JP2003345971A - Business operation presenting method and business operation presenting program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for creating an action program of the highest certainty to a customer to such an extent that its cost does not exceed its profit. <P>SOLUTION: Each customer is managed based on customer data including attribute information and channel admission information. A case DB (database) 18 stores case data which are customer data on customers about whom business operation is terminated and which also includes records on commodity sales. A CPU (central processing unit) 10 extracts K-pieces of case data approaching the attribute information and the admission information on a target customer from the DB 18, extracts each group of successful examples of commodities sold by each action from the case data, calculates the ratio of the number of successful examples to K as sales certainty for each group of the successful examples, and creates an action program. Thereafter, the CPU 10 employs an action program in which its cost falls below its profit and the sales certainty is maximized. Otherwise, channels of high cost are removed from channel admission on the customer and the processing is redone. <P>COPYRIGHT: (C)2004,JPO

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an effective sales activity guideline suitable for the attributes of a customer who sells a product such as a financial product (concludes a contract) when the customer receives an inquiry about the product. The present invention relates to a sales activity presentation method for automatically selecting and presenting it to a sales representative.

[0002]

2. Description of the Related Art For example, financial products such as life insurance and fire insurance are rarely designated and purchased by seeming customers at the counter of a financial institution such as an insurance company. When a customer with a customer contacts a financial institution, the sales representative of the financial institution asks for the customer's contact information (or when the customer visits the financial institution's web page). , And enter the contact information of the customer), and at a later date, the sales representative will perform business activities such as sending materials, explaining visits, and contacting the customer by telephone or e-mail based on the contact information. By repeating
At last, you can purchase (conclude a contract).

As described above, in order to sell this kind of product, one or more sales activities must be performed. And since each sales activity requires a cost,
Inefficient business activities should be avoided as much as possible. However, what kind of sales activities are efficient differs for each customer and is not constant.

[0004] Therefore, the sales representative must provide the most efficient (ie, most reliably linked to sales) action program (ie, the combination of channels, content, and timing of one or more sales activities) for each customer. ) Needs to be selected based on one's own experience while referring to the attributes (address, gender, age, family structure, etc.) of the customer.

[0005]

However, such a method of selecting a sales activity largely depends on the personal experience and skills of each salesperson. For this reason, inexperienced or inexperienced salespeople may repeat inefficient sales activities or conduct multiple approaches to customers who can sell products with only one effective sales activity, depending on the way of sales activities. May fail to sell products as a result of performing inappropriate sales activities for customers who need them, or conduct sales activities of a type that costs more than expected revenue from product sales. As a result, they often cannot contribute to the interests of the financial institution as a whole. It would take a lot of money and time to gain experience with such inexperienced or inexperienced salespeople to maintain and further improve the financial institution's overall profits.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and the problem is that the sales target customers can be most effectively provided that the cost does not exceed the profit expected from the conclusion of the sales target contract. A sales activity presentation method for creating a highly accurate action program (that is, a combination of channels, contents, and timings of one or more sales activities) using a computer, and presenting the same to a sales representative; and Another object of the present invention is to provide a business activity presentation program which is a computer program for causing a computer to execute such a business activity presentation method.

[0007]

According to the business activity presentation method and the business activity presentation program of the present invention devised to solve the above-mentioned problem, the computer processing device allows any one of the customers to send the information to the attribute thereof. When acquired, N (where N is an arbitrary natural number) of action programs are extracted by extracting a predetermined number of case data approximate to the attribute information from a database including the attribute information and the sales activity information of each customer. And an adoption process of adopting one of the N action programs created by the creation process that has a cost lower than the profit. If one action program cannot be adopted by the adoption process, the processing device resets the conditions for the customer and re-executes the processes after the creation process. In this way, the processing device outputs one action program finally adopted by the adoption process from the output device.

According to the present invention in which each process is executed as described above, the processing apparatus stores a predetermined number of case data approximate to the attribute information of the sales target customer into a case where the case data which is a record of past sales is stored. Extract from database. Then, the processing device creates N action programs based on the predetermined number of example data, and among the N action programs, adopts an action program whose cost is lower than profit and whose ratio is the maximum. I do. However, when the corresponding action program cannot be adopted, that is, when the cost of each action program is equal to or more than the profit, the processing device resets the conditions for the business target customer and starts the creation process. Start over. Therefore, according to the present invention, an action program with the highest probability is created for a sales target customer and presented to a sales representative within a range in which the cost does not exceed the profit expected from the conclusion of the sales target contract. Can be.

[0009]

Embodiments of the present invention will be described below with reference to the drawings. In the following embodiment, the business activity presentation method according to the present invention is applied for the sale of insurance products as financial products (conclusion of an insurance contract as a business target contract) by an insurance company as a financial institution. is there. In the following description, “contact” means access from a customer to an insurance company (access to a website for inquiries, requesting materials, etc., fax transmission of an inquiry form, telephone call to customer service, sales Etc.), and “action” refers to sales activities (sales of materials such as catalogs and quotes, explanations, greetings, etc.) from the sales staff of the insurance company to customers. Mail, fax, telephone, e-mail, visit, etc.). Further, "channel" means the means or medium used for the contact or action (specifically, website access, telephone, fax, e-mail, visit, etc.), and "timing" means the time from the contact or the previous action. Means the time until the next action. (System Configuration) FIG. 1 is a block diagram showing a schematic configuration of a computer network constructed for implementing the business activity presentation method in an insurance company. As shown in FIG. 1, the computer network includes an insurance company system 1 and a user terminal 2 which can transmit and receive data to and from each other via the Internet N. The insurance company system 1 is usually composed of one host computer and a plurality of terminals installed at the head office and each sales office of the insurance company, respectively. Here, in order to simplify the description, The description will be made assuming that the computer is one computer having all the functions.

Although only one user terminal 2 is shown in FIG. 1, there are actually a plurality of user terminals.
That is, the user terminal 2 refers to all computers that can connect to the Internet N and have a web browsing function.

The insurance company system 1 comprises a CPU 10, a communication adapter 11, and a RAM 1 connected to each other by a bus.
2, a hard disk 13, an input device 30, and a display 31. Of these, CPU1
Reference numeral 0 denotes a central processing unit that controls the entire insurance company system 1. The RAM 12 is a main storage device in which a work area is expanded when the CPU 10 executes various processes. The communication adapter 11 is an output device that interfaces with a line on which the Internet N is established (in some cases, an Internet backbone). The input device 30 is a pointing device such as a keyboard and a mouse for inputting various data to the CPU 10. Display 3
Reference numeral 1 denotes a display device that displays an image based on screen data generated by the CPU 10.

The hard disk 13 is a storage device for storing various programs and various data read and executed by the CPU 20. The data stored in the hard disk 13 includes screen data (hypertext) (not shown) for displaying various screens for a browser program (not shown) executed on the user terminal 2 and will be described later. Temporary file 17, case DB
(Data Base: database) 18, Contact CIF
(Customer Information File) 1
9, personal DB 20, primary discrimination parameter table 21, standardization table 22, cluster center table 2
3, a cluster-specific parameter table 24, a reference table 25, a rule DB 26, a process-specific channel cost DB 27, an action program generator file 28, and a product DB 29 are included. The programs stored in the hard disk include, besides the operating system 14, a WWW server program and some CGI (Common Gateway Interface) programs or Java (US Sun Microsys in the United States and other countries).
A customer reception server program 15 composed of a servlet (trademark or registered trademark of tems) and an action program creation program 16 are included.

