be collected, cleaned and updated by the information collection system as incremental data, and the total number of visit plans successfully executed in a historical period svpt will be accumulated. If the user cancels the visit plan, the cancellation reason will be selected, and the cancellation reason will be mapped to the corresponding coefficient in visitit_factor_config, and the total number of unexecuted visit plans in a historical period will be accumulated. ;
[0098] The dynamic tuning task is performed according to n days as a historical period. The tuning formula is:
[0099]
[0100] Among them, sr represents the success rate of visit plan execution in a historical period, ,
[0101] Indicates the value of a single coefficient in a historical period. The calculated visit cancellation rate, ;
[0102] Represents the total number of visit plans in a historical period, i=1, 2, 3, 4, 5, 6 or 7.
[0103] Arranged:
[0104]
[0105] In the present invention, due to the weight calculation coefficient The greater the cancellation rate caused by The more it needs to be reduced, the more the tuning range caused by a single coefficient will be diluted through the overall visit plan execution success rate sr. The higher the success rate, the smaller the adjustment range of a single coefficient each time, thereby more objectively adjusting the size of each weight factor, making the subsequently generated visit plan more objective. This is another important invention point of the present invention.
[0106] In one embodiment, a custom expansion factor is defined. For example, preferably, the number of devices of the customer is defined as the custom expansion factor, and then the visit priority weight value wvc is modified according to the number of devices of the customer. The formula is:
[0107]
[0108] Indicates the corrected visit priority weight value, The weight coefficient representing the number of devices of the customer, Indicates the number of devices of the customer.
[0109] For example, for beverage brand manufacturers, the number of terminal store equipment (refrigerators, freezers) is an important indicator. The business can customize a device number device_count. Sales representatives can report the number of devices when visiting stores. Administrators can The formula adds a correction formula as above, so that the corrected The value is more in line with the actual situation of the customer to generate an accurate visit plan, which is another important inventive concept of the present invention.
[0110] In the present invention, and ext can be expanded to multiple groups. In order to prevent the preset parameters of the algorithm formula from being diluted, causing large fluctuations in the calculation results, generally speaking, it is recommended to have a maximum of 3 groups for the same algorithm formula And ext, of course, can be extended more.
[0111] Figure 2 The invention shows an intelligent visit plan generation device for dynamically adjusting weight coefficients, the device comprising:
[0112] The collection unit 201 obtains a customer list customer_list to be visited according to a visit plan generation instruction sent by the user, the customer list customer_list includes at least one customer to be visited, and obtains the user's personal visit configuration person_config{vst,vet,ihv,vsl}, wherein vst represents the daily visit start time, vet represents the daily visit end time, ihv represents whether to automatically create a visit plan on holidays, and vsl represents the location information of the departure place of the visit;
[0113] The calculation unit 202 calculates the visit priority weight value of each customer to be visited in the list customer_list based on the departure location information of the user, the type of the customer to be visited, the location information of the customer to be visited and the visit weight factor vector, and arranges the customers to be visited in the list customer_list in descending order based on the visit priority weight value;
[0114] A generating unit 203 generates a list of visiting plans to be executed based on vst, vet, the location information of the departure place, the average visiting time of the customers to be visited and the location information of the customers to be visited;
[0115] The adjusting unit 204 adjusts each weight vector in the visit weight factor vector based on the visit plan executed by the user in a historical period.
[0116] The intelligent visit plan generation method of the present invention adopts multiple weight factors to participate in the calculation, and obtains corresponding weights based on the type of customers to be visited, forming a two-level weight composition pattern, which is convenient for flexible adjustment of weights and dynamic optimization of weight coefficients, so that the generated customer visiting method is more accurate and in line with actual application scenarios. This is one of the important inventive concepts of the present invention.
