CN-121998263-A - Business data analysis method and system for medicine marketing system

CN121998263ACN 121998263 ACN121998263 ACN 121998263ACN-121998263-A

Abstract

The invention relates to a business data analysis method and a business data analysis system for a medicine marketing system, which are used for collecting multi-source data of four main bodies corresponding to organizations, personnel, hospitals and products in medicine marketing business, constructing a linkage data model for associating the four main body multi-source data, building a clear, comprehensive and accurate unified data system for realizing the on-line of core data, building a core analysis layer and a diagnosis calibration layer based on a large language model, realizing the serial analysis of the four-dimensional data under the multi-granularity of the organizations, the personnel and the clients, combining an output display layer to generate and display a report which is hierarchical and drillable, visually presenting the analysis of each main body of the organizations, the personnel, the hospitals and the products, replacing the traditional manual analysis mode, rapidly outputting the business optimization suggestion which can fall to the ground, objectively measuring the operation health degree of the power-assisted business, promoting the marketing management to change from rough to fine, and improving the management efficiency and decision scientificity.

Inventors

XIE CHENGRUN
Cao Fenze
LIANG JINZHU
ZHAO YAN

Assignees

正大天晴药业集团股份有限公司

Dates

Publication Date: 20260508
Application Date: 20260409

Claims (10)