Hereinafter, these customer reception server programs 1
5 and the contents of the processing executed by the CPU 10 by the action program creation program 16 together with the contents of the data 17 to 29 described above and the contents of the screen displayed on the display 31 based on the screen data generated by the CPU 10, This will be described in detail. <Processing by Customer Reception Server Program> FIG. 2 is a flowchart showing processing executed by the CPU 10 according to the customer reception server program 15. This process is executed when an HTTP request message specifying a predetermined URL is received from the user terminal 2 operated by any user (potential customer).

In the first step S001 after the processing by the customer reception server program 15 starts, the CPU 10
Transmits HTML data for displaying the customer attribute information input screen as shown in FIG. 11 to the requesting user terminal 2. In the customer attribute information input screen, a list box 51 for selecting an insurance product for which the user (potential customer) intends to request data from among many insurance products sold by the insurance company, a name of the user, , An option button 53 for selecting gender, a list box group 54 for selecting a date of birth, a text box 55 for inputting a zip code, and a home address. Text box 56, a text box 57 for inputting a home telephone number, an option button 58 for selecting married or unmarried, and an option button 59 for selecting whether or not to be married if unmarried. , Select one or more channels that the user is authorized to use as a channel for action on himself Radio button 60, and "OK" button 61 of the eye, are included. HT for displaying the customer attribute information input screen
The ML data includes a setting for transmitting the information set in each of the items 51 to 60 to the customer reception server program 15 when the “OK” button 61 in the customer attribute information input screen is clicked on the user terminal 2. ,It has been incorporated.

In the next S002, the CPU 10
Wait for the information set in each of the items 51 to 60 in the customer attribute information input screen to be transmitted from the user terminal 2 that has transmitted the HTML data in 1. Then, upon receiving the information, the CPU 10 advances the processing to S003.

In S003, the CPU 10 sets an HT for displaying a psychological characteristic information input screen as shown in FIG.
The ML data is transmitted to the requesting user terminal 2. A large number (16 in the figure) of psychological test questions are listed on this psychological characteristic information input screen, and for each psychological test question, the answer is "No", "Slightly yes", "Pretty much" And an option button for selecting any one of “very good”. Here, for example, “CUBIC (trademark)” manufactured by AGP Inc. is used as the psychological test question. Further, the psychological information input screen includes an “OK” button 62. The HTML data for displaying the psychological characteristic information input screen includes the “OK” button 6 on the user terminal 2.
When "2" is clicked, the setting for transmitting the answer set to each option button for each psychological test question to the customer reception server program 15 is incorporated.

In the next S004, the CPU 10 executes the processing in S00.
In step 3, the user terminal 2 that has transmitted the HTML data waits for an answer for each psychological test question to be sent. Then, upon receiving the answer, the CPU 10 proceeds to S00.
Proceed to 5.

In S005, the CPU 10 transmits HTML data for displaying a killer question input screen as shown in FIG. 13 to the requesting user terminal 2. The killer question input screen includes a list box 63 for the user (potential customer) to select the purpose of requesting the data, an option button 64 for selecting whether the user is going to take out insurance, A text box group 65 for inputting the planned date when insurance is planned, and a plurality of questions prepared to determine whether or not the user has a possibility of purchasing insurance (purchasing insurance products). An option button 66 for selecting an answer for each of the (killer questions)
include. Further, this killer question input screen includes an “OK” button 67. Also, an HT for displaying this killer question input screen.
The ML data includes a setting for transmitting the information set in each of the items 63 to 66 to the customer reception server program 15 when the “OK” button 67 in the killer question input screen is clicked on the user terminal 2. It has been incorporated.

At the next step S006, the CPU 10
Wait for the information set in each of the items 63 to 66 in the killer question input screen to be transmitted from the user terminal 2 that has transmitted the HTML data in 5. Then, upon receiving the information, the CPU 10 advances the processing to S007.

In S007, the CPU 10 assigns a unique customer number to the customer.

In the next S008, the CPU 10 executes
2, the information received in S004 and S006, respectively,
It is stored in the temporary file 17 in the hard disk 13 in association with the customer number given in 007. When S008 is completed, the CPU 10 ends the processing by the customer reception server program 15. <Processing by Action Program Creation Program> FIGS. 3 to 5 show the CPU 10 according to an action program creation program 16 as a sales activity presentation program.
6 is a flowchart showing a process executed by the user. In this process, after the salesperson receives a contact (contact using a channel other than the website access, such as telephone, fax, e-mail, and visit, etc.) from the user (potential customer), the salesperson enters the input device. The operation starts from S101 by inputting a predetermined command by operating 30. This process is automatically started from S108 for each data of each customer stored in the temporary file 17 by the badge process.

In S101, the CPU 10 displays a customer attribute information input screen shown in FIG. This customer attribute information input screen has the same contents as those described in S001.

In the next step S102, the CPU 10 determines the items 51 to 6 on the customer attribute information input screen based on the information obtained by the sales representative from the user (potential customer).
While the information is set to 0, the process waits for the “OK” button 61 to be clicked. When the “OK” button 61 is clicked in a state where each information is set, the CPU 1
0 takes in the information set in each of the items 51 to 60, and advances the processing to S103.

In S103, the CPU 10 displays a psychological characteristic information input screen shown in FIG. This psychological identification information input screen has the same contents as those described in S003, and a “Skip” button 68
Contains.

In the next step S104, the CPU 10 clicks the "OK" button 62 with each check button set based on the answer obtained by the sales representative from the user (potential customer), or Waits for the sales representative to click the “Skip” button 68 because he was unable to hear the “. And
When the “OK” button 62 is clicked, the CPU 10
Fetches the answer of each psychological test question set to each option button, and advances the processing to S105. Also,
When the “Skip” button 68 is clicked, the CPU 1
If it is 0, the process proceeds to S105.

In S105, the CPU 10 displays a killer question input screen shown in FIG. This killer question input screen is S0
It has the same contents as those described in 05.

In the next step S106, the CPU 10 determines each item 6 on the killer question input screen based on the information obtained by the sales representative from the user (potential customer).
With the information set in 3-66, the "OK" button 6
Wait for 7 to be clicked. When the “OK” button 67 is clicked in a state where each information is set, C
The PU 10 fetches the information set in each of the items 63 to 66, and advances the processing to S107.

In S107, the CPU 10 assigns a unique customer number to the customer. After the completion of S107,
The CPU 10 advances the processing to S109.

On the other hand, in S 108, the CPU 10 reads information on the customer to be processed from the temporary storage file 17 onto the RAM 12. After the completion of S108, the CPU 1
If "0", the process proceeds to S109.

