Pharmaceutical sales assistance information processing device and method

Abstract

[Problem] To enable assistance of efficient marketing activities and sales promotion activities by a company such as a pharmaceutical company, thereby reducing marketing cost and optimizing the cost of medicines and the like to improve a social situation in which savable lives are not saved. [Solution] Provided is a device, a program, or a method comprising a pharmaceutical sales assistance information processing technique that includes: an acquisition unit that acquires data including a history or schedule of sales promotion activities including an attribute for each medical doctor, a mail magazine transmission to the medical doctor, an opening notification of the mail magazine, a web interview, or a face-to-face interview; a storage unit that stores a program and the data; a calculation unit that calls the program and the data, and calculates use or prescription probability by using a statistical method including logistic regression analysis, or calculates a recommended sales and marketing activity by a logit function using the use or prescription probability; and an output unit that indicates the recommended sales and marketing activity calculated by the calculation unit.

Description

[Revised in accordance with Rule 26, 23.04.2025] Medical sales support information processing device and method

[0001] The present invention relates to a medical sales support information processing device and method.

[0002] Rising pharmaceutical prices are a serious social problem, especially in developed countries, resulting in potentially savable lives being wasted. One underlying factor behind this is the current trend in which pharmaceutical companies invest significantly more in marketing and sales promotion activities than in research and development (R&D). As a result, in the United States, marketing expenses drive up drug prices, limiting access to the latest treatments and creating a social problem. In Japan, where reimbursement prices are set by the government, the low reimbursement prices of pharmaceuticals make the Japanese market less attractive. As a result, 143 drugs approved in Europe and the United States remain unapproved in Japan, with 86 of these still in development as of March 2023. Meanwhile, some marketing spending is not reaching the appropriate target, resulting in waste. One major factor behind this waste is the inability to identify optimal sales targets for products from the user's perspective. To reduce the waste caused by inefficient marketing and sales promotion activities and curb rising drug prices, improving the accuracy of marketing activities and optimizing resource allocation is effective. In the past, pharmaceutical and medical device marketing activities were primarily based on qualitative perspectives, relying on experience and intuition. However, the external environment is changing rapidly, and decision-making that relies on experience and intuition is highly uncertain. Therefore, in recent years, Customer Relationship Management (CRM) systems, which apply information technology to marketing activities, have been proposed and are being used in some areas.

[0003] JP 2022-042617 A

[0004] However, the success rate of sales support systems actually implemented is low. It has even been reported that up to 70% of implementations end in failure. This is partly due to the weak theoretical models used in conventional sales support systems, particularly because they were not based on theoretical models based on statistical processing that clearly targeted individual physicians, who are the decision-makers behind pharmaceutical product adoption. The present invention was developed to address the above-mentioned issues and provides a sales support technology based on theoretical models based on statistical processing that utilizes the latest logistic regression analysis and logit functions, etc., that clearly targeted individual physicians, who are the decision-makers behind pharmaceutical product adoption. The present invention has multiple functions to achieve this objective.

[0005] According to the present invention, there is provided a medical sales support information processing device including: an acquisition unit that acquires data including the attributes of each doctor, a history or schedule of sales promotion activities including sending e-mail newsletters to the doctors, notifications of opening of e-mail newsletters, web interviews, or face-to-face interviews, or the results of the sales promotion activities; a memory unit that stores a program and the data; a calculation unit that calls up the program and the data and calculates the probability of use or prescription using a statistical method including logistic regression analysis, or calculates recommended sales activities using a logit function using the probability of use or prescription; and an output unit that shows the recommended sales activities calculated by the calculation unit.

[0006] This invention can support pharmaceutical companies in carrying out efficient marketing and sales promotion activities, thereby reducing marketing costs and optimizing pharmaceutical costs, thereby contributing to improving the social situation in which lives that could be saved are not being saved.

[0007] 1 is a conceptual diagram showing the configuration of the medical sales support information processing device 1 when the server 10, customer information terminals, etc. are connected via a network N. FIG. 1 is a block diagram showing the functional configuration of the server 10 that causes the medical sales support information processing device 1 to function. FIG. 2 is a diagram showing the data structure of a doctor table 1012 for initial value input. FIG. 3 is a diagram showing the data structure of a product table 1013 for initial value input. FIG. 4 is a diagram showing the data structure of an ROI analysis input table 1014 for initial value input. FIG. 5 is a diagram showing the data structure of an activity simulation table 1015. FIG. 6 is a diagram showing the data structure of an ROI simulation table 1016. FIG. 7 is a diagram showing the data structure of a daily sales report consulting table 1017. FIG. 8 is a flowchart showing the operation of the activity simulation. FIG. 9 is a flowchart showing the operation in the case of selecting doctors with high similarity. FIG. 10 is a flowchart showing the operation of the ROI simulation. FIG. 11 is a chart showing the operation of the daily sales report consulting. FIG. 12 is an example of an Excel A FORM for uploading initial values ​​for the activity simulation. FIG. 13 is a block diagram showing the basic hardware configuration of a general computer 90 that constitutes the information processing device.

[0008] Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all drawings illustrating the embodiments, common components are designated by the same reference numerals, and repeated explanations will be omitted. Note that the following embodiments do not unduly limit the content of the present invention described in the claims. Furthermore, not all components shown in the embodiments are necessarily essential components of the present invention. Furthermore, each drawing is a schematic diagram and is not necessarily a precise illustration.

[0009] <Configuration of the Medical Sales Support Information Processing Device> The following description will be made with reference to the drawings related to the information processing device. As shown in FIG. 1 , the medical sales support information processing device 1 of the present invention is composed of a connected server 10 and a management information processing terminal 15, and can also be connected to optional information processing terminals 20 and 30, such as customer information processing terminals. Some or all of these may be connected via a network N. While the network N is disclosed assuming the Internet or a mobile phone carrier line, it is not limited thereto and may be, for example, a dedicated line. Similarly, the server 10 is disclosed as being connected to the management information processing terminal 15 via the network N, but some or all of the information may be stored within the management information processing terminal 15.

[0010] <Configuration of Information Processing Apparatus> FIG. 2 is a block diagram showing the functional configuration of the server 10. As shown in FIG.

[0011] Each information processing device is configured by a computer equipped with a processing unit for control and calculation and a storage device. The basic hardware configuration of the computer and the basic functional configuration of the computer realized by the hardware configuration will be described later. For each of the server 10, the management information processing terminal 15, and the optional information processing terminals 20 and 30, etc., descriptions that overlap with the basic hardware configuration and basic functional configuration of the computer will be omitted.

[0012] <Configuration of Server 10> The server 10 is an information processing device that acquires, stores, and, if necessary, calculates and provides information related to the medical sales support information processing device 1. The server 10 includes a storage unit 101 and a control unit 104.

[0013] <Configuration of Storage Unit 101 of Server 10 > The storage unit 101 of the server 10 includes an application program 1011 , a doctor table 1012 , a product table 1013 , an ROI table 1014 , an activity simulation table 1015 , an ROI simulation table 1016 , and a daily sales report consulting table 1017 .

[0014] <Configuration of Control Unit 104 of Server 10 > The control unit 104 of the server 10 includes an acquisition unit 1041 , a calculation unit 1042 , and an output unit 1043 .

[0015] <Functions of medical sales support information processing device>

[0016] <Optimal Sales Action Calculation Process> First, the present invention predicts and identifies doctors who are likely to use or prescribe a target drug, etc., and evaluates that probability in order to improve the accuracy of targeting sales targets. This allows for identifying and matching similar target customers, enabling effective resource allocation. In other words, it is possible to identify doctors who are predicted to have a desired range of usage or prescription probability, such as a high probability, and present that probability.

[0017] <Overview of the Optimized Sales Action Calculation Process> The optimized sales action calculation process is a technology that utilizes statistical methods, particularly logistic regression analysis, to quantify the results of sales activities for each doctor and derive the amount of activity required to maximize sales effectiveness.

[0018] The inventors of the present invention discovered that multiple regression logistic regression analysis can be applied to calculate optimal sales actions. Specifically, by calculating the maximum probability when the regression coefficients related to physician attributes among the 21 regression coefficients used in multiple regression logistic regression analysis are fixed as a general rule and the remaining regression coefficients related to sales activities are freely settable, it is possible to derive the maximum probability of use or prescription for each physician and calculate the minimum amount of sales activity that maximizes the probability of use or prescription for each individual sales activity. The fixed regression coefficients related to physician attributes are the attributes of each physician, while the freely set regression coefficients related to sales activities are the details of the sales activity. Therefore, the present invention provides an efficient device and method for calculating optimal sales actions by utilizing a logistics function (probability transformation) and its inverse function, the logit function (linear combination). Specifically, an optimal sales activity model is constructed using the following procedure. 1. 1. Data collection - Collect activity history data for each doctor (e.g., number of visits, number of times information sessions attended, number of times materials were provided, etc.) - Collect attribute information on doctors (e.g., specialty, work facility, prescription tendencies, etc.) - Predict or collect results of sales activities (e.g., drug adoption / prescription status) 2. Analysis using statistical methods - Use the collected data to statistically analyze the relationship between each activity and outcome (usage or prescription), and calculate a coefficient that shows the correlation between the activity and outcome - Apply the logit function to calculate the probability of use or prescription for each doctor 3. Construction of sales optimization model (model for each doctor with a desired range of usage or adoption probability) - For doctors with a low probability of use or prescription (e.g., less than 50%), quantitatively calculate which sales activities need to be increased and by how much to improve sales - For doctors with a certain probability of use or prescription (e.g., 50% or more), it is also possible to generate proposals to reduce unnecessary sales activities 4. Visualization and output of results - Analysis results are output to a file and made available for viewing by sales and marketing personnel - Optimization strategies for each sales activity are presented as a report, and an actionable action plan is presented Details of the process will be explained later.