[0117] In one embodiment, in the calculation unit 202, a corresponding visit weight factor vector visitit_factor_config is obtained based on the type of the customer to be visited. ;
[0118] Calculate the service weight value wvb, the formula is as follows:
[0119] ;
[0120] The distance weight value wvd is calculated based on the user's departure location information and the location information of the customer to be visited, and the formula is as follows:
[0121]
[0122] Based on wvb and wvd, the visit priority weight value of the customer to be visited is calculated. The formula is as follows:
[0123]
[0124] in, Indicates the time interval coefficient from the last visit to the present, with a value range of [0,1]; Indicates the time coefficient of the last order transaction, the value range is [0,1]; Indicates the coefficient of the last order transaction amount, with a value range of [0,1]; Indicates the customer level coefficient, with a value range of [0,1]; Represents the customer opinion coefficient, with a value range of [0,1]; Indicates the business weight value wvb coefficient, the value range is [0,1]; It represents the positioning distance weight wvd coefficient, with a value range of [0,1]; lvt represents the time interval from the last visit to the present, lot represents the time interval from the last order transaction to the present, loa represents the total amount of the last order transaction, cl represents the customer level, pol represents the customer opinion information level, d represents the distance between the user and the customer to be visited, which is calculated based on the positioning information of the user's departure place and the location information of the customer to be visited.
[0125] visitit_factor_config is the visit plan weight factor coefficient vector, which is used to bind the relationship and refine it into the secondary customer classification. In the same enterprise, employees with different roles may visit different customers. If an enterprise only uses one coefficient configuration, the accuracy will be greatly affected in scenarios with large differences in customers.
[0126] Taking brand manufacturers in the fast-moving consumer goods industry as an example, the company's customers are mainly classified into two types: terminal stores and channel distributors. The corresponding field staff role for store visits is the sales representative, and the corresponding role for channel distributor visits is the regional manager.
[0127] The number of terminal stores is two to three orders of magnitude larger than that of channel distributors. Moreover, terminal stores are divided into convenience stores, small supermarkets and large supermarkets according to customer levels. The visiting cycles of stores of different levels are also different, including weekly visits, biweekly visits and four-weekly visits. The corresponding last visit interval coefficient is The coefficient of time since the last order transaction Relatively large. However, there is very little opinion information from terminal stores, so the opinion coefficient Small. Sales representatives are usually responsible for visiting all terminal stores in an area, and the average number of terminal stores visited per day can reach more than 20. Route planning has a great impact on the efficiency of visits, so the positioning weight coefficient Larger, business weight coefficient Smaller.
[0128] The coefficient configuration of the terminal store is vistit_factor_config. A template configuration example is {0.7, 0.7, 0.5, 0.5, 0.1, 0.5, 1}
[0129] The dealer visit cycle is uncertain and relatively long. Usually, the dealer needs to visit when the brand needs to promote new policies or new cooperation. Therefore, the customer level coefficient Larger, customer opinion coefficient Larger, the last visit interval coefficient The coefficient of time since the last order transaction Relatively small. Dealers are usually scattered. Dealers of large companies may be distributed all over the country. Different dealers are relatively far away, so the positioning weight coefficient Smaller, overall business weight coefficient Larger.
[0130] A template configuration example of the channel distributor's coefficient configuration vistit_factor_config is {0.3, 0.3, 0.5, 0.9, 0.7, 1, 0.1}.
[0131] The present invention also maintains a data structure visit_customer_data, which contains 6 key parameter values {lvt, lot, loa, a, cl, pol}, which are specifically defined as follows:
[0132] lvt: Abbreviation of last_visit_time, which indicates the time interval from the last visit to the present, with the unit accurate to hours.
[0133] lot:last_order_time is the abbreviation of lot:last_order_time, which indicates the time interval from the last order transaction to the present, with the unit accurate to the hour.
[0134] loa: Abbreviation of last_order_amount, which indicates the total amount of the last order transaction, with the unit accurate to RMB.
[0135] a: Abbreviation of address, indicating the customer address location information, including longitude and latitude information (latitude, longitude).
[0136] cl: abbreviation of customer_level, indicating customer level, integer.
[0137] pol: abbreviation of public_opinion_level, indicating the level of customer opinion information, an integer.
[0138] Each parameter of visit_customer_data has a preset mapping relationship with the field value of customer_data_set in the data warehouse.