In S109, the CPU 10 checks whether or not the information about the customer to be processed includes psychological characteristic information (that is, information set on the psychological characteristic information input screen in FIG. 12). This process starts from S101, and the “Skip” button 68
Is determined to have been clicked, the information about the processing target customer does not include the psychological characteristic information, so the process proceeds from S109 to S111, otherwise, the processing for the processing target customer is performed. Since the information includes the psychological characteristic information, the process proceeds from S109 to S110.

At S110, CPU 10 executes a customer identification determining process. FIG. 6 is a flowchart showing the customer identification determining process subroutine executed in S110. In the first step S201 after entering this subroutine, as shown in FIG. 14, the CPU 10 responds to each psychological test question included in the psychological characteristic information and according to the score expansion rule defined in advance, according to each personality characteristic. Each time a score is awarded. In FIG. 14, the answer "1" indicates the answer "No" on the psychological characteristic information input screen, and similarly, the answer "2" indicates "Slightly so", and the answer "3" indicates "Somewhat so". "4" indicates "very much." And according to the scoring rules,
Those who answered "No" to the question "It does not bother to move actively" have personality traits "inner-closed" and "bearness" of "2", respectively, and have a personality trait of "2." Since the levels of “mood”, “self-esteem” and “cautious” are “1”, and the level of the other personality traits is “0”, a score corresponding to the level is given. Therefore, if another answer is given to the same question, a different score development will be made. When the score development based on the answers to all the psychological test questions included in the psychological characteristic information is completed, the CPU 10 advances the processing to S202.

In S202, the CPU 10 calculates the total (score) of the scores given to each personality trait as a result of the scoring in S201. Then, the personality characteristic having the largest calculated score is specified as the main personality of the processing target customer.

In the next S203, the CPU 10 proceeds to S20.
In step 2, a dummy variable “1” is assigned to the personality trait specified as the main personality, and a dummy variable “0” is assigned to all other personality traits. After the completion of S203, the CPU 10
Ends the customer characteristic determination processing subroutine, and proceeds from S110 to S112 in the main routine of FIG.
Proceed to.

On the other hand, in S111, the CPU 10 determines the address, age, gender, marriage among the customer attribute information included in the information on the processing target customer (that is, the information input on the customer attribute information input screen in FIG. 11). The most similar case data is extracted from the case DB 18 using “whether married” or “unmarried” as a key.

The case data stored in the case DB 18 is a record of business activities of customers who have been subjected to business activities in the past using the insurance company system 1, and each customer constituting the personal DB 20 described later. Has the same data structure as the data (in other words, the customer data for a customer whose sales activities have been concluded or abandoned is recorded with information about the concluded or abandoned business contract). , Is transferred from the personal DB 20 to the case DB 18 to become case data.) Specifically, as shown in FIG. 15, the case data is “customer number (No.)”, “customer attribute information” and “psychological characteristic information” as attribute information, and a record of a sales contract. It consists of items of "contracted information", "channel permission" as channel permission information, "floating layer information", "contact history", and "action program history" as history information. "Psychological characteristic information" in this case data includes S201,
Score calculated by performing S202,
Also, a dummy variable calculated by executing S203 is stored. In addition, the "contracted information" stores information for specifying an insurance product for which a contract has been concluded based on the customer data that was the basis of this case data. In the case data, it is blank. Further, of the information input on the customer attribute information input screen, information for selecting a channel available for information provision is stored in “channel allowance” in the case data, and other information is
Stored in “customer attribute information”. The “contact history” includes the first contact, the second contact,
For each of the third contacts, the timing (always "0" for the first contact), the channel and the contents (material request, estimate request, etc.) can be stored. Also, "Action program history"
As for the first action, the second action, and the third action, the timing, channel, and content of each action can be stored.

Thereafter, the CPU 10 combines the psychological characteristic information in the case data extracted from the case DB 18 with the information on the customer to be processed. After completion of S111, the CPU
10, the process proceeds to S112.

In S112, the CPU 10 determines whether or not there has been a contact from the processing target customer in the past based on whether or not the contact CIF 19 including the customer number in the information on the processing target customer is stored in the hard disk 13. Check. This contact CIF1
Reference numeral 9 denotes a file created in the final stage (S126) of processing by the action program creation program 16, and as shown in FIG.
o. )), “Customer attribute information” and “contact history”. This "customer attribute information" is the same as that of the case data, and the "contact history"
Same as the case data. When the corresponding contact CIF 19 is stored, the CPU 1
In step S113, the process reads the contact CIF 19 into the RAM 12 and combines the read contact CIF 19 with the data of the processing target customer, and advances the processing to step S114. On the other hand, when the corresponding contact CIF 19 is not stored, the CPU 10 advances the processing to S114 as it is.

In S114, the CPU 10 determines the information of the customer to be processed (the psychological characteristic information converted into the score and the dummy variable in S110 or the psychological characteristic information combined in S111, and the contact CIF1 read in S113).
8 is edited into a customer data format as shown in FIG.
Register at 20. If the customer data has already been registered in the personal DB 20 for the customer to be processed (that is, if the contact from the customer to be processed is the second time or later), the CPU 10 registers the new data with the new data. Add customer data that has been added.

In the next step S115, the CPU 10 executes a floating layer primary discrimination process for finding out customers who are unlikely to sell insurance products (conclude a contract). FIG. 7 shows this S
It is a flowchart which shows the floating layer primary determination processing subroutine performed at 115. In the first step S301 after entering this subroutine, the CPU 10 converts the answer to the question about marriage and the answer to the killer question included in the customer attribute information in the data of the processing target customer into dummy variables. That is, as shown in FIG. 18, for all the answers that can be selected for each question (question), a dummy variable “1” is assigned to the answer that the processing target customer has actually selected as an answer, and the answer is selected as an answer. A dummy variable “0” is assigned to the one not performed.

In the next step S302, the CPU 10 refers to the primary discrimination parameter table 21 shown in FIG. 19, which defines parameter values for all answers that can be selected for each question (question). For each of the available answers to (question)
The product of the dummy variable assigned in S301 and the parameter value defined in the primary discrimination parameter table 21 is calculated, and the total sum t of all products is calculated. Then, the floating layer accuracy is calculated by executing the following function (1) on the calculated sum t.

Floating layer accuracy = e ＾ t / (1 + e ＾ t) (where e is the base of natural logarithm) (1) In the next S303, the CPU 10 determines that the floating layer accuracy calculated in S302 is 0.5 Check if it is over. If the floating layer accuracy exceeds 0.5, the CPU 10 ends the floating layer primary determination processing subroutine as it is, and if it is 0.5 or less, the CPU 10 sets the parting layer determination flag in S304. Thereafter, the floating layer primary determination processing subroutine ends.

Referring back to FIG. 4, after completion of S115, the CPU 1
If it is 0, it is checked in S116 whether or not the parting layer discrimination flag is set. And the CPU 10
Indicates that if the parting layer discrimination flag is set,
The process advances to S117, and if the parting layer discrimination flag is not set, the process advances to S117.