[0019] <Cosine Similarity Analysis> The cosine similarity analysis of the present invention is outlined in the patent publication issued by the inventor (Japanese Patent No. 7636838). However, the present invention particularly relates to a system and method for quantifying the similarity between medical professionals based on their attribute data (e.g., age, specialty, certification, work style, etc.) and outputting it visually or in data format. The present system converts the attribute data into a numerical format, standardizes continuous variables, and calculates cosine similarity on a scale of 0 to 100. This allows the similarity between each medical professional to be presented in a format that is intuitively interpretable for the user. Specifically, the present system has the following steps: 1. Obtain medical professional attribute data in Excel or a similar format. 2. Perform data preprocessing to convert the attribute data into a numerical format and standardize continuous variables. 3. Calculate cosine similarity and generate a similarity score between each medical professional on a scale of 0 to 100. 4. The calculation results are output as an Excel file or database and provided in a format that can be used for analysis purposes. More specifically, the following is performed: A. Data input and preprocessing: Attribute data for each healthcare professional is received as input. Attribute data includes age, qualifications, specialty, number of activities, etc. Continuous variables are standardized and scale differences are adjusted. B. Similarity calculation: Cosine similarity is calculated based on the quantified attribute data. The cosine similarity calculation results are converted to a 0-100 scale and provided as a score that can be intuitively interpreted. C. Results output: The calculation results are saved in an Excel file or database. Similarity scores between each healthcare professional are visualized and can be used to optimize educational programs and propose team compositions. This invention enables advanced analysis based on the characteristics of healthcare professionals, promoting more efficient resource management and the discovery of new interactions. Furthermore, 1. Attribute-based similarities between healthcare professionals can be efficiently analyzed. 2. 2. The similarity score can be intuitively interpreted, making it possible to utilize high levels of agreement between medical professionals with scores of 70% or higher. 3. Educational programs and resource allocation can be optimized based on the analysis results. 4. Large amounts of data can be processed in a short time, providing highly accurate results.

[0020] <ROI Analysis> According to the present invention, the ROI of each sales promotion activity can be analyzed, enabling optimal budget allocation. Until now, multiple events (campaigns) have been implemented within a certain period (e.g., one year), and even if the effects were apparent, it was not possible to clearly determine which event contributed to sales performance and to what extent. Therefore, it was necessary to continue both efficient and inefficient events.

[0021] <ROI Analysis Details> ROI is an abbreviation for "Return on Investment" and means "return on investment." ROI analysis is an analysis of return on investment, that is, it shows the ratio of the resulting effect, i.e., sales, to the funds invested in sales support. According to the present invention, ROI is an index that shows the degree to which sales increased as a result of the event (campaign) invested in compared to before and after the investment. The present invention can calculate the ROI for each campaign, making it easy to determine whether or not it is desirable to continue each individual campaign, thereby contributing to sales efficiency.

[0022] In the past, it was difficult to quantitatively evaluate the contribution of each campaign to sales before and after a certain period, especially when multiple campaigns were run simultaneously within that period, and there was a lack of methods for doing so. The inventors of the present invention have discovered a regression analysis model that fits ROI analysis, which quantitatively evaluates the contribution of each campaign to sales when multiple campaigns are run in the same fiscal year, and have invented a technology that uses the regression coefficients calculated by the regression analysis model to calculate the impact of each campaign on sales increase. In other words, the present invention provides an efficient method and system for performing regression analysis on sales increase data and campaign data, quantitatively evaluating the impact of each campaign, and calculating ROI.

[0023] Specifically, the present invention further calculates the ROI for each campaign based on sales data and campaign data in the following steps: 1. Obtaining the Increase in Sales: The increase in sales is calculated based on the annual sales performance for the target year and the sales performance for the previous year. If the difference in sales performance for individual physicians is significant, the growth rate (before and after the relevant period) may also be used. 2. Assessing Impact Using Regression Analysis: The inventors of the present invention found that for each client running a campaign, the relationship between the implementation (zero-one) of each campaign for each physician, the amount of investment in each campaign, and the impact on the increase in sales for each physician closely matches the following multiple regression equation (1). They also found that the regression coefficients multiplied by each investment in each term of this equation indicate the weight, or impact, of each investment.

[0024]

[0025] In other words, in the above formula (1), y is the amount of sales increase, m is the number of investments made, and x １ x ２ ... is the investment amount for each campaign, and the calculated regression coefficient represents the weight, or influence, of each investment.

[0026] Moreover, the above formula (1) can be expressed as the following formula (2).

[0027]

[0028] The regression coefficients of the above formula (2) can be calculated using the least squares method. The calculation formula for each regression coefficient (= the impact of each campaign) is the following formula (3). The calculation method may be a general method in the least squares method that calculates the regression coefficient β as follows, which minimizes the loss function.

[0029] In this way, a combination of the regression coefficient vector β=(β0 β1 . . . βm) is calculated.

[0030] Based on the regression coefficients obtained by the above regression analysis, the relative influence of each campaign n is calculated using the following formula (4): m is the number of campaigns (a natural number), and n is any natural number between 1 and m.

[0031]

[0032] Calculating total ROI: Calculate the total ROI by dividing the total increase in sales by the total investment for all campaigns. 4. Calculating ROI for each campaign: Calculate the ROI for each campaign by multiplying the relative impact of each campaign by the total ROI. 5. Visualizing and exporting results: Export the analysis results in Excel format and use them for decision-making.

[0033] <Daily Report Analysis> According to the present invention, the present invention further analyzes the daily reports of sales representatives, analyzes doctors' prescription trends and market feedback, and makes it possible to summarize and propose strategies.

[0034] <Details of Daily Report Analysis> Sales daily reports are recognized as an important source of information for the field, even in the pharmaceutical industry. However, the vast amount of data in these reports is not being fully utilized, resulting in the following issues: 1. Biased analysis by humans: When sales representatives and managers analyze daily reports, they tend to emphasize only the parts that are subjectively convenient, resulting in a lack of objective data analysis. 2. Insufficient reflection of analysis results in tactics and strategies: Daily report information is not effectively reflected in sales activities or strategic planning, and the value of daily reports is not fully utilized. 3. Wasted resources: Inappropriate analysis of daily reports leads to inefficient sales activities and unnecessary costs, which ultimately lead to increased pharmaceutical prices. To solve these issues, the present invention utilizes generative AI to automatically and objectively analyze sales daily reports and propose optimal sales activities tailored to the field.

[0035] This invention applies generative AI (e.g., a general-purpose generative AI server such as ChatGPT) to sales daily report data to extract important information, perform quantitative and qualitative analysis, and automatically generate sales strategies. This reduces the human burden while also contributing to optimizing sales activities and preventing pharmaceutical price increases. The main features of this invention are as follows: 1. Automatic data analysis: Significant sales-related data is extracted from sales daily reports and objective analysis is performed using generative AI. This enables analysis that eliminates traditional human bias. The general-purpose generative AI server can be a commonly available one, such as ChatGPT. 2. Sales optimization proposals: Based on the analysis results, tactics and strategies appropriate to the on-site situation are automatically generated. 3. Cost reduction: Using general-purpose generative AI technology to streamline traditional human analysis processes and eliminate unnecessary sales activities results in overall cost reduction.

[0036] To achieve the above functions, in the present invention, initial values ​​are set in the medical sales support information processing device. These initial values ​​are stored in the storage unit 104 of the server 10 as a doctor table 1012, a product table 1013, and an ROI table 1014. The doctor table 1012, the product table 1013, and the ROI table 1014 will be described in detail below.

[0037] FIG. 3 shows the data structure of the doctor table 1012. The doctor table 1012 is a table that stores and manages affiliation and attribute information for doctors who have the authority or actual authority to decide whether to adopt or prescribe medicines or medical devices. The administrator of the medical sales support information processing device inputs this information about doctors into the doctor table, which stores the information in records in the doctor table 1012. This allows users of the medical sales support information processing device to use the services of the present invention. The doctor table 1012 is a table with a facility doctor code as a primary key and columns for doctor no., doctor name, facility no., facility name, attribute 1, attribute 2, attribute 3, attribute 4, attribute 5, attribute 6, attribute 7, attribute 8, and registration date and time D, which is the date and time when this information was input.

[0038] The facility physician code is a field that stores identification information (e.g., numbers and symbols) related to the sales activity target, assigned to each combination of a physician and the facility to which it belongs. The facility physician code is a field that has a unique value assigned to each combination of a physician and a facility. The physician facility code may be a combination of a physician number and a facility number. The physician name and facility name are the personal name of the physician or the name of the facility, such as a hospital, indicated by the physician number and facility number, respectively. The physician name and facility name may be entered as any character string, such as an abbreviated name or nickname. Attributes 1 to 8 are fields that store attributes that may be related to the sales activity of the sales activity target. Publicly available information, such as age, gender, affiliated academic societies, and affiliated research groups, is often registered, but information not publicly available, such as the physician's email newsletter subscriptions and other information that is of interest to the physician, may also be used. The information registered in the records of the doctor table 1012 can be changed by an administrator, or it may be changed by information entered by the user of the medical sales support information processing device at the time of use and recorded in the activity simulation table 1015, described below. All input and output information to the medical sales support information processing device, including initial values, is recorded in an input and output record table (not shown), allowing the administrator to trace the history.

[0039] FIG. 4 shows the data structure of the product table 1013. The product table 1013 stores and manages information such as regression coefficients calculated in advance as a result of a logistics regression analysis performed on pharmaceuticals and other products targeted for sales activities. The product table 1013 is also a management table within the server. Therefore, it has fields for product number and product name, allowing for the distinct management of information for multiple users. The product table 1013 has columns for product number, product name, regression coefficients β01, β02, β03, β04, β05, β06, β07, and β08 corresponding to attributes 1 to 8, regression coefficients β09, β10, β11, β12, β13, β14, β15, β16, β17, β18, β19, β20, and β21 corresponding to activities 1 to 13, and a registration date and time P for the input date and time.

[0040] The product number is identification information that identifies the medical product targeted by multiple users of the medical sales support information processing device. The product name is the name of the product represented by the product number. Logistic regression analysis is performed for each individual product number, and regression coefficients, etc. are determined in advance. The product number is an item for which a unique value is assigned for each individual product. Note that individual users do not need to provide the product number each time they use the present invention. The regression coefficients, etc. for each individual product number are determined by logistic regression analysis based on pre-collected information on the attributes of each doctor for each individual product number, the type and frequency of sales activities, and the probability of product use and prescription. In addition, the technology described in Japanese Patent No. 7418877, a factor analysis method invented and patented by the inventor of the present invention, can also be utilized. The regression coefficients, etc. input as initial values ​​are pre-processed, such as by filling in missing data and processing outliers, as in general logistic regression analysis. Furthermore, a sufficient amount of data should be secured, and sufficient consideration should be given to avoiding multicollinearity and overfitting. As a more advanced method of logistic regression analysis, analysis using penalized logistic regression or multiclass logistic regression may also be performed. The regression coefficients obtained in this way (regression coefficients) corresponding to physician attributes 1 to 8 are stored in columns β01 to β08, and those corresponding to sales activities 1 to 13 are stored in columns β09 to β21. The date and time when these values ​​were input into the medical sales support information processing device is stored in the registration date and time P column.

[0041] The outline of the factor analysis method (Japanese Patent No. 7418877 (JP7418877B) obtained by the inventor of the present invention as described above) is a sales support technology based on factor analysis of doctor data.