[0139] You can use the intelligent visit plan generator to generate a visit plan, and use the weight calculation engine to calculate the visit priority weight value. The intelligent visit plan generator sorts the customer_list according to the results returned by the weight calculation engine, combines the average visit time and the in-transit time intransit_time, creates the corresponding visit plan, and reports the total number of visit plans vpt created within the cumulative period (n days) to the weight calculation engine. Each time a visit plan is created, the remaining visit time for the day is calculated. When the time is insufficient, the remaining customer visit plans will no longer be created.
[0140] In the present invention, the corresponding coefficient configuration vector visitit_factor_config is selected according to the customer type {frist_type, second_type}, thereby forming a two-level weight configuration mode, so that the weight vector is more in line with the actual situation of the customer, and the generated visit plan is more in line with objective needs. On the basis of the above, the present invention proposes a specific formula for calculating the visit priority weight value of the customer to be visited. After actual use, the generated visit plan meets actual needs and can improve the customer's visit efficiency. This is another important invention point of the present invention.
[0141] In one embodiment, in the acquisition unit 201, the acquired data is cleaned and then stored in a unified data structure.
[0142] In one embodiment, in the adjustment unit 204, the user queries the list of visit plans to be executed, and the employee chooses whether to execute according to the actual situation. If so, the business data generated during the visit will be collected, cleaned and updated by the information collection system as incremental data, and the total number of visit plans successfully executed in a historical period svpt will be accumulated. If the user cancels the visit plan, the cancellation reason will be selected, and the cancellation reason will be mapped to the corresponding coefficient in visitit_factor_config, and the total number of unexecuted visit plans in a historical period will be accumulated. ;
[0143] The dynamic tuning task is performed according to n days as a historical period. The tuning formula is:
[0144]
[0145] Among them, sr represents the success rate of visit plan execution in a historical period, ,
[0146] Indicates the value of a single coefficient in a historical period. The calculated visit cancellation rate, ;
[0147] Represents the total number of visit plans in a historical period, i=1, 2, 3, 4, 5, 6 or 7.
[0148] Arranged:
[0149]
[0150] In the present invention, due to the weight calculation coefficient The greater the cancellation rate caused by The more it needs to be reduced, the more the tuning range caused by a single coefficient will be diluted through the overall visit plan execution success rate sr. The higher the success rate, the smaller the adjustment range of a single coefficient each time, thereby more objectively adjusting the size of each weight factor, making the subsequently generated visit plan more objective. This is another important invention point of the present invention.
[0151] In one embodiment, a custom expansion factor is defined, for example, the number of devices of the customer is defined as the custom expansion factor, and then the visit priority weight value wvc is modified according to the number of devices of the customer, and the formula is:
[0152]
[0153] Indicates the corrected visit priority weight value, The weight coefficient representing the number of devices, Indicates the number of devices of the customer.
[0154] For example, for beverage brand manufacturers, the number of terminal store equipment (refrigerators, freezers) is an important indicator. The business can customize a device number device_count. Sales representatives can report the number of devices in the store when visiting the store. Administrators can The formula adds a correction formula as above, so that the corrected The value is more in line with the actual situation of the customer to generate an accurate visit plan, which is another important inventive concept of the present invention.
[0155] In the present invention, and ext can be expanded to multiple groups. In order to prevent the preset parameters of the algorithm formula from being diluted, causing large fluctuations in the calculation results, generally speaking, it is recommended to have a maximum of 3 groups for the same algorithm formula And ext, of course, can be extended more.
[0156] Compared with the prior art, the present invention has outstanding features and advantages:
[0157] For example, the formula based on the building two-dimensional map in the Chinese patent with publication number CN115081956A is limited to concentrated customer visits in urban buildings; the formula of the Chinese patent with publication number CN117557059A is mainly based on the customer's business information matrix, without considering factors such as customer location and visit time. The weight calculation formula of the present invention adds the order and customer opinion dimensions to the default weight factor, and the formula can be expanded to support enterprises to customize weight factors according to actual conditions (for example, beverage companies can customize weight factors: equipment quantity, including freezers and refrigerators, etc.), which is more versatile and suitable for visit plan weight calculation in various industries and scenarios.