In S117, the CPU 10 determines whether the floating layer
The next determination process is performed. FIG. 8 is a flowchart showing the floating layer secondary determination processing subroutine executed in S117. First S4 after entering this subroutine
In 01, the CPU 10 performs advance preparations. That is, CP
U10 refers to the standardization table 22 to standardize the gender, age, and unmarried status of the customer to be processed. That is, here, the gender of the processing target customer is the parameter value Z1.
(Male: 1, female: 2), and the age is a parameter value Z2 (20s: 20, 30s: 30, 40s: 40,
..), And the unmarried person is converted into a parameter value Z3 (married: 1, single: 2). The standardization table 22 shown in FIG. 20 stores the average (μ) and standard deviation (δ) of the gender, age, and unmarried parameters calculated in advance based on all the case data registered in the case database 18. ) Are listed. C
PU10, for each gender, age, unmarried,
By calculating (Z-μ) / δ, the parameters Z1 to Z3 are standardized.

In the next S402, the CPU 10 proceeds to S40.
1. The customer attribute of the processing target customer indicated by each parameter value standardized in 1 and the cluster center table 23
To each cluster (cluster distance)
Is calculated. Since the calculation of the cluster distance is a well-known statistical processing method, the details thereof will not be described, but the cluster center table 23 shown in FIG.
In the example, four clusters (cluster no.
1-4) are prepared.

In the next S403, the CPU 10 proceeds to S40.
The cluster with the shortest cluster distance from the customer attribute calculated in 2 is specified as the cluster to which the customer belongs.

In the next S404, the CPU 10 executes
As shown in FIG.
The dummy variable of the psychological characteristic information in the customer data stored in 0, and the dummy variable when the action program history exists in the customer data (specific values of the timing, channel, and content of each action are The value “1” assigned to the registered one, the value “0” assigned to the unregistered specific one), and the dummy variable of the contact history in the contact CIF 19 read in S113 (each time). Of the contact timing, channel, and content of the contact (1) assigned to a contact whose specific value is registered, and value (0) assigned to a contact whose specific value is not registered.

In the next S405, based on the result of the combination in S404 and the parameter table 21 for each cluster,
The score Y is calculated. As shown in FIG. 23, this cluster-specific parameter table includes, for each cluster, each personality trait, timing for each action in the action program history, channel, contents, timing for each contact in the contact history, channel, This is a table in which parameter values are defined for each of the contents. The CPU 10 reads out all the parameters of the belonging class specified in S403 from the cluster-specific parameter table 21. And S40
Each dummy variable included in the combination result in step 4 is multiplied by a corresponding parameter among all parameters read from the cluster-specific parameter table 21. Finally, the sum of the products calculated for each dummy variable is calculated as a score Y.

In the next step S406, the CPU 10 proceeds to step S40.
The score Y calculated in 5 is compared with each reference value (b1 = 3, b2 = 1) registered in the reference table shown in FIG.

In the next step S407, the CPU 10 determines the score Y
Is checked whether or not exceeds the reference value b1 (= 3). When the score Y exceeds the reference value b1 (= 3), the CPU 10 determines that there is a high possibility that the customer to be processed purchases an insurance product (concludes a contract) regardless of business activities. After the recognition and the setting of the prospective layer discrimination flag in S408, the floating layer secondary discrimination processing ends. On the other hand, when the score Y is less than or equal to the reference value b1 (= 3), the CPU 10 advances the processing to S409.

In S409, the CPU 10 checks whether or not the score Y is equal to or more than the reference value b2 (= 1) and equal to or less than the reference value b1 (= 3). Then, the score Y becomes the reference value b
If not less than 2 (= 1) and not more than the reference value b1 (= 3),
The CPU 10 recognizes that the customer to be processed may purchase an insurance product (conclude a contract) depending on the business activities, and sets the floating layer discrimination flag in S410. The next determination processing ends. On the other hand, when the score Y is less than the reference value b2 (= 1), the CPU 10 recognizes that there is no possibility that the customer purchases an insurance product (concludes a contract), and the CPU 10 determines in S411. After setting the parting layer discrimination flag,
The determination processing ends.

Returning to FIG. 4, after completion of S117, the CPU 1
If it is 0, it is checked in step S118 whether or not the parting layer discrimination flag is set. And the CPU 10
Indicates that if the parting layer discrimination flag is set,
The process proceeds to 121, and if the parting layer discrimination flag is not set, the process proceeds to S119.

In S121, the CPU 10 displays on the display 31 that the customer belongs to the parting layer. In the next S122, the floating layer accuracy calculated in S115 (S302) and S115 (S304) or S1
17 (S411),
It is registered in the “floating layer information” included in the customer data of the customer in the personal DB 20. Upon completing S122, the CPU 10 advances the processing to S126.

On the other hand, in S119, the CPU 10 displays on the display 31 that the customer belongs to the prospect layer or the floating layer.
Display on the screen. In the next S120, S115 (S3
02) and the floating layer accuracy calculated in S117 (S4
08) The estimated layer discrimination flag set in step S08 or S117
The floating layer discrimination flag set in (S410) is added to the “floating layer information” included in the customer data for the customer in the personal DB 20, and the belonging cluster specified in S117 (S403) is set to the customer for the customer. Register in "customer attribute information" included in the data.

In the next S123, the CPU 10 executes an action program creation process. 9 and 10
Is a flowchart showing an action program creation processing subroutine executed in S123. In the first step S501 after entering this subroutine, the CPU 1
0 sets the proposal rule initial value based on the customer data for the processing target customer registered in the personal DB 20 and the contact CIF 19 read in S113. As shown in FIG. 25, this proposal rule initial value includes a customer number (No.) included in “customer attribute information” in the customer data of the processing target customer, a main character of the psychological characteristic information, and a contact CIF 19. "Contact history" in the "Action program history" and "Channel allowed" in the customer data for the customer to be processed
Includes Furthermore, this proposal rule initial value includes
A variable in which the timing, channel and content of each contact recorded in the "contact history" are parameterized into three-digit numbers, "action program history"
For each action recorded in, a variable that parameterized the timing, channel, and contents into three-digit numbers, and whether or not OK set for each channel recorded in “Channel Allowed” A "contact status" consisting of parameterized variables is included.