[0042] 5 is a diagram showing the data structure of the ROI table 1014. The ROI table 1014 is a table for inputting initial values ​​for conducting an ROI analysis for each campaign when a medical customer or the like conducts a sales campaign for a product such as a pharmaceutical product that is the target of sales activities into the medical sales support device. That is, the ROI table 1014 is composed of calculations such as whether or not each campaign was conducted for each doctor who is a customer of the customer or the like, the monthly sales for each doctor, the monthly sales for each doctor in the previous year for each doctor, and the amount of sales increase calculated from these. The ROI table 1014 has the following columns: Customer_ID, which represents each doctor; Campaign_1, 2, 3, etc., which represent whether or not campaigns 1 to 3 (the numbers are examples; the same applies below) were implemented for each customer doctor; the year and month, which stores each month's sales to each doctor in the implementation year; the year and month, which stores each month's sales in the previous year; the total annual sales calculated from these; Incremental sales increase; and Incremental sales increase rate (%).

[0043] Customer_ID is identification information that identifies each doctor who is a customer, such as a client, or the facility to which the doctor belongs. This field is set with a unique value for at least each doctor. Campaign_1, 2, 3, etc. store a zero or one to indicate whether or not each of Campaigns 1 to 3 was implemented for each Customer_ID, and the year / month column stores the actual sales figures for each Customer_ID for the year the campaign was implemented, as well as the actual sales figures for the previous year. The total sales for the year calculated from these are stored in the TOTAL column, the sales increase amount in the Incremental column, and the sales increase rate in the Incremental (%) column. These values ​​may be input by creating an Excel list and uploading them all at once to the medical sales support information processing device.

[0044] <Configuration of control unit 104 of server 10> The control unit 104 of the server 10 includes an acquisition unit 1041, a calculation unit 1042, and an output unit 1043. The control unit 104 executes an application program 1011 stored in the storage unit 101 to realize the functions of each functional unit.

[0045] The acquisition unit 1041 acquires information input in advance as initial values ​​by the administrator of the medical sales support information processing device, and stores the information in the doctor table 1012, product table 1013, and ROI table 1014 of the storage unit 101. This completes the initial registration, and the information stored in the doctor table 1012, product table 1013, or ROI table 1014 becomes available for use. The acquisition unit 1041 also acquires information input by the user of the medical sales support information processing device, and stores the information in the activity simulation table 1015, ROI simulation table 1016, and daily sales report consulting table 1017 of the storage unit 101. This enables the calculation unit to perform calculations.

[0046] The calculation unit 1042 extracts information from the activity simulation table 1015, ROI simulation table 1016, and sales daily report consulting table 1017 acquired by the acquisition unit 1041 and stored in the memory unit 101, performs calculations using logit functions and the like using the application program 1011, and stores the results in designated parts of the activity simulation table 1015, ROI simulation table 1016, and sales daily report consulting table 1017.

[0047] The output unit 1043 extracts information from the activity simulation table 1015, ROI simulation table 1016, and daily sales report consulting table 1017, including the results calculated by the calculation unit 1042 and stored in the storage unit 101, and displays the information on a monitor display, prints it, or writes it as an electronic file in an appropriate format such as Excel for the user of the medical sales support information processing device. It may also display information from the doctor table 1012, product table 1013, or ROI table 1014 required by the user on a monitor display, print it, or write it as an electronic file in an appropriate format such as Excel. The output unit 1043 may be configured to display on a monitor display, print, or write out to an electronic file in an appropriate format such as Excel the information in the doctor table 1012, product table 1013, and ROI table 1014 in the memory unit 101, which is information that the administrator of the medical sales support information processing device has input in advance as initial values, and to display on a monitor display, print, or write out to an electronic file in an appropriate format such as Excel the information stored in the input / output history table.

[0048] The medical sales support information processing device may include a communication unit that executes communication processing with the customer's information processing terminal 20, 30 or any other information processing terminal.

[0049] <Configuration of Information Processing Terminals 20 and 30> The customer's information processing terminals 20 and 30 may be any terminal that can transmit the necessary information to the user of the medical sales support information processing device, and may be a server, desktop PC, laptop PC, mobile terminal, dedicated terminal, etc. Furthermore, the customer's information processing terminals 20 and 30 may have a function to display the information received by the user of the medical sales support information processing device from the server 10, but are not limited to this.

[0050] <Configuration of Storage Unit 201 of Information Processing Terminal 20 / 30> The storage unit 201 (not shown) of the information processing terminal 20 / 30 includes an application program 2011 and a terminal table 2012 (not shown).

[0051] The application program 2011 may be pre-stored in the storage unit 201, or may be downloaded from a web server operated by a service provider via a communication IF. The application program 2011 includes an application such as a web browser application. The application program 2011 may include an interpreter-type programming language such as JavaScript (registered trademark) that is executed on a web browser application stored in the information processing terminal 20 / 30.

[0052] <Operation of the Medical Sales Support Information Processing Device> Each process of the medical sales support information processing device will be described in detail below. FIG. 6 shows an activity simulation table 1015 used in the <Optimized Sales Action Calculation Process>. FIG. 9, which relates to this, is a flowchart showing the operation of the <Optimized Sales Action Calculation Process>. FIG. 10 is a flowchart showing the operation of the <Cosine Similarity Analysis> process. FIG. 7 shows an ROI simulation table 1016 used in the <ROI Analysis> process. FIG. 11, which relates to this, is a flowchart showing the operation of the <ROI Analysis> process. FIG. 8 shows a daily sales report consulting table 1017 used in the <Daily Sales Report Analysis> process. FIG. 12, which relates to this, is a chart showing the operation of the <Daily Sales Report Analysis> process.

[0053] <Activity Simulation Table 1015 for Optimized Sales Action Calculation Processing> Figure 6 is a diagram showing the data structure of the activity simulation table 1015 used in the <Optimized Sales Action Calculation Processing>. Note that Figure 6 is an example of a table for management purposes and is not, in principle, displayed to customers. The activity simulation table 1015 is a table into which the attributes of doctors who were actually approached by a user of the medical sales support information processing device for products such as pharmaceuticals that are the subject of sales activities and the details of the sales activities conducted toward those doctors are input for the <Optimized Sales Action Calculation Processing>, and which stores and manages the calculated <Optimized Sales Actions>. The activity simulation table 1015 is stored on a server. The activity simulation table 1015 has columns for product number, product name, facility doctor code, doctor no., doctor name, facility no., facility name, use or prescription probability "before" p0, x01, x02, x03, x04, x05, x06, x07, x08 as explanatory variables corresponding to attributes 1 to 8 to be multiplied by the regression coefficients obtained by logistic regression analysis, x09, x10, x11, x12, x13, x14, x15, x16, x17, x18, x19, x20, x21 as explanatory variables corresponding to activities 1 to 13, the registration date and time a0 of the above, use or prescription probability "after" p1 for storing the results after the <optimized sales action calculation process>, R1, R2, R3 for storing the names of recommended activities, and the registration date and time a1 for storing the results after the <optimized sales action calculation process>. Note that if there are not enough columns, additional columns may be added.

[0054] The product number is identification information that identifies the medical product that the user of the medical sales support information processing device intends to use. The product name is the name of the product represented by the product number. However, this is used only by the management information processing terminal and the server 10 and does not need to be known to the user. The <Optimized Sales Action Calculation Process> is performed for each individual product number. The product number is an item for which a unique value is set for each individual product, and corresponds to the product number in the product table 1013.

[0055] The use or prescription probability "before" p0 is the probability that the target product will be used or prescribed by the doctor in question, calculated from x01, x02, x03, x04, x05, x06, x07, and x08 as explanatory variables corresponding to attributes 1 to 8, and x09, x10, x11, x12, x13, x14, x15, x16, x17, x18, x19, x20, and x21 as explanatory variables corresponding to activities 1 to 13, which are stored in advance in activity simulation table 1015, and the regression coefficients stored as initial values ​​in the row having the same product number in product table 1013. On the other hand, the use or prescription probability "after" p1 is the probability that the target product will be used or prescribed by the doctor in question after the <optimized sales action calculation process> is performed. Furthermore, R1, R2, and R3, which store the names of recommended activities, are names of activities that contribute greatly to achieving the use or prescription probability "after" p1, but more activity names may be recorded by adding a recording column, not limited to the three with the highest contributions. The registration date and time a1 stores the date and time when the results after the <optimized sales action calculation process> were stored.

[0056] <Details of the Optimized Sales Action Calculation Process> Fig. 9, which is related to Fig. 6, is a flowchart showing the operation of the <Optimized Sales Action Calculation Process>. Details of the <Optimized Sales Action Calculation Process> will be explained below.

[0057] In step S100 of Figure 9, the control unit 104 of the server 10 acquires information for the doctor table 1012 and the product table 1013 input from the management information processing terminal 15, and stores the information as initial values ​​in each column of the doctor table 1012 and the product table 1013.

[0058] In step S110, the control unit 104 of the server 10 acquires information for the activity simulation table 1015 that the customer has collected from any terminal, such as the customer's information processing terminal 20 or 30, and stores the information in each column for before the <optimized sales action calculation process> in the activity simulation table 1015. Specifically, from any terminal, the following are stored: the target product name, facility doctor code, facility name, doctor name, x01, x02, x03, x04, x05, x06, x07, and x08 as explanatory variables corresponding to attributes 1 to 8, x09, x10, x11, x12, x13, x14, x15, x16, x17, x18, x19, x20, and x21 as explanatory variables corresponding to activities 1 to 13, and the registration date and time a0 for the above.

[0059] In step S120, the control unit 104 of the server 10 compares the doctor identification information and the explanatory variables for the doctor's attributes stored in the activity simulation table 1015 (i.e., the facility doctor code, facility name, doctor name, and the explanatory variables x01, x02, x03, x04, x05, x06, x07, and x08 corresponding to attributes 1 to 8) with the corresponding records stored in the doctor table 1012 (step S130). If there are any discrepancies, the record in the doctor table is overwritten with the information stored in the activity simulation table 1015 (step S135). The process then returns to step S120 and executes step S120 again. If there are no discrepancies, the process proceeds to step S140 without any further action. Because all input / output information is recorded in an input / output history table (no diagram), even if it is later discovered that the overwritten information was incorrect, it is possible to determine what the original data was.