[0158] The tuning logic of the present invention is also different from the prior art. The formula coefficient tuning of the Chinese patent with publication number CN115081956A is based on the information of the actual execution of the visit (only positive feedback data is used, and the situation where the visit plan is not executed is not considered), while the coefficient dynamic tuning method of the present invention covers both positive and negative feedback. It not only uses the information of the actual execution of the visit (including: visit success rate, last visit time, etc.), but also collects the reasons why the visit plan was not executed for each coefficient, so as to be able to perform comprehensive tuning. The coefficients of the Chinese patent with publication number CN117557059A do not have a dynamic tuning mechanism, and its business scenario is to assign a group of customers to a group of salesmen, while the scenario of the present invention is to calculate the optimal visit sequence for a group of customers that each salesman is responsible for.
[0159] In one embodiment of the present invention, a computer storage medium is provided, on which a computer program is stored. When the computer program on the computer storage medium is executed by a processor, the above method is implemented. The computer storage medium can be a hard disk, DVD, CD, flash memory or other memory.
[0160] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing the present application, the functions of each unit can be implemented in the same or multiple software and / or hardware.
[0161] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which can be stored in a storage medium, such as ROM / RAM, a disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the devices described in each embodiment of the present application or some parts of the embodiments.
[0162] Finally, it should be noted that the above embodiments are only intended to illustrate rather than limit the technical solutions of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the present invention can still be modified or replaced by equivalents. Any modification or partial replacement that does not depart from the spirit and scope of the present invention should be included in the scope of the claims of the present invention.
Claims
1. A method for generating intelligent visit plans with dynamic adjustment of weight coefficients. It is characterized in that The method includes: The collection step is to obtain a customer list customer_list to be visited according to the visit plan generation instruction sent by the user, and the customer list customer_list includes at least one customer to be visited, and obtain the user's personal visit configuration person_config{vst,vet,ihv,vsl}, wherein vst represents the daily visit start time, vet represents the daily visit end time, ihv represents whether to automatically create a visit plan on holidays, and vsl represents the location information of the departure point of the visit; A calculation step, calculating a visit priority weight value of each customer to be visited in the list customer_list based on the user's departure location information, the type of the customer to be visited, the location information of the customer to be visited and the visit weight factor vector, and arranging the customers to be visited in the list customer_list in descending order based on the visit priority weight value; A generation step, generating a list of visit plans to be executed based on vst, vet, departure location information, average visit time of customers to be visited and location information of customers to be visited; An adjustment step of adjusting each weight vector in the visit weight factor vector based on the visit plan executed by the user in a historical period; Wherein, in the calculation step, the corresponding visit weight factor vector visit_factor_config is obtained based on the type of the customer to be visited; Calculate the service weight value wvb, the formula is as follows: ; The distance weight value wvd is calculated based on the user's departure location information and the location information of the customer to be visited, and the formula is as follows: Based on wvb and wvd, the visit priority weight value of the customer to be visited is calculated. The formula is as follows: in, Indicates the time interval coefficient from the last visit to the present, with a value range of [0,1]; Indicates the time coefficient of the last order transaction, the value range is [0,1]; Indicates the coefficient of the last order transaction amount, with a value range of [0,1]; Indicates the customer level coefficient, with a value range of [0,1]; Represents the customer opinion coefficient, with a value range of [0,1]; Indicates the business weight value wvb coefficient, the value range is [0,1]; represents the location distance weight wvd coefficient, with a value range of [0,1]; lvt represents the time interval from the last visit to the present, lot represents the time interval from the last order transaction to the present, loa represents the total amount of the last order transaction, cl represents the customer level, pol represents the customer opinion information level, and d represents the distance between the user and the customer to be visited, which is calculated based on the location information of the user's departure place and the location information of the customer to be visited; Among them, in the adjustment step, the user queries the list of visit plans to be executed, and the employee chooses whether to execute according to the actual situation. If so, the business data generated during the visit will be collected, cleaned and updated by the information collection system as incremental data, and the total number of visit plans successfully executed in a historical period svpt will be accumulated. If the user cancels the visit plan, the cancellation reason will be selected, and the cancellation reason will be mapped to the corresponding coefficient in visitit_factor_config, and the total number of unexecuted visit plans in a historical period will be accumulated. ; The dynamic tuning task is performed according to n days as a historical period. The tuning formula is: Among them, sr represents the success rate of visit plan execution in a historical period, , Indicates the value of a single coefficient in a historical period. The calculated visit cancellation rate, ; represents the total number of visit plans in a historical period, i = 1, 2, 3, 4, 5, 6 or 7; The visit priority weight value wvc is modified according to the customer's custom expansion factor, and the formula is: ; Indicates the corrected visit priority weight value, Indicates the weight coefficient of the custom expansion factor, Indicates a custom expansion factor, so that the modified The value is more in line with the customer's actual situation to generate an accurate visit plan.