In the next step S502, the CPU 10 proceeds to step S50.
The proposed rule initial value set in step 1 is compared with each record registered in the rule DB 26. This rule DB
As shown in FIG. 26, each record registered in 26 includes customer attribute information, contact status, contact history, action program history, and channel of a customer who has purchased an insurance product (concluded a contract) in the past. Includes tolerance. Further, each record includes a "channel rule" item relating to the channel of each action, a "content rule" item relating to the content, and an "approach rule" item relating to the timing.
In the item of “channel rule”, the channel used for the first action when the insurance product can be sold at the time of executing the first action (coincides with the first channel in the “action program history”:
Channels with priority 1), channels that were next candidates (channels with priority 2), channels that were candidates next (channels with priority 3), and the case where the insurance product could be sold at the time of executing the second action Channels of priority 1 to 3 used for the first action, channels of priority 1 to 3 used for the second action, 1 in the case where the insurance product could be sold at the time of executing the third action Priority 1 to 3 used for the second action
Channel, priority 1 used for second action
Columns for writing channels 1 to 3 and channels of priority 1 to 3 used for the third action are included. Similarly, the item of the “content rule” includes the content used for the first action when the insurance product was sold at the time of executing the first action (the first content in the “action program history”). Matches: priority 1 content), next candidate content (priority 2 content), next candidate content (priority 3 content), insurance product can be sold when the second action is executed In this case, the contents of the priorities 1 to 3 used for the first action, the contents of the priorities 1 to 3 used for the second action, and the insurance product could be sold at the time of executing the third action Contents of priorities 1 to 3 used for the first action, contents of priorities 1 to 3 used for the second action, 3
A column for writing the contents of the priorities 1 to 3 used for the second action is included. Similarly, the “approach rule” item includes the first item in the case where the insurance product could be sold at the time of executing the first action.
The timing used for the second action (coincides with the first timing in the "action program history": the timing of priority 1), the timing of the next candidate (timing of priority 2), the timing of the next candidate (priority) Timing 3), the timings of the priorities 1 to 3 used for the first action when the insurance product could be sold at the time of executing the second action,
Timing of priority 1 to 3 used for the second action, timing of priority 1 to 3 used for the first action when the insurance product can be sold at the time of executing the third action, 2 Columns for writing the timings of priority 1 to 3 used for the third action and the timings of priority 1 to 3 used for the third action are included.

Then, the CPU 10 sells the insurance product (enrollment contract) at the time when the first action is executed in the record in which the initial attribute of the proposal rule set in S501 matches the customer attribute and the contact status is close. Record indicating that the insurance product was successfully sold, the record indicating that the insurance product was sold when the second action was executed (contract was signed), and the insurance product was sold when the third action was executed ( Records indicating the case where the subscription contract has been concluded) are specified. Then, from the identified records, the rules (channel rules, content rules, approach rules) used for the first action when the insurance product can be sold (contracted) when the first action is executed. ), The first and second rules when the insurance product can be sold (subscription contract) when the second action is executed, and the insurance product is sold (subscription 1) to 3 when it is completed
The second rule is extracted as shown in FIG.
If there is no applicable rule, the CPU 10
Leave the rules unextracted.

In the next step S503, the CPU 10 determines in step S50.
By combining the rules extracted in S502 with the proposed rule initial values set in step 1, a proposed rule search value as shown in FIG. 28 is created. If there is a rule that cannot be extracted in S503, the CPU 10 blanks the rule column in the rule search value.

In the next S504, the CPU 10
All the case data stored in B18 are plotted in a multidimensional space whose axes are the values of each item constituting the case data. The scale of each axis is set by reflecting the weight of the corresponding variable (when the weight is increased, the scale is set wide, and when the weight is reduced, the scale is set narrow).

In the next step S505, the CPU 10 determines in step S50.
The proposed rule search value created in 3 is plotted in the same multidimensional space.

In the next step S506, the CPU 10 sets a value obtained by adding 1 to the number of times of execution of the action included in the "action program history" in the customer data for the customer to be processed, as a variable j. Next, the CPU 10 executes the loop processing of S507 to S512 in order to create an action program for selling an insurance product (concluding a contract) at the time of executing the j-th to third actions, respectively. First step S507 after entering this loop
Then, the CPU 10 calculates k pieces of case data in order from the shortest distance to the processing target customer in the multidimensional space, which is calculated based on the value of each item from the customer attribute to the j-th action. Is extracted from the case DB 18 (corresponding to a case extraction process). Here, k is a constant indicating a predetermined number, and is set to, for example, 200.

In the next step S508, the CPU 10 executes
Case data extracted in step 7 (example of insurance product sold at the time of execution of the jth action, example of insurance product sold at the time of subsequent action execution, example of insurance product not being sold finally ) Is further extracted from the case data (success example group) when the insurance product is sold (subscription contract is concluded) at the time of the j-th action execution.

In the next step S509, the CPU 10 determines in step S50
The channels, timings, and contents in each of the first to jth actions included in the case data extracted in step 8 are totaled. At this time, the rule (channel rule, content rule, approach rule) for each time defined in the proposed rule search value is referred to, and a higher priority is given to a larger point. Then, each time, the channel, the timing, and the content having the maximum assigned points are selected. The CPU 10 creates the channel, timing, and content for each of the first to jth actions selected as described above as an action program.

In the next step S510, the CPU 10 proceeds to step S50.
A purchase probability (ratio) is calculated by executing the following function (2) based on the number of case data (the number of subscriber data) extracted in 8.

[0064]     Purchase accuracy (%) = number of subscriber data / k × 100 (2) These steps S509 and S510 correspond to the creation processing.

In the next step S511, the CPU 10 sets the variable j
Check if is equal to 3. And still the variable j
Does not reach 3, the variable j is incremented by one in S512, and the process returns to S507. On the other hand, when the variable j reaches 3 as a result of repeating the processing loop from S507 to S512, the CPU 10
The process proceeds from S511 to S513. FIG. 29 shows an example of 3 (= N) action programs created when the first contact is made (that is, when the number of executed actions = 0) (the insurance product is sold at the time of the first action execution). An action program “a” that is likely to be able to be performed, an action program “b” that is likely to be able to sell an insurance product at the time of executing the second action, and an action program that is likely to be able to sell the insurance product at the time of executing the third action 5 shows the program “c”).

In S513, the CPU 10 executes S5
At 07, the average premium (average annual fee) and the average number of years of maintenance of the requested insurance product are calculated based on k case data extracted from the case DB 18 respectively.

In the next S514, the CPU 10 proceeds to S51.
The profit is calculated by multiplying the average premium calculated in step 3 by the average maintenance period.

In the next step S515, the CPU 10 refers to the process-specific channel-specific cost DB 27 for each of the action programs created in the respective S509, and
The total cost required to execute the action program is calculated. Cost DB2 by process and channel
FIG. 7 shows a list of costs corresponding to combinations of each channel and each content, as shown in FIG.
The CPU 10 reads out the cost of each action included in each action program from the process-specific channel-specific cost DB 27, and calculates the total as the total cost required to execute the action program.

In the next S516, the CPU 10 proceeds to S51.
It is checked whether the profit calculated in step 4 exceeds the cost calculated in step S515 for all the action programs, and whether the purchase accuracy calculated in step S510 exceeds a predetermined threshold. Then, the profit calculated in S514 is calculated for all the action programs in S51.
If the purchase probability calculated in S510 exceeds the cost calculated in S510 and the purchase probability calculated in S510 exceeds a predetermined threshold, the CPU 10 determines in S517 that the purchase probability calculated in S510 is the maximum. To terminate the action program creation processing (corresponding to the recruitment processing).

On the other hand, if the profit calculated in S514 is equal to or less than the cost calculated for any of the action programs, or if any of the action program purchase probabilities is equal to or less than the threshold, the CPU 10 executes the processing. From S516 to S518. In S518, the CP
U10 checks whether there is at least one action program in which the profit calculated in S514 exceeds the cost calculated in S514. If there is no action program in which the calculated profit exceeds the cost, the CPU 10 advances the processing to S521.