[0060] In step S140, the <optimized sales action calculation process> is performed. Prior to this, the control unit 104 of the server 10 first reads out the regression coefficients and the like stored in the product table 1013 and corresponding to the product numbers in the activity simulation table 1015, and calculates the regression coefficients x01, x02, x03, x04, x05, x06, x07, and x08 as explanatory variables corresponding to attributes 1 to 8 in the activity simulation table 1015, and x09, x10, x11, x12, x13, x14, x15, x16, and x17 as explanatory variables corresponding to activities 1 to 13. The same product numbers are multiplied by β01, β02, β03, β04, β05, β06, β07, β08, β09, β10, β11, β12, β13, β14, β15, β16, β17, β18, β19, β20, β21 stored in the product table 1013, and the probability of use or prescription "before" p0 is calculated using a logit function, which is stored in the corresponding column of the activity simulation table 1015. In this case, an example of the formula for calculating p0 using the logit function is as follows: (5) (where pi = p0), but the formula is not limited to this.

[0061]

[0062] Furthermore, in step S140, the combination of the usage or prescription probability "after" p1, which is the maximum probability that the target product will be used or prescribed to the target doctor after the <optimized sales action calculation process> is performed, and the x values ​​related to the activity to obtain it are calculated and recorded. That is, while x01, x02, x03, x04, x05, x06, x07, and x08, which are explanatory variables corresponding to attributes 1 to 8 identifying the doctor stored in the activity simulation table 1015, are left fixed, x09, x10, x11, x12, x13, x14, x15, x16, x17, x18, x19, x20, and x21, which are explanatory variables corresponding to activities 1 to 13, are changed, and the combination of x09 to x21 that maximizes the usage or prescription probability "after" p1 is calculated. This calculated value is stored in the input / output history table, and at the same time, in step S150, the name of the activity with the greatest contribution, the use or prescription probability "after" p1, and the calculation date and time thereof are stored in the columns of the use or prescription probability "after" p1, R1, R2, R3, and registration date and time a1 of the activity simulation table 1015.

[0063] When calculating the combination of x09 to x21 that maximizes the use or prescription probability "after" p1, p1 is calculated by the above formula (5). ｉ = p １ To calculate the combination of x09 to x21 that maximizes the use or prescription probability "after" p1, in addition to the formula shown in this specification, a general method for calculating the optimal combination using a logistic function or logit function may be used, or the maximum likelihood method may be used. The activity with the greatest impact is selected from the terms with β09 to β21 (terms related to activities) in the following formula (6):

[0064]

[0065] That is, for each term of the following formula (7), for example, three items with the largest calculated values ​​may be selected, and the names of the activities corresponding to these items may be stored as the top three activities with the highest contributions.

[0066]

[0067] Furthermore, the targets for this <optimized sales action calculation process> may be limited to only those doctors whose usage or prescription probability "before" p0 is low, by setting a threshold (for example, less than 50%).

[0068] In step S160, the control unit 104 of the server 10 may output, via the output unit, the use or prescription probability "after" p1, which is the probability that the target product will be used or prescribed to the doctor after the <optimized sales action calculation process> is performed, a combination of the value of x related to the activity to achieve that probability, and the name of the activity that contributes most. The output may be displayed on a monitor display, printed, or written to an electronic file.

[0069] Table for Cosine Similarity Analysis No dedicated table is required for the cosine similarity analysis.

[0070] <Details of Cosine Similarity Analysis> Figure 10 is a flowchart showing the operation of the <Cosine Similarity Analysis> calculation process. Details of the <Cosine Similarity Analysis> are explained below. The above <Optimized Sales Action Calculation Process> makes it possible to analyze the similarity for doctors who have a high probability of using or prescribing a drug. Therefore, it can be said that the <Cosine Similarity Analysis> after the <Optimized Sales Action Calculation Process> is performed produces a synergistic effect.

[0071] In step S200 of FIG. 10, the control unit 104 of the server 10 acquires information for the doctor table 1012 and the product table 1013 from the management information processing terminal 15 and stores the information in each column of the doctor table 1012 and the product table 1013 as initial values.

[0072] In step S210, the control unit 104 of the server 10 acquires, through the acquisition unit, the facility doctor code stored in the product table 1013 after performing the <Optimized Sales Action Calculation Process> or the facility doctor code specified by the customer from any terminal such as the customer's information processing terminal 20 / 30, as information representing the similar source.

[0073] In step S220, the calculation unit of the control unit 104 of the server 10 calculates the cosine similarity between the attribute information of the doctor associated with the facility doctor code acquired in step S210 and the attribute information of each doctor stored in the doctor table. The results of these similarities may be recorded in the input / output history table. Furthermore, in step S230, the control unit 104 of the server 10 may store the results in the doctor table 1012. A storage column may be added to the doctor table 1012.

[0074] In step S240, the control unit 104 of the server 10 selects a doctor with a high degree of similarity from the result of step S220. Note that the selection of a doctor with a high degree of similarity is just one example, and a desired range of similarity can also be specified for selection.

[0075] In step S240, the output unit 1043 of the control unit 104 of the server 10 outputs information about doctors who have the desired similarity. The output may be displayed on a monitor display, printed, or written to an electronic file.

[0076] Figure 7 shows the data structure of an ROI simulation table 1016 used for ROI analysis. The ROI simulation table 1016 stores and manages campaign activities conducted by customers for products such as pharmaceuticals targeted for sales activities, sales results for the target period for performance verification, sales results for the previous year, and ROI analysis results. The ROI simulation table 1016 has columns for customer no., customer name, investment amount for each campaign activity conducted for campaigns 1 to 5, total number of campaigns conducted, absolute sum of relative impacts for the conducted campaigns, calculated relative impacts for campaigns 1 to 5, monthly settlement 1 (Jan. to Dec., Total) which is sales results for the target period, monthly settlement 0 (Jan. to Dec., Total) which is sales results for the previous year, sales increase amount which is the increase in sales results, total ROI, ROI for each campaign for campaigns 1 to 5, and the registration date and time r in which these are stored. The ROI analysis according to the present invention evaluates the impact of the entire campaign by statistically aggregating and analyzing the individual customer data stored in the ROI simulation table 1016. Therefore, the "ROI per campaign" and "relative impact" listed in the table are not values ​​for each individual customer, but aggregated indicators based on the overall analysis.

[0077] A customer is a company that uses a medical sales support information processing device, such as a pharmaceutical company or a medical device manufacturer / distributor. The customer no. is an item in which a unique value corresponding to the customer name is set. A campaign is a special event held to promote sales to doctors, and although it incurs expenses, it is held in the hope of increasing sales. The amount invested in the campaign activities held is the actual amount invested by the customer in each campaign. Even if the same campaign is held multiple times in the same target year, each is considered to be a separate campaign event by assigning a number or other identifier to distinguish them. The number of times held is the total number of campaigns held by the customer in the year.

[0078] The regression coefficient β value corresponding to each campaign is calculated by the ROI analysis. The sum of the absolute values ​​of the regression coefficient β values ​​corresponding only to the implemented campaigns shown in the ROI simulation table 1016 is stored in the "Implemented β Value Absolute Value Sum" field. The relative influence is calculated by using the actual absolute value sum of the β values ​​as the denominator and the β value of each campaign as the numerator, and is stored.

[0079] The actual sales figures for each month of the target product from January to December of the target year are read from the ROI table 1014, which is the default table, and stored in Monthly Settlement 1 (January-December column), with the total amount stored in the Total 1 column. The actual sales figures for each month of the target product from January to December of the year prior to the target year are stored in Monthly Settlement 0 (January-December column), with the total amount stored in the Total 0 column. The amount obtained by subtracting the Total 0 column from the Total 1 column is then stored in the Sales Increase column. Note that this sales increase amount may be acquired by the acquisition unit based on the numerical value entered by the user at the time of use. As long as the corresponding months of the current and previous years are stored, it is acceptable for some months to have missing sales figures, and it is not necessary for all December figures to be filled in. In this case, numerical values ​​such as campaign investment amounts will also be calculated for the entered sales period.

[0080] The total ROI column stores the value obtained by dividing the increase in sales by the total amount invested in the campaigns. The ROI columns for each campaign, Campaigns 1 to 5, store the percentage obtained by dividing the funds invested in each campaign by the value obtained by multiplying the increase in sales by the relative impact. The registration date and time r stores the date and time when the ROI was calculated. Note that the number of campaigns is not limited to five, and columns for Campaigns 1 to 5, etc. may be added as needed.

[0081] Furthermore, if it is possible to calculate the total cost of regular sales activities that users normally conduct with doctors over the target period, this can be treated as the investment amount for one campaign. If this is possible, it will be possible to grasp the factors that affect the increase in sales more comprehensively and consider whether each other campaign is more or less efficient than regular sales activities.

[0082] <ROI Analysis Details> Fig. 11, which is related to Fig. 7, is a flowchart showing the operation of <ROI analysis>. Details of <ROI analysis> are explained below. In step S300 of Fig. 11, the control unit 104 of the server 10 acquires information for the ROI table 1014 from the management information processing terminal 15 and stores the information.

[0083] In steps S310 and S320, the control unit 101 of the server 10 acquires the information of the ROI table 1014, which is the initial value, and stores it in the columns for the investment amount for each campaign activity carried out, the number of activities carried out, monthly settlement 0, and monthly settlement 1.

[0084] The number of campaigns implemented and the amount of investment in each campaign activity for Campaigns 1 to 5 are acquired from a specified terminal and stored. Note that this information may be acquired using general-purpose AI or prompts using Python (registered trademark), etc.

[0085] In step S330, the calculation unit 1042 of the control unit 104 of the server 10 applies each of the acquired values ​​to the aforementioned formula to calculate the regression coefficient for each campaign activity. The control unit 104 of the server 10 stores the sum of the absolute values ​​of these regression coefficients in the "total absolute β value of implementation" field. The calculation unit 1042 of the control unit 104 of the server 10 also stores the value obtained by dividing the regression coefficient β value corresponding to the campaign 1-5 fields in the ROI simulation table 1016, where a numerical value is stored in the "investment amount for each implemented campaign activity" field, by the value in the "total absolute β value of implementation" field, in the corresponding campaign field for each campaign 1-5 corresponding to the relative influence. The calculation unit 1042 of the control unit 104 of the server 10 also stores the value obtained by dividing the sales increase amount by the total investment amount for campaign activities in the "total ROI" field.

[0086] In step S340, the calculation unit of the control unit 104 of the server 10 stores the value obtained by multiplying the sales increase amount by the relative impact of each campaign, divided by the investment amount for each campaign, in the campaign columns for campaigns 1 to 5 of the ROI. That is, the calculation result of the following formula (8) is stored.

[0087]

[0088] In step S350, the output unit of the control unit 104 of the server 10 outputs the ROI of each campaign. The output may be displayed on a monitor display, printed, or written to an electronic file.