2. The method according to claim 1, It is characterized in that In the collection step, the collected data is cleaned and then stored in a unified data structure.
3. An intelligent visit plan generation device with dynamically adjusted weight coefficients, It is characterized in that The device includes: The collection unit generates a visit plan instruction sent by the user, obtains a customer list customer_list to be visited, the customer list customer_list includes at least one customer to be visited, and obtains the user's personal visit configuration person_config{vst,vet,ihv,vsl}, wherein vst represents the daily visit start time, vet represents the daily visit end time, ihv represents whether to automatically create a visit plan on holidays, and vsl represents the location information of the departure place of the visit; a calculation unit, which calculates a visit priority weight value of each customer to be visited in the list customer_list based on the departure location information of the user, the type of the customer to be visited, the location information of the customer to be visited and the visit weight factor vector, and arranges the customers to be visited in the list customer_list in descending order based on the visit priority weight value; A generating unit generates a list of visiting plans to be executed based on vst, vet, departure location information, average visiting time of customers to be visited and location information of customers to be visited; an adjusting unit, adjusting each weight vector in the visit weight factor vector based on the visit plan executed by the user in a historical period; Wherein, in the calculation unit, the corresponding visit weight factor vector visit_factor_config is obtained based on the type of the customer to be visited. ; Calculate the service weight value wvb, the formula is as follows: ; The distance weight value wvd is calculated based on the user's departure location information and the location information of the customer to be visited, and the formula is as follows: Based on wvb and wvd, the visit priority weight value of the customer to be visited is calculated. The formula is as follows: in, Indicates the time interval coefficient from the last visit to the present, with a value range of [0,1]; Indicates the time coefficient of the last order transaction, the value range is [0,1]; Indicates the coefficient of the last order transaction amount, with a value range of [0,1]; Indicates the customer level coefficient, with a value range of [0,1]; Represents the customer opinion coefficient, with a value range of [0,1]; Indicates the business weight value wvb coefficient, the value range is [0,1]; represents the location distance weight wvd coefficient, with a value range of [0,1]; lvt represents the time interval from the last visit to the present, lot represents the time interval from the last order transaction to the present, loa represents the total amount of the last order transaction, cl represents the customer level, pol represents the customer opinion information level, and d represents the distance between the user and the customer to be visited, which is calculated based on the location information of the user's departure place and the location information of the customer to be visited; Among them, in the adjustment unit, the user queries the list of visit plans to be executed, and the employee chooses whether to execute according to the actual situation. If so, the business data generated during the visit will be collected and cleaned by the information collection system as incremental data and updated and saved, and the total number of visit plans successfully executed in a historical period svpt will be accumulated. If the user cancels the visit plan, the cancellation reason will be selected, and the cancellation reason will be mapped to the corresponding coefficient in visitit_factor_config, and the total number of unexecuted visit plans in a historical period will be accumulated. ; The dynamic tuning task is performed according to n days as a historical period. The tuning formula is: Among them, sr represents the success rate of visit plan execution in a historical period, , Indicates the value of a single coefficient in a historical period. The calculated visit cancellation rate, ; represents the total number of visit plans in a historical period, i = 1, 2, 3, 4, 5, 6 or 7; The visit priority weight value wvc is modified according to the customer's custom expansion factor, and the formula is: ; Indicates the corrected visit priority weight value, Indicates the weight coefficient of the custom expansion factor, Indicates a custom expansion factor, so that the modified The value is more in line with the customer's actual situation to generate an accurate visit plan.
4. The device according to claim 3, It is characterized in that In the acquisition unit, the acquired data is cleaned and then stored in a unified data structure.