On the other hand, if the profit calculated in S514 exceeds the cost calculated for some of the action programs, the CPU 10 calculates the profit calculated in S514 in S519 in S515. It is checked whether there is any action program whose cost is higher than the cost and the purchase accuracy calculated in S510 is higher than the threshold. Then, if there is an action program in which the calculated profit exceeds the cost and the purchase probability exceeds the threshold, the CPU 10 determines whether the
At 520, among the action programs whose calculated profits exceed the cost and whose purchase probabilities exceed the threshold value, those having the highest purchase probabilities are adopted, and the action program generation processing is ended (the adoption processing is terminated). Equivalent).

On the other hand, if there is no action program in which the calculated profit exceeds the cost and the purchase probability exceeds the threshold, the CPU 10 advances the processing from S519 to S521.

In S521, the CPU 10 executes S504-
It is checked whether or not the loop processing in S519 has been completed for the first time. Then, when the loop processing is completed for the first time, in S532, the CPU 10 determines, among the channels for which “OK” included in the “channel allowance” in the proposal rule search value created in S503 is set. , The value of the highest cost is changed to “NO”, and the corresponding channel flag included in the “contact status” is reset. After the completion of S532, the CPU
10 returns the process to 504 and starts the action program creation again. On the other hand, if the loop processing has been completed for the second time, the CPU 10 proceeds from S521 to S521.
Proceed to 522.

In S522, the CPU 10 adds the value obtained by adding “1” to the average maintenance years calculated in S513, to the variable m.
Substitute for

In next step S523, the CPU 10 subtracts the age included in the “customer attribute” in the customer data of the processing target customer from 60 (that is, the value of the customer is 60).
(The remaining years until the child turns old) is substituted for the variable n. S5
After the completion of Step 23, the CPU 10 advances the processing to S524. Note that the numerical value “60” used here is a constant indicating the payout expiration year, and thus a numerical value other than “60” may be used.

In S524, the CPU 10 multiplies the average premium calculated in S513 by a variable m,
Calculate revenue.

In the next S525, the CPU 10 proceeds to S52.
It is checked whether the profit calculated in step 4 exceeds the cost calculated in step S515 for all action programs. Then, when the calculated profit exceeds the cost calculated for each of the action programs, the CPU 10 proceeds to S526 and proceeds to S51.
The action program whose purchase accuracy calculated at 0 is the maximum is adopted, and the action program creation processing ends (corresponding to the adoption processing).

On the other hand, if the profit calculated in S524 is equal to or less than the cost calculated for any of the action programs, the CPU 10 proceeds from S525 to S525.
Proceed to 527. In S527, the CPU 10 determines in S52
It is checked whether there is any one of the action programs in which the profit calculated in 4 exceeds the cost calculated in S515. If the profit calculated in S524 exceeds the cost calculated for some of the action programs, the CPU 10 determines in S528 that the purchase probability is the highest among the action programs in which the calculated profit exceeds the cost. The action program creation process is terminated (corresponding to the adoption process).

On the other hand, if there is no action program in which the profit calculated in S524 exceeds the cost, the CPU 10 proceeds from S527 to S529.
Proceed to. In S529, the CPU 10 checks whether the variable m is larger than the variable n. And the variable m
Is less than or equal to the variable n, the CPU 10 proceeds to S53
At 0, the process returns to S524 after incrementing the variable m by one. On the other hand, when the variable m exceeds the variable n, the CPU 10 determines in S531
The action program in which the amount obtained by subtracting the cost calculated in S515 from the profit calculated in S524 is the largest is adopted, and the action program creation processing ends (corresponding to the adoption processing).

Returning to FIG. 5, after completion of S123, the CPU 1
0 in S124 (S517, S52).
0, S526, S528 or S531), the action program generator file 28 is created by applying the date and time of the current contact by the customer to be processed to the action program adopted in S531. As shown in FIG. 31, an example of the action program generator file 28 is as follows. The execution date and the action schedule (channel and content) of each action defined by the action program are scheduled for each customer. The stored file.

In the next S125, the CPU 10 proceeds to S12
4. The action program screen is displayed on the display 31 based on the data on the customer to be processed included in the action program generator file 28 created in step 4 (when this processing is started by bad processing, the action program screen is displayed). To print out). This action program screen is, as shown in FIG.
8 is a screen for displaying the content indicated by the data in 8 on a weekly basis.
After completion of S125, the CPU 10 advances the processing to S126.

In S126, the CPU 10 reflects the contents of the current contact by the customer to be processed in the “contact history” and “customer number” and “customer attribute information” in the customer data for the customer to be processed. Is created and stored in the hard disk 13. If the contact CIF 19 for the processing target customer already exists, the CP
U10 updates the contact CIF 19 by adding the content of the current contact to the "contact history" in the contact CIF 19. After the completion of S126, the CPU 10 ends the processing according to the action program creation program.

According to the present embodiment configured as described above, the floating layer primary discriminating process of S115 and the floating layer secondary discriminating process of S117 are executed, so that the insurance product is independent of the business activity. The customers (close-out layer) who are unlikely to purchase (conclude a subscription contract) are screened.
Therefore, it is possible to prevent waste of performing business activities at a cost for such customers.

On the other hand, a customer determined not to belong to the closing layer (that is, a customer who is likely to purchase an insurance product irrespective of business activities: a prospective class and a customer who can purchase insurance products regardless of business activities) Potential customer: floating layer)
PU 10 is based on customer attributes, including its personality traits,
From the rule DB 26, rules (channel rules, content rules, and approach rules) that can sell insurance products at the time of the first action execution to rules that can sell insurance products at the time of the third action execution are extracted (S502). These rules describe the channels, timings, and content that should be used for each action.
Each of the options defines priority 1 to priority 3.

Next, the CPU 10 sets “customer attribute information” for the customer, “channel allowance” listing channels allowed to be used by the customer, “contact history” which is a record of the contact of the customer, and The customer data including the “action program history”, which is a record of the action for the customer, is compared with the case data stored in the case DB 18 for the customer whose sales activities have been completed (S504 to S505). Then, k pieces of case data whose “customer attribute information”, “channel allowance” and “action program history” up to the first action are close to the customer data of the customer are extracted (S507), and the k case data are extracted. Is calculated as the sales probability (S508, S510), and the ratio of the case data (success example group) in which the insurance product can be sold at the time of executing the first action included in the success example is calculated for each action included in the success example group. An action program that sells insurance products at the time of the first action execution by collecting channels, timings and contents according to the priorities in the above rules, and selecting the channel, timing and contents with the highest points for each action. Create
Similarly, an action program for selling the insurance product at the time of executing the second action and its sales accuracy, and an action program for selling the insurance product at the time of executing the third action and its sales accuracy are respectively created (S50).
9).

Next, the CPU 10 calculates an average profit that can be expected by selling the insurance product to be sold based on the average premium and the average life of the insurance product to be sold (S513, S514). And if the cost of any action program you create is less than your revenue,
Adopt the action program with the highest sales accuracy (S
516, S517) Even in other cases, if there is an action program in which the cost is lower than the profit and the purchase probability exceeds the threshold, the action program with the highest purchase probability is adopted (S518 to S518). S5
20). If the action program could not be adopted even by the above algorithm, the channel with the highest cost among the channels set to “OK” in “allowed channel” in the proposed rule search value is set to “NG”. Then, the process is started again from the creation of the action program (S532, S504).