[0089] <Daily Sales Report Consulting Table 1017 for Daily Sales Report Analysis> Figure 8 shows the data structure of the daily sales report consulting table 1017 used in the <Daily Sales Report Analysis> process. The daily sales report consulting table 1017 is a table into which daily sales report text data of activities actually conducted by MRs (Medical Representatives, hereinafter the same) or others with doctors approached for products such as pharmaceuticals that are the subject of sales activities is input in order to perform <Daily Sales Report Analysis>, and which stores and manages reports including summaries and countermeasure recommendations as a result of the <Daily Sales Report Analysis>. The daily sales report consulting table 1017 is stored on a server. The daily sales report consulting table 1017 has columns for product number, product name, doctor facility code, doctor no., doctor name, facility no., facility name, daily report data, AI_response, and registration date and time t recording the AI_response.

[0090] The product number is identification information that identifies the medical product that the user of the medical sales support information processing device is trying to use. However, the product number is not usually displayed on the screen, but is used for internal processing. The product name is the name of the product represented by the product number. <Sales Daily Report Analysis> is performed for each individual product number. The product number is an item for which a unique value is set for each individual product, and corresponds to the product number in the product table 1013.

[0091] The daily report data column stores the daily report text data of the activities that MRs and others actually conducted with doctors they approached regarding products such as pharmaceuticals that are the target of sales activities before the <daily report analysis> process is performed. Reports including summaries and countermeasure recommendations are created using an external general-purpose generation AI server. The general-purpose generation AI server can be a general one such as ChatGPT. The created report including summaries and countermeasure recommendations is stored in the AI_response column. The registration date and time column stores the date and time when the data was stored in the AI_response column after the <daily report analysis> process was performed.

[0092] <Details of Daily Sales Report Analysis> Fig. 12, which is related to Fig. 8, is a chart showing the operation of the <Daily Sales Report Analysis> process. Details of the <Daily Sales Report Analysis> will be explained below.

[0093] In step S12 of Figure 12, the acquisition unit 1041 of the control unit 104 of the server 10 acquires information (S11) of the daily sales report (text) for the target period for which the customer collected information from any terminal such as the customer's information processing terminal 20, 30, and stores it in the daily report data column of the daily sales report consulting table 1017.

[0094] The acquisition unit 1041 of the control unit 104 of the server 10 identifies the doctor facility code (S14), target period (S15), etc. from the text information entered in the daily report data column or from any terminal such as the customer's information processing terminal 20 / 30.

[0095] The acquisition unit 1041 of the control unit 104 of the server 10 analyzes the text stored in the daily report data using natural language processing (NLP) technology, extracts important information from the vast amount of daily reports and summarizes it concisely (summary generation function), objectively identifies on-site problems (issue extraction function), and creates prompts including instructions for generating appropriate sales strategy proposals (improvement proposal function), and sends a generation request to the generation AI server (S17). The generation server converts the specified summary and items into text, embeds it in a specified form (drawings not attached), and sends a response to the server 10 (S19). The control unit 104 of the server 10 stores the response in the AI_response column of the daily sales report consulting table 1017, and simultaneously stores the date and time of storage in the AI_response column in the registration date and time t column.

[0096] In step S22, the output unit 1043 of the control unit 104 of the server 10 outputs the content of the AI_response field in a predetermined form. The output may be sent to the customer's information processing terminal 20 or the like (S21) to be displayed on a monitor display (S22), or may be printed or written to an electronic file.

[0097] 14 is a block diagram showing the basic hardware configuration of a typical computer 90. The computer 90 includes at least a processor 901, a main storage device 902, an auxiliary storage device 903, and a communication IF 991 (interface). These components are electrically connected to one another by a communication bus 921.

[0098] The processor 901 is hardware for executing an instruction set written in a program, and is composed of an arithmetic unit, a register, a peripheral circuit, and the like.

[0099] The main storage device 902 is used to temporarily store programs, data to be processed by the programs, etc. For example, it is a volatile memory such as a DRAM (Dynamic Random Access Memory).

[0100] The auxiliary storage device 903 is a storage device for saving data and programs, such as a flash memory, a hard disk drive (HDD), a magneto-optical disk, a CD-ROM, a DVD-ROM, or a semiconductor memory.

[0101] The communication IF 991 is an interface for inputting and outputting signals for communicating with other computers via a network using a wired or wireless communication standard. The network is composed of the Internet, a LAN, various mobile communication systems constructed using wireless base stations, etc. For example, the network includes 3G, 4G, and 5G mobile communication systems, LTE (Long Term Evolution), and wireless networks (e.g., Wi-Fi (registered trademark)) that can connect to the Internet via a predetermined access point. In the case of a wireless connection, communication protocols include, for example, Z-Wave (registered trademark), ZigBee (registered trademark), Bluetooth (registered trademark), etc. In the case of a wired connection, the network also includes a direct connection via a USB (Universal Serial Bus) cable or the like.

[0102] It is also possible to virtually realize the computer 90 by distributing all or part of each hardware configuration across multiple computers 90 and interconnecting them via a network. In this way, the concept of the computer 90 includes not only a computer 90 housed in a single housing but also a virtualized computer system.

[0103] <Basic Functional Configuration of Computer 90> A description will be given of the functional configuration of the computer realized by the basic hardware configuration (FIG. 14) of the computer 90. The computer includes at least the functional units of a control unit, a storage unit, and a communication unit.

[0104] The functional units of the computer 90 can also be realized by distributing all or part of the functional units among multiple computers 90 interconnected via a network. The computer 90 is a concept that includes not only a single computer 90 but also a virtualized computer system.

[0105] The control unit executes processing in accordance with the programs by the processor 901 reading various programs stored in the auxiliary storage device 903 and loading them into the main storage device 902. The control unit can realize the functions of functional units that perform various types of information processing depending on the type of program. This allows the computer to function as an information processing device that performs information processing.

[0106] The storage unit is realized by a main storage device 902 and an auxiliary storage device 903. The storage unit stores data, various programs, and various databases. The processor 901 can allocate a storage area corresponding to the storage unit in the main storage device 902 or the auxiliary storage device 903 in accordance with the programs. The control unit can cause the processor 901 to add, update, and delete data stored in the storage unit in accordance with the various programs.

[0107] A database refers to a relational database, which manages data sets called tables and masters in a tabular format structurally defined by rows and columns, by associating them with each other. In a database, a table is called a table or master, a column in a table is called a column, and a row in a table is called a record. In a relational database, relationships between tables and masters can be set and associated. Typically, each table and each master has a column set as a primary key to uniquely identify a record, but setting a primary key to a column is not required. The control unit can cause the processor 901 to add, delete, or update records in specific tables and masters stored in the storage unit according to various programs.

[0108] Each table, database, and master in the present invention may include any data structure in which information is structurally defined (such as a list, dictionary, associative array, or object). Data structures also include data that can be considered as a data structure by combining data with functions, classes, methods, etc. written in any programming language.

[0109] The communication unit is realized by the communication IF 991. The communication unit realizes the function of communicating with other computers 90 via a network. The communication unit can receive information transmitted from other computers 90 and input the information to the control unit. The control unit can cause the processor 901 to execute information processing on the received information in accordance with various programs. Furthermore, the communication unit can transmit information output from the control unit to other computers 90. Specific examples of implementation

[0110] As specific examples of the implementation, first, second and third embodiments will be described below.

[0111] <Embodiment 1> Company A is a pharmaceutical company that developed a new drug, Drug P, an anti-obesity drug, and launched it two years ago. Company A's MRs (Medical Representatives; same below) and wholesalers have conducted various sales activities by visiting doctors at various hospitals and other facilities, and as a result, have gained a 15% market share. Clinical prescription results for Drug P have been good, and Company A believes it can still increase sales. However, for some reason, some doctors are reluctant to prescribe Drug P, and the company believes that it would be a waste to allocate sales resources uniformly to such doctors. Therefore, the company decides to utilize the present invention.

[0112] In the medical sales support information processing device of the present invention, in step S100 of Figure 9, the control unit 104 of the server 10 obtains information for the doctor table 1012 and the product table 1013 from the management information processing terminal 15 and stores the information in each column of the doctor table 1012 and the product table 1013, with the initial values ​​already having been entered, as data on each hospital facility since the new drug P was launched, the attributes of the doctors affiliated with those facilities, the details of sales activities, and the results (usage or prescription).

[0113] Company A's MR will compile in an Excel A form as much information as they know about the attributes of each hospital facility and doctor they are going to approach about the new drug P, as well as the planned sales activities they are going to carry out.

[0114] In the medical sales support information processing device of the present invention, in step S110, the control unit 104 of the server 10 acquires the information necessary for the <optimized sales action calculation process> for the activity simulation table 1015 and stores it in each column of the activity simulation table 1015.

[0115] Specifically, the facility doctor code, facility name, doctor name, attributes (gender, age, affiliated academic society, etc.), planned activities 1 to 13 (future sales activities such as face-to-face interviews, online interviews, email awareness campaigns, and email newsletter distribution), product names, product numbers, other affiliation information that have been stored in advance on the server, and regression coefficients for logistic regression analysis that have been calculated in advance using another technology and appropriately preprocessed are obtained.

[0116] In step S120, the control unit 104 of the server 10 compares the doctor identification information and the explanatory variables for the doctor's attributes stored in the activity simulation table 1015 (i.e., the facility doctor code, facility name, doctor name, and the explanatory variables x01, x02, x03, x04, x05, x06, x07, and x08 corresponding to attributes 1 to 8) stored in the activity simulation table 1015 with the corresponding records stored in the doctor table 1012 (step S130). If there are any discrepancies, the record in the doctor table 1012 is overwritten with the information stored in the activity simulation table 1015 (step S135). Then, the process returns to step S120 and executes step S120 again. If there are no discrepancies, the process proceeds to step S140 without any further action. Because all input / output information is recorded in the input / output history table, even if it is later discovered that the overwritten information was incorrect, it is possible to determine what the original data was.

[0117] In step S140, the control unit 104 of the server 10 performs the <Optimization Sales Action Calculation Process>. Prior to this, the control unit 104 of the server 10 first reads out the regression coefficients and the like stored in the product table 1013 and corresponding to the product numbers in the activity simulation table 1015, and calculates the regression coefficients and the like as explanatory variables corresponding to attributes 1 to 8 in the activity simulation table 1015, such as x01, x02, x03, x04, x05, x06, x07, and x08, as well as the explanatory variables corresponding to activities 1 to 13, such as x09, x10, x11, x12, x13, x14, x15, x16, and x17. The same product numbers are multiplied by β01, β02, β03, β04, β05, β06, β07, β08, β09, β10, β11, β12, β13, β14, β15, β16, β17, β18, β19, β20, β21 stored in product table 1013, and the use or prescription probability "before" p0 is calculated using a logit function and stored in the corresponding column of activity simulation table 1015. The above registration date and time a0 is also stored.