1. A business data analysis method facing a medicine marketing system is characterized by comprising the following steps: Step A, based on preset systems in the medicine marketing business, acquiring multi-source data of four main bodies corresponding to organizations, personnel, hospitals and products respectively, wherein the multi-source data comprises performance data, cost data, distribution data and interaction data, and then entering step B; Step B, cleaning the collected multi-source data, constructing a linkage data model for associating the multi-source data of four main bodies based on four-level unique association keys corresponding to the four main bodies of the organization, personnel, hospital and product, and then entering the step C; Step C, based on multi-source associated data of multi-source data of four main bodies of an associated organization, personnel, a hospital and a product under a linkage data model, respectively aiming at the four main bodies of the organization, the personnel, the hospital and the product, applying a preset quantitative evaluation algorithm of the production health degree, quantitatively grading according to each dimensionality of performance data, cost data, distribution data and interaction data, quantitatively analyzing the production health degree aiming at the main bodies to obtain comprehensive grading of the production health degree of the main bodies, and then entering the step D; And D, comprehensively scoring the main body production health of four main bodies of organizations, personnel, hospitals and products, inputting multi-source associated data of four main body multi-source data under a linkage data model into a target large language model, driving the target large language model to perform intelligent diagnosis by combining with a structured prompt under the application of a preset medical scene, outputting a preliminary diagnosis conclusion comprising abnormality identification, excellent case extraction, comprehensive diagnosis and optimization suggestion, and checking the preliminary diagnosis conclusion through a preset multi-dimensional cross-validation mechanism to obtain a final full-dimensional production diagnosis conclusion.
2. The method of claim 1, wherein in the step A, a single sales representative, a single doctor, and a single product are taken as collection objects, real-time data collection is performed for multi-source data of a dynamic data type, batch data collection is performed for multi-source data of a static basic data type, and compliance verification is performed for the collected multi-source data according to marketing compliance requirements of the pharmaceutical industry.
3. The method of claim 1, wherein in the step A, each index of the multi-source data corresponding to the organization comprises office identification O-ID, performance goal and actual value, total cost, SL plan cost, cost rate, total number and grading of clients, doctor interaction rate, regional market capacity, performance ring ratio and comparability data; The indexes of the multi-source data corresponding to the personnel comprise sales representative P-ID, affiliated office identification O-ID, personal performance, SL plan expense and actual expenditure, doctor number and grading, interaction frequency type and object, target grading doctor coverage rate, visit compliance and examination period data; The indexes of the multi-source data corresponding to the hospital comprise hospital H-ID, belonging office identification O-ID, hospital grade, cooperation product Pr-ID, performance contribution, SL plan expense rate, doctor number distribution, interaction times and forms, participation coverage rate and department subdivision data; The indexes of the multi-source data corresponding to the product comprise Pr-ID of the product, the identification O-ID of the office, the achievement rate, the SL planning cost, the cost rate, the adaptation hospital grade and department, the interactive conversion data and the category attribute.
4. The method for analyzing business data for pharmaceutical marketing system according to claim 1, wherein the step B comprises the steps of B1 to B3; Step B1, orderly executing abnormal data regularization elimination, data standardization unification and missing value medical service association complementation aiming at the acquired multi-source data to realize cleaning and updating of the acquired multi-source data, and then entering step B2; Step B2, forming preset four-level unique association keys corresponding to four main bodies of an organization, personnel, a hospital and a product by using an office identification O-ID index corresponding to the organization, a sales representative P-ID index corresponding to the personnel, a hospital H-ID index corresponding to the hospital and a product Pr-ID index corresponding to the product, and establishing P-ID O-ID、H-ID O-ID、Pr-ID Inheritance relationship of O-ID and P-ID H-ID、P-ID Pr-ID、H-ID The association relation of Pr-ID is then entered into step B3; Step B3, linking a preset multisource data acquisition time dimension t, and establishing a linkage data model according to the following formula; M_{t,(O,P,H,Pr),k}=V_{t,(O,P,H,Pr),k}+Tag_{(O-ID,P-ID,H-ID,Pr-ID)}; Wherein O represents organization, P represents personnel, H represents products, and (O, P, H, pr) represents four main body association combinations of organization, personnel, hospitals and products, k represents k indexes under the four main body association combinations of organization, personnel, hospitals and products, tag_ { (O-ID, P-ID, H-ID, pr-ID) represents association labels of four-level unique association keys of the four main bodies of organization, personnel, hospitals and products, V_ { t, (O, P, H, pr), k } represents acquired data values of k indexes under the four main body association combinations of organization, personnel, hospitals and products corresponding to time dimension t, M_ { t, (O, P, H, pr), k } represents a linkage data matrix of k indexes under the four main body association combinations of organization, personnel, hospitals and products.
5. The business data analysis method for the pharmaceutical marketing system of claim 1, wherein the step C is based on multi-source associated data of multi-source data of four main bodies of an organization, a person, a hospital and a product under a linkage data model, and the multi-source associated data is respectively aimed at the four main bodies of the organization, the person, the hospital and the product according to the following formula: calculating to obtain the comprehensive score of the health degree of the main body production, wherein, Represent the first The main body production health degree comprehensive score of each main body, 、、、 Sequentially represent the first The preset weights corresponding to the performance data, the cost data, the distribution data and the interaction data under the individual main bodies, 、、、 Sequentially represent the first Quantized scores of performance data, cost data, assignment data, interaction data under individual subjects.
6. The business data analysis method for the pharmaceutical marketing system of claim 5, wherein the step C further comprises the steps of comprehensively grading the health degree of the production of the main body based on the main body of each main body of the organization, the personnel, the hospital and the product, and quantitatively grading the performance data, the cost data, the distribution data and the interaction data under each main body, and realizing the accurate positioning of the abnormal point and the reproducible extraction of the excellent case according to the following abnormal point identification rule and excellent case screening rule; the abnormal point identification rule and the excellent case screening rule comprise a preset main body production health comprehensive scoring abnormal threshold corresponding to an abnormal main body, a preset index abnormal threshold corresponding to each core index under multi-source associated data, a preset main body production health comprehensive scoring threshold corresponding to an excellent case and a target core index threshold under multi-source associated data.
7. The business data analysis method for the pharmaceutical marketing system of claim 1, wherein in the step D, the preliminary diagnosis conclusion is checked by a preset multi-dimensional cross-validation mechanism as follows to obtain a final full-dimensional production diagnosis conclusion; A quantization index verification mechanism for verifying whether the identified abnormal or excellent case in the preliminary diagnosis conclusion accords with a preset threshold quantization judgment rule, if so, retaining the identified abnormal or excellent case in the preliminary diagnosis conclusion, otherwise, refusing the corresponding identified abnormal or excellent case in the preliminary diagnosis conclusion; the main body association verification mechanism is used for verifying whether the abnormal root analyzed in the preliminary diagnosis conclusion is consistent with the standard quantitative data of the associated main body based on the linkage data model, if so, the abnormal root analyzed in the preliminary diagnosis conclusion is reserved, and otherwise, the abnormal root analyzed in the preliminary diagnosis conclusion is refused; And the medical service logic verification mechanism is used for verifying whether the optimization suggestions in the preliminary diagnosis conclusion accord with service reality or not based on medical marketing service rules and industry compliance requirements, if so, retaining the optimization suggestions in the preliminary diagnosis conclusion, and otherwise, rejecting the optimization suggestions in the preliminary diagnosis conclusion.
8. The business data analysis method for medicine marketing system of claim 1, further comprising the step E, after the step D is executed, entering the step E; And E, generating and displaying a hierarchical and drillable report based on the linkage data model and the final full-dimension production diagnosis conclusion, wherein the report supports the detail respectively drilled from an organization level to a personnel level, a hospital level and a product level, and visually presents the comprehensive score of the main body production health of each main body of the organization, personnel, the hospital and the product, and each core index, the identified abnormality and the optimization suggestion under the multi-source associated data.
9. The system for realizing the business data analysis method for the medicine marketing system is characterized by comprising a data acquisition layer, a data processing layer, a core analysis layer and a diagnosis calibration layer, wherein the data acquisition layer is used for executing the step A, acquiring multi-source data of four main bodies of a preset system in a medicine marketing business, which correspond to organizations, personnel, hospitals and products respectively, the data processing layer is used for executing the step B, constructing a linkage data model for associating the multi-source data of the four main bodies, the core analysis layer is used for executing the step C, obtaining comprehensive scores of the health of the main bodies of the organizations, the personnel, the hospitals and the products respectively, the diagnosis calibration layer is used for executing the step D, executing intelligent diagnosis by applying a target large language model, outputting a preliminary diagnosis conclusion, and checking through a preset multi-dimensional cross validation mechanism to obtain a final full-dimensional diagnosis conclusion.
10. The system for realizing the business data analysis method for the pharmaceutical marketing system of claim 9, further comprising an output display layer for generating and displaying a hierarchical, drillable report based on the linkage data model and the final full-dimension production diagnosis conclusion, wherein the report supports the detail of drilling down from an organization level to a personnel level, a hospital level and a product level respectively, and visually presents the comprehensive scores of the production health of the organization, personnel, the hospital and the main bodies of the products, and the core indexes, the identified anomalies and the optimization suggestions under the multi-source associated data.