If the action program cannot be adopted even if the process is started again, the CPU 10 sets the customer within the remaining years until the customer turns 60 years old.
The maintenance period is extended by one year, and the recalculation of the profit is repeated (S522 to S524, S529, S530). Each time the profit is recalculated in this way, the CPU 10 executes the above-mentioned algorithm for adopting the action program and tries to adopt the action program (S525 to S52).
8). As a result, if any of the action programs cannot be adopted by the above algorithm even if the maintenance years are extended to the remaining years until the customer turns 60, the CPU 10 has the maximum (revenue-cost). An action program is adopted (S531).

As described above, any one of the action programs is adopted. As long as the cost does not exceed the profit expected from sales of insurance products, the action program with the highest sales accuracy for the customer is determined. , Will be automatically adopted. By conducting sales activities in accordance with the action program adopted in this way, even inexperienced or inexperienced salespeople can efficiently sell insurance products without wasting cost (conclusion of enrollment contract). can do.

(Supplementary Note 1) A business activity presentation method for presenting an action program for a sales target customer using a computer having a processing device for executing a program, a storage device, and an output device, wherein the processing device is one of In the case where the attribute information is obtained from the customer, a predetermined number of case data approximate to the attribute information is extracted from a database including the attribute information of each customer and the business activity information, thereby obtaining N (where N is an arbitrary number). And a recruiting process of generating one action program whose cost is lower than the profit from among N action programs created by the generating process. If one action program could not be adopted by processing,
A business activity wherein the conditions for the customer are reset, the processing after the creation processing is re-executed, and one action program finally adopted by the adoption processing is output from the output device. Presentation method.

(Supplementary Note 2) Using a computer having a processing device, a storage device, and an output device for executing a program, one or more combinations of channels, contents, and timings of one or more sales activities for a sales target customer. A business activity presentation method that presents an action program,
The processing device generates, for each customer, customer data including attribute information thereof, channel permission information for specifying a channel usable for business activities, and history information for recording the channel, timing and content of each business activity. At the same time, the storage device stores a database that stores customer data on individual customers whose sales activities have been completed and also includes case data including records of business contracts. When the information is acquired, a case extraction process of extracting a predetermined number of case data approximating the attribute information and the channel permission information from the database stored in the storage device, and the case data extracted by the case extraction process. From all the case data, a group of successful cases or N that were able to conclude a contract in the first sales activity
(However, N is an arbitrary natural number) Each group of successful cases for which a contract was concluded in the third business activity is extracted, and for each group of successful cases, the ratio of the number of successful cases to the predetermined number is calculated. Creation processing for creating N action programs by selecting the most similar channel, timing, and content used for each business activity for each business activity, and N created by this creation processing
Out of the plurality of action programs, the cost is lower than the profit, and the adoption process adopting the one action program in which the ratio is the largest is executed, and when one action program cannot be adopted by the adoption process, Excludes the channel with the highest cost from the channel permission information for the customer, re-executes the processing after the case extraction processing, and finally outputs one action program adopted by the adoption processing to the output device. Business activity presentation method characterized by output from

(Supplementary note 3) In the recruitment processing, if any of the N action programs has a profit exceeding the cost, only the action program in which the ratio exceeds a predetermined threshold is used. The business activity presentation method according to claim 2, wherein

(Supplementary note 4) The business activity according to Supplementary note 2 or 3, wherein the processing unit calculates the profit by multiplying an average annual charge by an average maintenance period of the business object contract in the recruitment processing. Presentation method.

(Supplementary Note 5) If one of the action programs cannot be adopted even after the case extraction process and the adoption process are re-executed, the processing device may reduce the average annual fee to calculate the profit. 5. The business activity presentation method according to claim 4, wherein the number of years to be multiplied is increased from the average maintenance years, and the recruitment process is executed again.

(Supplementary Note 6) The business activity presentation method according to Supplementary Note 2, wherein the attribute information includes information on the character of the customer.

(Supplementary Note 7) The business activity presentation method according to Supplementary Note 2, wherein the attribute information includes information for specifying a belonging cluster specified by classifying a customer attribute into a cluster.

(Supplementary Note 8) For each processing device in a computer having a processing device, a storage device, and an output device for executing data processing by executing a computer program, the attribute information is used for each customer and used for business activities. Generating customer data including channel permission information specifying possible channels, and history information recording the channel, timing and content of each sales activity, and storing, in the storage device, customer data for each customer whose sales activity has been completed; Then, when a database storing case data including records of business contracts is stored and attribute information and channel permission information are obtained from any customer, a case approximate to the attribute information and channel permission information is obtained. A case extraction process for extracting a predetermined number of data from the database stored in the storage device. Out of all the case data extracted by this case extraction processing, a group of successful cases through which the contract was concluded in the first sales activity or a contract is concluded in the Nth (where N is an arbitrary natural number) sales activities Each successful case group is extracted, the ratio of the number of successful cases to the predetermined number is calculated for each successful case group, and the channel, timing, and contents used for the sales activity for each business activity Out of the N action programs created by selecting the most similar ones from among the N action programs, and from the N action programs created by the creation processing, the cost is lower than the profit and the ratio is the largest. And a recruitment process that employs one of the action programs is executed.If the recruitment process fails to adopt the one action program, , By excluding the channel with the highest cost from the channel permission information for the customer, re-executing the processing after the case extraction processing, and finally, from the output device, the one action program adopted by the adoption processing. A sales activity presentation program characterized by being a computer program to be output.

(Supplementary Note 9) When the attribute information is acquired from one of the customers for the processing device in the computer having the processing device, the storage device, and the output device for executing the data processing by executing the computer program. Is extracted from a database including attribute information of each customer and business activity information by a predetermined number of case data approximated to the attribute information, thereby obtaining N (where N is an arbitrary natural number).
And executing a recruiting process for creating one action program and a recruiting process for adopting one of the N action programs created by the regenerating process that has a cost lower than the profit. If the action program cannot be adopted, the conditions for the customer are reset, the processes after the creation process are re-executed, and the one action program finally adopted by the A business activity presentation program, which is a computer program output from an output device.

(Supplementary Note 10) In a sales activity presentation device that presents one action program with respect to a sales target customer, an input device into which attribute information of each customer is input, and the attribute information of any customer is input by the input device. When input, a predetermined number of case data approximating the attribute information is extracted from a database including the attribute information of each customer and the business activity information, thereby obtaining N (where N is an arbitrary natural number) action programs. A means for creating an action program, a means for adopting one action program whose cost is lower than the profit out of N action programs created by the means for creating, and an action program adopted by the means for employing means If not, reset the conditions for the customer and restart the creation means That a restart means operating activities presentation device characterized by comprising an output device for outputting an action program adopted by finally the adopted means.