[0118] Furthermore, in step S140, the system calculates and records the "after" use or prescription probability p1, which is the maximum probability that the target product will be used or prescribed to the doctor after the <Optimized Sales Action Calculation Process> is performed, and the combination of x values ​​related to the activity to obtain this probability. That is, while x01, x02, x03, x04, x05, x06, x07, and x08, which are explanatory variables corresponding to attributes 1 to 8 identifying the doctor stored in the activity simulation table 1015, remain fixed, the values ​​of x09, x10, x11, x12, x13, x14, x15, x16, x17, x18, x19, x20, and x21, which are explanatory variables corresponding to activities 1 to 13, are opened, and the combination of x09 to x21 that maximizes the "after" use or prescription probability p1 is calculated using the logit function described above, etc. This calculated value is stored in the input / output history table, and at the same time, in step S150, the names of the three activities with the highest contribution, the use or prescription probability "after" p1, and the calculation date and time are stored in the columns of the use or prescription probability "after" p1, R1, R2, R3, and registration date and time a1 of the activity simulation table 1015.

[0119] In step S160, the control unit 104 of the server 10 displays on the monitor display via the output unit 1043 the use or prescription probability “after” p1, which is the maximum probability that the target product will be used or prescribed by the doctor after the <optimized sales action calculation process> is performed, as well as the combination of x values ​​related to the activities required to achieve this probability and the names of the three activities with the greatest contribution.

[0120] The results showed that even if the currently planned activities of eight interview visits, 12 email awareness campaigns, and two mailings of information were carried out, the probability that new drug P would be adopted and prescribed by Doctor D, whom Company A's MR was trying to approach, would only be 44%. However, if 19 interview visits, 15 email awareness campaigns, and five mailings of information were carried out, the probability that new drug P would be adopted and prescribed by Doctor D would rise to 83%, and the above three activities would be the activities that would make a significant contribution to this. On the other hand, for Doctor F, who the MR from Company A is also trying to approach, even if the currently planned activities of 11 home visits, 15 email awareness campaigns, and 7 mailings of materials are carried out, the probability that Doctor D will adopt and prescribe new drug P is only 21%. Even if the activities are maximized to 29 home visits, 29 email awareness campaigns, and 13 mailings of materials, which would achieve the highest predicted probability, the probability that Doctor D will adopt and prescribe new drug P is only increased to 26%.

[0121] Upon seeing this, the MR at Company A writes the details into an Excel file and decides to prioritize activities with Doctor D over Doctor F.

[0122] In addition, the MR manager of Company A wants to know whether each of the sales campaigns that have been implemented this fiscal year has been effective and how efficient each campaign's approach has been, so he decides to carry out the ROI analysis provided by the present invention.

[0123] Specifically, to conduct a campaign ROI analysis, the medical sales support information processing device of the present invention acquires information on sales results for the current and previous years, the total number of campaigns implemented, and the investment amount for each campaign, and after carrying out the procedure of the ROI analysis flowchart, it is possible to display the investment efficiency for each campaign. According to this, it is clear that out of the total ROI of 43.6%, the ROI of Campaign 3 was 217% efficient, but Campaign 1 was 38%, which was below the total ROI value, and Campaign 2 was only 23%, dragging down the overall ROI.

[0124] Therefore, the MR manager at Company A decides to focus on campaign 3 from now on, cancel campaign 2, and run campaign 1 for another year to see how the ROI changes.

[0125] At the very least, by canceling Campaign 2, which is inefficient for both the company and its customers, it appears possible to reduce investment costs.

[0126] <Embodiment 2> Company B is a pharmaceutical company that developed a new anti-allergy drug, AL, for the treatment of hay fever and launched it one and a half years ago. Due to the increasing number of patients and subtle changes in allergens year by year, new anti-allergy drugs for the treatment of hay fever are being launched by various companies almost every year. Sales of long-standing treatments are declining, making it important to switch to new drugs. Company B's medical representatives and wholesalers have not been fully engaged in sales activities, such as visiting doctors at hospitals and other facilities. However, a competitor, Company C, launched an anti-allergy drug with very similar efficacy and mechanism of action three years ago as an anti-allergy drug. This anti-allergy drug still has a low market share of less than 10%, and Company B would like to target the same market. While the medical sales support information processing device of the present invention has only a small amount of registered activity data for the company's new drug, AL, it has accumulated a considerable amount of data on anti-allergy drugs for the treatment of hay fever, including anti-allergy drugs. Therefore, the present invention will be utilized.

[0127] As described above, data on the attributes of each hospital facility and its affiliated physicians, as well as sales activities and results (usage or prescription) for anticoagulants and other products similar to the new drug AL since their launch, are registered in advance by the administrator of the medical sales support information processing device of the present invention. Each time this information becomes available in the past, in step S100 of the medical sales support information processing device of the present invention, the control unit 104 of the server 10 acquires information for the doctor table 1012 and the product table 1013 from the management information processing terminal 15 and stores it in each column of the doctor table 1012 and the product table 1013, providing sufficient storage information as initial values. Data on anticoagulants and other similar drugs similar to the new drug AL is then integrated in preprocessing and registered as initial values ​​with new similar product numbers.

[0128] Company B's MR will compile in an Excel A form as much information as they know about the attributes of each hospital facility and doctor that they are planning to approach regarding the new drug AL, as well as the planned sales activities that they are planning to carry out.

[0129] In the medical sales support information processing device of the present invention, in step S110, the control unit 104 of the server 10 acquires information for these activity simulation tables 1015 and stores it in each column required to perform the <optimized sales action calculation process> of the activity simulation table 1015.

[0130] Specifically, the facility doctor code, facility name, doctor name, attributes (gender, age, affiliated academic society, etc.), planned activities 1 to 13 (future sales activities such as face-to-face interviews, online interviews, email awareness campaigns, and email newsletter distribution), product names, product numbers, other affiliation information that have been stored in advance on the server, and regression coefficients for logistic regression analysis that have been calculated in advance using another technology and appropriately preprocessed are obtained.

[0131] In step S120, the control unit 104 of the server 10 compares the information identifying the doctor stored in the activity simulation table 1015 with explanatory variables regarding the doctor's attributes and performs any necessary updates based on the called application program 1011.

[0132] In step S140, the <Optimization Sales Action Calculation Process> is performed. Prior to this, the control unit 104 of the server 10 first reads the regression coefficients and the like corresponding to the aforementioned similar product numbers in the activity simulation table 1015 stored in the product table 1013, and multiplies the explanatory variables corresponding to attributes 1 to 8 in the activity simulation table 1015 and the explanatory variables corresponding to activities 1 to 13 by β01, β02, β03, β04, β05, β06, β07, β08, β09, β10, β11, β12, β13, β14, β15, β16, β17, β18, β19, β20, and β21 stored in the product table 1013, each for the same product number, and calculates the usage or prescription probability "before" p0 using a logit function in the same manner as in embodiment 1, and stores this in the corresponding column of the activity simulation table 1015. The above registration date and time a0 are also stored.

[0133] Furthermore, in step S140, the combination of the use or prescription probability "after" p1, which is the maximum probability that the target product will be used or prescribed to the doctor after the <optimized sales action calculation process> is performed, and the value of x related to the activity to achieve it is calculated and recorded.

[0134] In step S160, the control unit 101 of the server 10 displays on the monitor display via the output unit 1043 the use or prescription probability “after” p1, which is the maximum probability that the target product will be used or prescribed by the doctor after the <optimized sales action calculation process> is performed, as well as the combination of x values ​​related to the activities to achieve this and the name of the activity with the highest contribution.

[0135] According to the information, even if the currently planned activities of three online interviews, three email awareness campaigns, and two mailings of materials are carried out for Doctor G, who the MR from Company B is trying to approach, the probability that Doctor G will adopt and prescribe the new drug AL will only be 38%, but if 22 online interviews, 17 email awareness campaigns, and seven mailings of materials are carried out, the probability that Doctor G will adopt and prescribe the new drug AL will rise to 89%, and the above three activities will be activities that will greatly contribute to this. On the other hand, for Doctor H, who is also being approached by Company B's MR, even if the currently planned activities of 9 interview visits, 13 email awareness campaigns, and 9 mailings of materials are carried out, the probability that Doctor D will adopt and prescribe the new drug AL is only expected to be 29%; and even if the activities are maximized to 39 interview visits, 25 email awareness campaigns, and 9 mailings of materials, which would achieve the highest predicted probability, the probability that Doctor G will adopt and prescribe the new drug AL is only expected to increase to 31%.

[0136] Furthermore, in order to search for doctors who are likely to be able to take an efficient approach instead of Doctor H and who have high similarity in attributes with Doctor G, who will have a higher prescription probability based on the above results, the MR of Company B clicks a button on the screen for obtaining a similarity analysis list from the customer information processing terminal 30 of the medical sales support information processing device of the present invention. The medical sales support information processing device of the present invention obtains this information, executes the flowchart of <Cosine Similarity Analysis> (Figure 10), and then displays a list of doctors who are highly similar to Doctor G. Note that recommended activities and usage or prescription probability can also be calculated for these highly similar doctors, just as for Doctor G.

[0137] The MR from Company B will write the above information into an Excel file and immediately begin sales activities to Doctor G and several other doctors who are highly similar to Doctor G.

[0138] Third Embodiment Company T is a medical device manufacturer and distributor that developed and launched a new digital microscope five years ago, combining the functions of a compact camera with high image quality and 8K resolution with the functions of a microscope. The company sells the device to physicians and dentists (hereinafter collectively referred to as "physicians"). While sales were steady initially, sales performance over the past year has been disappointing. However, the launch of a new product is still some way off, and the company hopes to recover sales by narrowing down the target audience and enhancing sales activities. The medical sales support information processing device of the present invention has accumulated a fair amount of registered data on the activities and results of Company T's digital microscope, so the company will begin using the present invention.

[0139] As described above, the attributes of each hospital facility and its doctors, as well as sales activities and results (usage or prescription) of Company T's digital microscope since its launch, have been registered in advance by the administrator of the medical sales support information processing device of the present invention. In the medical sales support information processing device of the present invention, in step S100 of Fig. 9, whenever this information became available in the past, the control unit 104 of the server 10 acquires information for the doctor table 1012 and the product table 1013 from the management information processing terminal 15 and stores it in each column of the doctor table 1012 and the product table 1013.