Description

Business data analysis method and system for medicine marketing system Technical Field The invention relates to a business data analysis method and system for a medicine marketing system, and belongs to the technical field of intelligent analysis of medicine marketing data. Background The core competitiveness of medicine marketing is that 'resource accurate release' and 'input-output efficiency', along with the trend of strict supervision and aggravation of market competition in the medicine industry, enterprises are urgently required to optimize marketing resource allocation through data driving, and the yield rate of return is improved. The medical marketing business relates to a multi-level main body, wherein an organization level (office) needs to balance the overall performance and the cost, a personnel level (sales representative) needs to improve the personal resource utilization efficiency, a hospital level (cooperation client) needs to realize the deep operation of high-value clients, and a product level needs to monitor the performance and the cost investment rationality of different products. The performance, cost, distribution and interaction data generated by each subject are the core basis for evaluating the production efficiency. The traditional medicine marketing analysis mode relies on manual arrangement of scattered data, only can realize the basic presentation of 'how much to perform and how much to charge', and can not answer the core questions such as 'whether the charge input is reasonable', 'whether the interaction behavior is effective', 'why part of personnel/hospitals are prominent', and the like. For example, abnormal persons or hospitals who cannot quickly identify "high cost low performance" are also difficult to replicate the excellent experience of "low cost high growth", resulting in waste of resources and inefficiency of management. With the popularization of large data technology in the application of the pharmaceutical industry, the enterprise increasingly urgent demands on 'full-dimension, accurate and floor-type' production analysis, the prior art such as a sales data statistics system and a general marketing analysis tool which are commonly used in the pharmaceutical industry, has the core scheme that the system comprises a pharmaceutical sales data statistics system and a general marketing analysis tool, wherein the pharmaceutical sales data statistics system is mainly connected with a sales management system and a cost reimbursement system, collects performance and cost basic data of organizations and personnel and client distribution data, is used for providing basic statistical forms such as performance achievement rate, cost occupation ratio, regional ranking and the like, supports single-dimension data screening and export, has no multi-main data association analysis capability, only presents data results and does not comprise abnormal identification, cause analysis and optimization suggestions; The general marketing analysis tool supports structured data uploading and integration, but does not design a special data interface aiming at a multi-dimensional data architecture in the aspect of medicine marketing, has poor suitability, is used for providing general planning cost rate calculation and trend analysis functions, but lacks customized models such as 'interactive behavior and performance association', 'classified customer production analysis' and the like aiming at medicine scenes, can only carry out basic visual charts (bar charts and line charts), cannot realize multi-level drilling (such as from tissue drilling to personnel and hospitals), and has weak decision support capability, so the technical schemes provide basic data statistics tools for medicine marketing, are limited by data integration depth and analysis model universality, cannot realize 'full-dimensional production diagnosis' and 'floor decision support', and are difficult to meet the fine marketing management requirements of medicine enterprises. The prior art thus currently has several disadvantages: 1. The data integration dimension is insufficient, and the prior art only covers part of main body data and does not realize the full dimension association of organizations, personnel, hospitals and products. The causal link is that the medicine marketing production effect is influenced by multi-main body linkage (such as personnel interaction behavior directly influences hospital performance, product characteristics determine hospital suitability), and data fragmentation leads to failure in establishing 'input-behavior-output' complete logic- & gt analysis conclusion one-sided, failure in accurately positioning production problem root cause- & gt resource optimization lacks data support, so that input is wasted. 2. The analysis model lacks of medical customization, and the prior art adopts a general analysis model and is not suitable for the characteristics of multiple medicine marketing subject