(Supplementary Note 11) In a sales activity presentation device that presents one action program, which is a combination of channels, contents, and timings of sales activities for one or more times, the attribute information of each customer Means for generating customer data including channel permission information for specifying channels usable for sales activities, history information for recording the channel, timing and content of each sales activity; A storage device storing a database storing customer data on customer and case data including records of business contracts; an input device for inputting attribute information and channel permission information of each customer; When the customer attribute information and channel permission information are input, Case extracting means for extracting a predetermined number of case data approximated to the information from the database stored in the storage device; and among all the case data extracted by the case extracting means, a first business activity The successful case groups through which the contract was concluded in the business activities N through N (where N is an arbitrary natural number) are extracted, and the number of successful cases is determined for each successful case group. Creation means for calculating a ratio to the number, and for each of the sales activities, selecting the most similar channel, timing and content used for the sales activity, thereby creating N action programs, Among the N action programs created by this creation means, one action program whose cost is lower than profit and whose ratio is the largest If one action program cannot be adopted by the adoption means adopting the ram and the adoption means processing, the highest cost channel is excluded from the channel permission information for the customer, and the case extraction means is re-executed. A business activity presentation device comprising: a restarting means for starting; and an output device for finally outputting one action program adopted by the adopting means.

[0100]

According to the sales activity presentation method of the present invention configured as described above, within the range where the cost does not exceed the profit expected from the conclusion of the sales target contract, the highest probability is given to the sales target customer. An action program can be created using a computer and presented to a sales representative.

[Brief description of the drawings]

FIG. 1 is a block diagram of a network system according to an embodiment of the present invention;

FIG. 2 is a flowchart showing processing by a customer reception server program.

FIG. 3 is a flowchart showing processing by an action program creation program;

FIG. 4 is a flowchart showing processing by an action program creation program;

FIG. 5 is a flowchart showing processing by an action program creation program;

FIG. 6 is a flowchart showing a customer characteristic determination processing subroutine executed in S110 of FIG. 4;

FIG. 7 is a flowchart showing a floating layer primary determination processing subroutine executed in S115 of FIG. 4;

FIG. 8 is a flowchart showing a floating layer secondary determination processing subroutine executed in S117 of FIG. 4;

9 is a flowchart showing an action program creation processing subroutine executed in S123 of FIG. 5;

FIG. 10 is a flowchart showing an action program creation processing subroutine executed in S123 of FIG. 5;

FIG. 11 is a diagram showing a customer attribute information input screen.

FIG. 12 is a diagram showing a psychological characteristic information input screen.

FIG. 13 is a diagram showing a killer question input screen.

FIG. 14 is a table showing a deployment example of answers to psychological test questions

FIG. 15 is a diagram showing a structure of case data in a case DB;

FIG. 16 is a diagram showing a data structure of a contact CIF.

FIG. 17 is a diagram showing a structure of customer data in a personal DB.

FIG. 18 is a table showing dummy variables for the responses of killer questions.

FIG. 19 is a table showing a primary discrimination parameter table.

FIG. 20 is a table showing a standardization table;

FIG. 21 is a table showing a cluster center table;

FIG. 22 is a diagram showing an example of a dummy variable.

FIG. 23 is a table showing a cluster-specific parameter table;

FIG. 24 is a table showing a reference table;

FIG. 25 is a diagram showing an example of a proposal rule initial value.

FIG. 26 is a diagram showing a data structure of a record in a rule DB.

FIG. 27 is a diagram showing an example of an extracted rule;

FIG. 28 is a diagram showing an example of a proposed rule search value.

FIG. 29 is a diagram showing an example of a generated action program.

FIG. 30 is a diagram showing a data structure of a process-specific channel-specific DB;

FIG. 31 shows an example of data of an action program generator file.

FIG. 32 shows an action program screen.

[Explanation of symbols]

1 insurance company system 2 User terminal 10 CPU 11 Communication adapter 12 RAM 13 Hard Disk 15 Customer reception server program 16 Action Program Creation Program 17 Temporary files 18 Case DB 19 Contact CIF 20 Personal DB 21 Primary discrimination parameter table 22 Normalization table 23 Cluster Center Table 24 Cluster Parameter Table 25 Reference table 26 Rule DB 27 Cost DB by Process and Channel 28 Action Program Generator File 29 Product DB 30 input device 31 Display

Claims (5)

[Claims] 1. A business activity presentation method for presenting an action program for a sales target customer using a computer having a processing device for executing a program, a storage device, and an output device, wherein the processing device includes: If you get the attribute above from
A creation process for creating N (where N is an arbitrary natural number) action programs by extracting a predetermined number of case data approximating the attribute information from a database including attribute information and sales activity information of each customer. And performing a recruiting process of adopting one action program whose cost is lower than the profit from among the N action programs created by the creation process. If one action program cannot be adopted by the recruitment process, Resets the conditions for the customer, re-executes the processes after the creation process, and finally outputs one action program adopted by the adoption process from the output device. Sales activity presentation method. 2. An action program which is a combination of channels, contents, and timings of one or more sales activities for a sales target customer using a computer having a processing device, a storage device, and an output device for executing the program. A business activity presentation method, wherein the processing device displays, for each customer, attribute information, channel permission information for specifying a channel usable for the business activity, a channel, timing, and content of each business activity. Along with generating customer data including history information to be recorded, the storage device stores a database which stores customer data on individual customers whose sales activities have been completed and which stores case data including records of sales target contracts, When the attribute information and the channel permission information are obtained from any customer, the attribute information and the channel A case extraction process of extracting a predetermined number of case data approximated to the file permission information from the database stored in the storage device; and, among all the case data extracted by the case extraction process, Successful cases where a contract was concluded in the sales activities through N (where N is an arbitrary natural number) are extracted, and the number of successful cases where a contract was concluded in the third sales activity is extracted. Creation processing for creating N action programs by calculating the ratio to the predetermined number and selecting the most similar channel, timing and content used for each sales activity for each sales activity And one of the N action programs created by the creation process, wherein the cost is lower than the profit and the ratio is the largest. And if an action program cannot be adopted by the adoption process, the channel with the highest cost is excluded from the channel acceptance information for the customer, and the case extraction process is performed. A business activity presentation method, wherein the following processing is re-executed, and one action program finally adopted by the adoption processing is output from the output device. 3. The processing apparatus according to claim 1, wherein the processing includes:
3. The business activity presentation method according to claim 2, wherein when any one of the N action programs exceeds the cost, only the action program in which the ratio exceeds a predetermined threshold is used. . 4. The processing device according to claim 1, wherein:
The business activity presentation method according to claim 2 or 3, wherein the profit is calculated by multiplying an average maintenance period of the business contract by an average annual fee. 5. When the attribute information is obtained from any of the customers for the processing device in a computer having a processing device, a storage device, and an output device for executing data processing by executing a computer program,
A creation process of creating N (where N is an arbitrary natural number) action programs by extracting a predetermined number of case data approximate to the attribute information from a database including the attribute information and the sales activity information of each customer; And executing, from the N action programs created by the creation process, a recruitment process that adopts one action program whose cost is lower than the profit. If the recruitment process fails to adopt one action program, Is a computer program for resetting the conditions for the customer, re-executing the processes after the creation process, and finally outputting one action program adopted by the adoption process from the output device. A sales activity presentation program characterized by the following.