[0140] Company T's MR will compile, in an Excel A form, as much as they know about the attributes of each hospital facility and doctor that they are going to approach about Company T's digital microscope, as well as the planned sales activities that they are going to carry out.

[0141] In the medical sales support information processing device of the present invention, in step S110, the control unit 104 of the server 10 acquires information for the activity simulation table 1015 uploaded by the sales representative of Company T from the management information processing terminal and the server 10, and stores the information in each column of the activity simulation table 1015 for carrying out the <optimized sales action calculation process>.

[0142] Specifically, the facility doctor code, facility name, doctor name, attributes (gender, age, affiliated academic society, etc.), planned activities 1 to 13 (future sales activities such as face-to-face interviews, online interviews, email awareness campaigns, and email newsletter distribution), product names, product numbers, other affiliation information that have been stored in advance on the server, and regression coefficients for logistic regression analysis that have been calculated in advance using another technology and appropriately preprocessed are obtained.

[0143] In step S120, the control unit 104 of the server 10 compares the information identifying the doctor stored in the activity simulation table 1015 with explanatory variables regarding the doctor's attributes and performs any necessary updates based on the called application program 1011.

[0144] In step S140, the <Optimization Sales Action Calculation Process> is performed. Prior to this, the control unit 104 of the server 10 first reads the regression coefficients and the like corresponding to the aforementioned similar product numbers in the activity simulation table 1015 stored in the product table 1013, and multiplies the explanatory variables corresponding to attributes 1 to 8 in the activity simulation table 1015 and the explanatory variables corresponding to activities 1 to 13 by β01, β02, β03, β04, β05, β06, β07, β08, β09, β10, β11, β12, β13, β14, β15, β16, β17, β18, β19, β20, and β21 stored in the product table 1013, each with the same product number, and calculates the usage or prescription probability "before" p0 using a logit function in the same way as in embodiment 1, and stores this in the corresponding column of the activity simulation table 1015. The above registration date and time a0 are also stored.

[0145] Furthermore, in step S140, the combination of the usage or prescription probability "after" p1, which is the maximum probability that the target product will be used by the doctor after the <optimized sales action calculation process> is performed, and the value of x related to the activity to obtain it, is calculated and recorded.

[0146] In step S160, the control unit 104 of the server 10 displays on the monitor display via the output unit 1043 the use or prescription probability "after" p1, which is the maximum probability that the target product will be used by the doctor after the <optimized sales action calculation process> is performed, as well as the combination of x values ​​related to the activities required to achieve this and the names of the seven activities with the highest contribution (calculated in step 150).

[0147] According to the information, even if the currently planned activities of the sales representative from Company T who is trying to approach Doctor J, namely three face-to-face interviews, three email promotions, and two mailings of materials, are carried out, the probability that Doctor J will adopt and prescribe Company T's digital microscope will only be 44%, but if six online interviews, seven email promotions, and six mailings of materials are carried out, the probability that Doctor J will adopt and use Company T's digital microscope will rise to 66%, and the above three activities will be the activities that will contribute greatly to this.

[0148] On the other hand, for Dr. K, who the sales representative from Company T is also trying to approach, even if the currently planned activities of 11 online interviews, 15 email promotions, and 9 mailings of materials are carried out, the probability that Dr. K will adopt and prescribe Company T's digital microscope is only expected to be 23%. Even if the activities are further increased to 49 in-person interviews, 29 email promotions, and 19 mailings of materials, which would achieve the maximum predicted probability, the probability that Dr. K will adopt Company T's digital microscope is only expected to increase to 25%.

[0149] The sales manager of Company T collects the daily sales reports submitted by multiple sales representatives at Company T every day and transmits the analysis of the sales reports in a specified Excel FORM (not shown in the drawings) from the customer information processing terminal 40 of the medical sales support information processing device of the present invention. The medical sales support information processing device of the present invention acquires this data and, after carrying out the steps described in the chart for <sales sales report analysis>, is able to display a summary of the sales reports and information such as future improvement measures.

[0150] Company B's sales representative and / or sales manager can then write the above information into an Excel file and immediately put into practice focused sales activities aimed at the selected doctors and new sales strategies devised based on the improvement measures.

[0151] <Actions and Effects> The present invention provides the following actions and effects.

[0152] 1. Reduction of sales costs: Costs can be reduced by eliminating unnecessary sales activities and focusing on effective activities. 2. Optimization of sales resources: By creating optimal sales strategies for each physician, it is possible to efficiently allocate limited sales personnel. 3. Improvement of performance prediction: A data-based probabilistic approach improves the accuracy of sales performance predictions. 4. Strengthening relationships with physicians: It becomes possible to approach physicians in a way that meets their needs, building better relationships.

[0153] <Additional Notes> The matters described in the above embodiments will be added below.

[0154] (Supplementary Note 1) A medical sales support information processing device comprising: an acquisition unit that acquires data including attributes of each doctor, history or schedule of sales promotion activities including sending e-mail newsletters to doctors, notifications of opening of e-mail newsletters, web interviews or face-to-face interviews, or results of sales promotion activities; a storage unit that stores programs and data; a calculation unit that calls up the programs and data and calculates the probability of use or prescription using a statistical method including logistic regression analysis, or calculates recommended sales activities by a logit function using the probability of use or prescription; and an output unit that shows the recommended sales activities calculated by the calculation unit.

[0155] (Supplementary Note 2) The medical sales support information processing device according to Supplementary Note 1, wherein the logistic regression analysis includes penalized logistic regression or multi-class logistic regression.

[0156] (Supplementary Note 3) The medical sales support information processing device according to Supplementary Note 1 or 2, wherein the statistical method includes the use of a machine learning model.

[0157] (Supplementary Note 4) The medical sales support information processing device according to any one of Supplementary Notes 1 to 3, wherein the calculation of the recommended sales activities is performed using a linear or non-linear prediction model.

[0158] (Appendix 5) A medical sales support information processing device as described in any of Appendices 1 to 4, wherein the calculation unit further performs sales promotion activities for doctors and sales promotion activities based on the above-mentioned data using ROI analysis, which quantifies the relationship between investment amount and results and calculates ROI by applying either statistical methods, machine learning models, or regression analysis methods.

[0159] (Appendix 6) A medical sales support information processing device described in any of Appendices 1 to 5, wherein the calculation unit further calculates the similarity between doctors using cosine similarity analysis for doctors with a desired range of usage or prescription probability based on the doctor's attributes, classifies the doctors based on the calculated similarity, and uses it to list the names of doctors within the desired similarity range.

[0160] (Appendix 7) The medical sales support information processing device according to any one of Appendices 1 to 6, wherein the calculation unit further analyzes the sales activity record text using NLP and machine learning models, summarizes it using BERT, TF-IDF, LDA, or a supervised learning algorithm, extracts specific description items, or calculates a recommended strategy by associating it with the aforementioned data.

[0161] (Supplementary Note 8) The medical sales support information processing device according to any one of Supplementary Notes 1 to 7, wherein the recommended sales activities include calculation of the type of sales activity and the amount of increase or decrease.

[0162] (Supplementary Note 9) The medical sales support information processing device according to any one of Supplementary Notes 1 to 8, wherein the output unit outputs or displays an electronic file on a screen.

[0163] (Supplementary Note 10) The medical sales support information processing device according to any one of Supplementary Notes 1 to 9, wherein the output unit is further capable of writing out or displaying on a screen an electronic file including recommended tactics or recommended actions for each sales activity.

[0164] (Supplementary Note 11) The medical sales support information processing device according to any one of Supplementary Notes 1 to 10, wherein the medical sales are sales of pharmaceuticals.

[0165] (Supplementary Note 12) The medical sales support information processing device according to any one of Supplementary Notes 1 to 10, wherein the medical sales are sales of medical equipment.

[0166] (Supplementary Note 13) A computer program for causing the acquisition unit, storage unit, calculation unit, or output unit according to any one of Supplementary Notes 1 to 12 to perform an operation.

[0167] (Supplementary Note 14) A medical sales support method comprising the steps of: acquiring data including attributes for each doctor, a history or schedule of sales promotion activities including sending e-mail newsletters to doctors, notifications of opening of e-mail newsletters, web interviews, or face-to-face interviews, or the results of the sales promotion activities; storing a program and data; calling up the program and data, and calculating the probability of use or prescription using a statistical method including logistic regression analysis, or calculating recommended sales activities with a logit function using the probability of use or prescription; and displaying the recommended sales activities calculated by the calculation unit.

[0168] (Supplementary Note 15) The method according to Supplementary Note 14, comprising the calculation or computation according to any one of Supplements 2 to 8, or the output to an electronic file or the display on a screen according to any one of Supplements 9 to 10.

[0169] (Appendix 16) The medical sales support method according to appendix 14 or 15, wherein the medical sales are sales of pharmaceuticals.

[0170] (Supplementary Note 17) The medical sales support method according to Supplementary Note 14 or 15, wherein the medical sales are sales of medical equipment.

[0171] (Supplementary Note 18) A computer including a memory that is a non-transitory computer-readable medium that stores a computer program, and a processor that executes the computer program, wherein the processor stores data in the memory including attributes for each doctor, a history or schedule of sales promotion activities including sending e-mail newsletters to doctors, email newsletter open notifications, web interviews, or face-to-face interviews, or the results of the sales promotion activities, for supporting medical sales; and the processor, after inputting the data, calls the program and the data, calculates the probability of use or prescription using a statistical method including logistic regression analysis, or calculates recommended sales activities using the probability of use or prescription with a logit function, and outputs the calculated recommended sales activities.

[0172] (Supplementary Note 19) The computer of Supplementary Note 18, wherein the logistic regression analysis includes penalized logistic regression or multi-class logistic regression.

[0173] (Supplementary Note 20) The computer of Supplementary Note 18 or 19, wherein the statistical method includes the use of a machine learning model.

[0174] (Supplementary Note 21) The computer according to any one of Supplementary Notes 18 to 20, wherein the calculation of the recommended sales activities is performed using a linear or non-linear prediction model.

[0175] (Appendix 22) The computer of any of Appendices 18 to 21, characterized in that sales promotion activities for doctors and said data-based sales promotion activities are performed using ROI analysis that quantifies the relationship between investment amount and results and calculates ROI by applying either statistical methods, machine learning models, or regression analysis methods.

[0176] (Appendix 23) A computer described in any of Appendices 18 to 22, characterized in that it uses cosine similarity analysis to calculate the similarity between doctors who have a desired range of usage or prescription probability based on the doctor's attributes, classifies the doctors based on the calculated similarity, and uses it to list the names of doctors within the desired similarity range.

[0177] (Appendix 24) The computer of any of appendices 18 to 23, characterized in that the recorded text of sales activities is analyzed using NLP and machine learning models, and summarized using BERT, TF-IDF, LDA, or supervised learning algorithms, and specific descriptions are extracted or associated with the data to calculate a recommendation strategy.

[0178] (Supplementary Note 25) The computer according to any one of Supplementary Notes 18 to 24, wherein the recommended sales activities include calculation of the type of sales activity and the amount of increase or decrease.

[0179] (Supplementary Note 26) The computer according to any one of Supplements 18 to 25, which is capable of writing out or displaying electronic files on a screen.

[0180] (Supplementary Note 27) A computer according to any one of Supplements 18 to 26, capable of writing out or displaying on a screen an electronic file containing recommended tactics or recommended actions for each sales activity.

[0181] (Supplementary Note 28) The computer according to any one of Supplementary Notes 18 to 27, wherein the medical sales are sales of pharmaceuticals.

[0182] (Supplementary Note 29) The computer according to any one of Supplementary Notes 18 to 27, wherein the medical sales are sales of medical equipment.

[0183] (Supplementary Note 30) A medical sales support method comprising: causing a computer for medical sales support to acquire data including attributes for each doctor, a history or schedule of sales promotion activities including sending e-mail newsletters to doctors, notifications of opening of e-mail newsletters, web interviews, or face-to-face interviews, or the results of said sales promotion activities; storing a program and data; calling up the program and data; calculating the probability of use or prescription using a statistical method including logistic regression analysis; or calculating recommended sales activities using the probability of use or prescription with a logit function; and outputting the calculated recommended sales activities.

[0184] (Appendix 31) A method as described in Appendix 30, which includes causing a computer for medical sales support to perform a calculation or computation as described in any of Appendices 19 to 25, or output to an electronic file or display on a screen as described in any of Appendices 26 to 27.

[0185] (Supplementary Note 32) The medical sales support method according to Supplementary Note 30 or 31, wherein the medical sales are sales of pharmaceuticals.

[0186] (Supplementary Note 33) The medical sales support method according to Supplementary Note 30 or 31, wherein the medical sales are sales of medical equipment.

[0187] (Supplementary Note 34) A non-transitory computer readable medium that stores data in the memory including the attributes of each doctor, the history or schedule of sales promotion activities including sending e-mail newsletters to said doctors, email newsletter open notifications, web interviews or face-to-face interviews, or the results of sales promotion activities, in order to support medical sales; and after inputting the data, a processor calls up a program and the data, calculates the probability of use or prescription using a statistical method including logistic regression analysis, or calculates recommended sales activities by a logit function using the probability of use or prescription, and outputs the calculated recommended sales activities.

[0188] As a result of the above, the present invention can support pharmaceutical companies in realizing efficient sales activities, thereby reducing sales costs and optimizing pharmaceutical costs, thereby contributing to improving the social situation in which lives that could be saved are not being saved.

[0189] Although the embodiments of the present invention have been disclosed above, the present invention is not limited to these and can be modified as appropriate within the scope of the technical idea of ​​the invention.

[0190] This invention can be applied not only to sales activities in the pharmaceutical industry, but also to insurance sales, medical equipment sales, and other industries that require face-to-face sales, and can be used as technology to support the formulation of optimal sales strategies based on data.

[0191] REFERENCE SIGNS LIST 1 Medical sales support information processing device 10 Server 15 Management information processing terminal 20 Customer information processing terminal 30 Customer information processing terminal

Claims

1. A medical sales support information processing device comprising: an acquisition unit that acquires data including the attributes of each doctor, the history or schedule of sales promotion activities including sending e-mail newsletters to said doctors, notifications of e-mail newsletter opening, online interviews or face-to-face interviews, or the results of said sales promotion activities; a storage unit that stores a program and said data; a calculation unit that calls up said program and said data and calculates the probability of use or prescription using a statistical method including logistic regression analysis, or calculates recommended sales activities using the probability of use or prescription with a logit function; and an output unit that indicates the recommended sales activities calculated by said calculation unit.

2. The medical sales support information processing device according to claim 1, wherein the logistic regression analysis includes penalized logistic regression or multi-class logistic regression.

3. The medical sales support information processing device according to claim 1 or 2, wherein the statistical method includes the use of a machine learning model.

4. The medical sales support information processing device according to any one of claims 1 to 3, wherein the calculation of the recommended sales activities is performed using a linear or non-linear prediction model.

5. A medical sales support information processing device according to any one of claims 1 to 4, wherein the calculation unit further performs ROI analysis on the sales promotion activities for the doctors and the sales promotion activities based on the data, quantifying the relationship between investment amount and results and calculating ROI by applying either a statistical method, a machine learning model, or a regression analysis method.

6. A medical sales support information processing device according to any one of claims 1 to 5, wherein the calculation unit further calculates the similarity between doctors who have a usage or prescription probability within a desired range based on the doctor's attributes using cosine similarity analysis, classifies the doctors based on the calculated similarity, and uses it to list the names of doctors within the desired similarity range.

7. The medical sales support information processing device according to any one of claims 1 to 6, wherein the calculation unit further analyzes the sales activity record text using NLP and machine learning models, summarizes it using BERT, TF-IDF, LDA, or a supervised learning algorithm, extracts specific description items, or calculates a recommended strategy by associating it with the data.

8. The medical sales support information processing device according to any one of claims 1 to 7, wherein the recommended sales activities include calculation of the type of sales activity and the amount of increase or decrease.

9. The medical sales support information processing device according to any one of claims 1 to 8, wherein the output unit can write out an electronic file or display it on a screen.

10. A medical sales support information processing device according to any one of claims 1 to 9, wherein the output unit is further capable of writing out or displaying on a screen an electronic file containing recommended tactics or recommended actions for each sales activity.

11. A medical sales support information processing device according to any one of claims 1 to 10, wherein the medical sales are sales of pharmaceuticals.

12. A medical sales support information processing device according to any one of claims 1 to 10, wherein the medical sales are sales of medical equipment.

13. A computer program for causing the acquisition unit, storage unit, calculation unit, or output unit according to any one of claims 1 to 12 to perform its operations.

14. A medical sales support method comprising the steps of: acquiring data including attributes for each doctor, a history or schedule of sales promotion activities including sending e-mail newsletters to said doctors, notifications of e-mail newsletter open status, online interviews, or face-to-face interviews, or the results of said sales promotion activities; storing a program and said data; calling up said program and said data and calculating the probability of use or prescription using a statistical method including logistic regression analysis, or calculating recommended sales activities using the probability of use or prescription with a logit function; and displaying the recommended sales activities calculated by said calculation unit.

15. The method according to claim 14, comprising the calculation or computation according to any one of claims 2 to 8, or the output to an electronic file or the display on a screen according to any one of claims 9 to 10.

16. The medical sales support method according to claim 14 or 15, wherein the medical sales are sales of pharmaceuticals.

17. The medical sales support method according to claim 14 or 15, wherein the medical sales are sales of medical equipment.

18. A computer including a memory that is a non-transitory computer-readable medium for storing a computer program, and a processor that executes the computer program, wherein the processor stores in the memory data for medical sales support, including attributes for each doctor, a history or schedule of sales promotion activities including sending e-mail newsletters to the doctors, notifications of e-mail newsletter open status, online interviews, or face-to-face interviews, or the results of the sales promotion activities, and after inputting the data, the processor calls up the program and the data, calculates the probability of use or prescription using a statistical method including logistic regression analysis, or calculates recommended sales activities using the probability of use or prescription with a logit function, and outputs the calculated recommended sales activities.

19. The computer of claim 18, wherein the logistic regression analysis comprises penalized logistic regression or multi-class logistic regression.

20. The computer of claim 18 or 19, wherein the statistical techniques include the use of machine learning models.

21. The computer according to any one of claims 18 to 20, wherein the calculation of the recommended sales activities is performed using a linear or non-linear prediction model.

22. A computer according to any one of claims 18 to 21, characterized in that the sales promotion activities for the physicians and the sales promotion activities based on the data are carried out using ROI analysis that quantifies the relationship between investment amount and results and calculates ROI by applying either statistical methods, machine learning models, or regression analysis methods.

23. A computer according to any one of claims 18 to 22, characterized in that for doctors who have a desired range of usage or prescription probability based on the attributes of the doctors, cosine similarity analysis is used to calculate the similarity between the doctors, classify the doctors based on the calculated similarity, and use it to list the names of doctors within the desired similarity range.

24. A computer according to any one of claims 18 to 23, characterized in that the sales activity records are analyzed using NLP and machine learning models, and summarized using BERT, TF-IDF, LDA, or supervised learning algorithms, and specific descriptions are extracted or associated with the data to calculate a recommendation strategy.

25. The computer according to any one of claims 18 to 24, wherein the recommended sales activities include calculation of the type of sales activity and the amount of increase or decrease.

26. A computer according to any one of claims 18 to 25, capable of writing out or displaying electronic files on a screen.

27. A computer according to any one of claims 18 to 26, which is capable of writing out or displaying on a screen an electronic file containing the recommended tactics or recommended actions for each of the sales activities.

28. The computer according to any one of claims 18 to 27, wherein the medical sales are sales of pharmaceuticals.

29. The computer of any of claims 18 to 27, wherein the medical sales are sales of medical equipment.

30. A medical sales support method comprising: causing a computer for medical sales support to acquire data including the attributes of each doctor, the history or schedule of sales promotion activities including the sending of e-mail newsletters to said doctors, e-mail newsletter open notifications, online interviews or face-to-face interviews, or the results of said sales promotion activities; storing a program and said data; retrieving said program and said data; calculating the probability of use or prescription using a statistical method including logistic regression analysis; or calculating recommended sales activities using the probability of use or prescription with a logit function; and outputting the calculated recommended sales activities.

31. The method according to claim 30, which includes causing a computer for medical sales support to perform the calculation or computation described in any one of claims 19 to 25, or to output to an electronic file or display on a screen described in any one of claims 26 to 27.

32. The medical sales support method according to claim 30 or 31, wherein the medical sales are sales of pharmaceuticals.

33. The medical sales support method according to claim 30 or 31, wherein the medical sales are sales of medical equipment.

34. A non-transitory computer readable medium that, in order to provide medical sales support, stores in the memory data including the attributes of each doctor, the history or schedule of sales promotion activities including sending e-mail newsletters to said doctors, email newsletter open notifications, web interviews, or face-to-face interviews, or the results of said sales promotion activities, and after inputting the data, the processor calls up the program and the data, calculates the probability of use or prescription using a statistical method including logistic regression analysis, or calculates recommended sales activities using the use or prescription probability with a logit function, and outputs the calculated recommended sales activities.

Images