Business support device

Abstract

To accurately extract customers who are expected to have needs for predetermined financial products and support efficient business activities in financial institutions.SOLUTION: A business support device comprises: acquisition means for acquiring transaction data of customers to be updated in association with the use of financial services by customers; relevance degree acquisition means for acquiring a relevance degree being a degree of a likelihood of purchasing a financial product by a candidate customer, by using at least feature quantities of transaction data of customers who have purchased financial products in the past and feature quantities of most recently updated transaction data of the candidate customer; and sales destination priority list creation means for listing in the priority order based on the relevance degree for the candidate customers, for whom the relevance degree has been acquired by the relevance degree acquisition means.SELECTED DRAWING: Figure 7

Description

[Technical Field]

[0001] The present invention relates to a sales support device. [Background technology]

[0002] In recent years, financial institutions have begun to pay attention to EBM (Event-Based Marketing), which proposes financial products that customers need at the right time. EBM is a marketing method that detects changes in customer (target) attribute information and trading behavior from a customer database, predicts these as opportunities (events) that will trigger financial needs, and then proposes the financial products that customers want.

[0003] Traditionally, EBM-like methods have been used to identify potential customers by applying arbitrary thresholds to statistical data such as relatively static and unchanging customer attribute information, deposit balances, and past financial purchase history. For example, a quantitative threshold for deposit amount is used to identify customers with deposits of 10 million yen or more, and sales approaches are then made to these potential customers, such as investment trust products. However, since there are customers with high needs even with deposits of less than 10 million yen, this still results in missing out on customers with potential needs. Furthermore, because customer groups do not change much, it is difficult to accurately capture daily changes in customer needs, making it difficult to accurately identify potential customers.

[0004] As an example of technology related to EBM for financial institutions, Patent Document 1 describes a method for supporting the development of new business partners, which extracts as target companies those companies whose current registered company-specific information differs from their previous company-specific information registered before the current company-specific information, and distributes information about the target companies to personnel in charge to support sales activities.

[0005] Furthermore, for example, Patent Document 2 describes a system that includes a means for classifying accounts into main accounts that receive or are making specified continuous transfers based on past transfer history, and sub-accounts that do not make specified continuous transfers; a means for determining that a large deposit made to a main account is a deposit with a specified purpose based on attribute data of the main account; a means for determining that a large deposit made to a sub-account is a candidate deposit that may be a deposit with a specified purpose; and a means for determining that a candidate deposit made to a sub-account is a deposit with the same purpose as the large deposit made to the main account, based on attribute data of the sub-account, if the source and transfer date of the candidate deposit match the source and transfer date of the large deposit made to the main account.

[0006] Furthermore, for example, Patent Document 3 discloses a corporate sales support system in which a corporate extraction means extracts corporations to be marketed to from among the corporations involved in each payment indicated by corporate payment information indicating payments made between corporations. The corporate extraction means extracts at least one of corporations that have made payments to new business partners within a target period and corporations that have received payments from new business partners within a target period, based on the combination of the receiving and receiving corporations in each payment made within a target period and the combination of the receiving and receiving corporations in each payment made within a past period prior to the target period. The corporate sales support system is described in which, when a corporation is extracted by the corporation extraction means, a sales support information output means outputs sales support information to support sales to the extracted corporation. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Patent No. 6208907 [Patent Document 2] Patent No. 5850592 [Patent Document 3] Patent No. 5416852 Summary of the Invention [Problem to be solved by the invention]

[0008] The inventions described in the above Patent Documents 1-3 focus on changes in customer information held by financial institutions (for example, company-specific information, balances of main and sub-accounts, and corporate payment information), establish certain rules based on anticipated needs as extraction conditions, and when a change in customer information that fits the rules is detected in the customer database, extract that customer as a target customer.

[0009] However, the criteria for extracting target customers are generalized rules established in advance by the system administrator based on anticipated needs. Therefore, it goes without saying that it is not possible to extract target customers who are predicted to have high needs despite not meeting the rules. In other words, as customer lifestyles have become increasingly diverse in recent years, it is difficult to identify potential prospective customers in detail by simply using generalized rules for changes in customer characteristics.

[0010] The present invention has been proposed in consideration of the above points, and in one aspect, aims to accurately extract customers who are expected to have needs for specified financial products and to support efficient sales activities at financial institutions. [Means for solving the problem]

[0011] In order to solve the above problems, the sales support device of the present invention is a sales support device for a financial institution, and comprises an acquisition means for acquiring transaction data of customers at the financial institution, an association determination means for determining the association between first transaction data of the customer at the time of purchasing a financial product in the past and second transaction data of the customer at the current time, an extraction means for extracting customers whose first transaction data and second transaction data are determined to be associated by the association determination means, and an output means for outputting the customers extracted by the extraction means.

[0012] In addition, in order to solve the above-mentioned problems, the sales support device of the present invention is a sales support device for a financial institution, and includes: an acquisition means for acquiring at least two or more different types of transaction data of customers at the financial institution; an association determination means for determining an association between a pattern of first transaction data of the customer at the time when the customer previously purchased a financial product and a pattern of second transaction data of the customer at the present time; an extraction means for extracting the customers whose first transaction data pattern and second transaction data pattern are determined to be associated by the association determination means; and an output means for outputting the customers extracted by the extraction means. [Effects of the Invention]

[0013] According to one aspect of the embodiment of the present invention, it is possible to accurately extract customers who are expected to have needs for a predetermined financial product, thereby supporting efficient sales activities at financial institutions. [Brief explanation of the drawings]

[0014] [Figure 1] 1 is a diagram illustrating an example of the configuration of a sales support system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram illustrating an example of a hardware configuration of a management server according to the present embodiment. [Figure 3] FIG. 2 is a diagram illustrating an example of the software configuration of a management server according to the present embodiment. [Figure 4A] An example of customer information held by the financial institution system 10 (for example, a core system that manages various types of deposits such as ordinary deposits and fixed term deposits) is shown below. [Figure 4B] An example of balance information held by a financial institution system 10 (for example, a core system that manages various types of deposits such as ordinary deposits and fixed term deposits) is shown below. [Figure 4C] An example of transaction statement information held by a financial institution system 10 (for example, a core system that manages various types of deposits such as ordinary deposits and fixed term deposits) is shown. [Figure 4D]An example of investment trust information held by a financial institution system 10 (a sales management system that manages sales results of financial products such as investment trusts) is shown. [Figure 4E] An example of asset management policy information held by the financial institution system 10 (a front compliance system for managing and proposing financial products according to the asset management needs of each customer) is shown below. [Figure 4F] An example of negotiation history information held by the financial institution system 10 (a history system that manages negotiation history) is shown below. [Figure 5] 10 is a flowchart showing a potential customer extraction process according to the present embodiment. [Figure 6] This shows the pattern distribution of each transaction data for the date and time of past investment trust purchases. [Figure 7] FIG. 10 shows a diagram for explaining the relationship between pattern distributions of each transaction data. [Figure 8] FIG. 10 shows a diagram for explaining the relationship between pattern distributions of each transaction data. [Figure 9] FIG. 10 shows a diagram for explaining the relationship between the pattern distribution of indicators based on each transaction data. [Figure 10] 10 shows an example of a sales destination priority list according to the present embodiment. [Figure 11] 10 shows an example of a sales product priority list (modification) according to the present embodiment. [Figure 12] FIG. 10 shows a diagram for explaining the relationship between pattern distributions of each transaction data. DETAILED DESCRIPTION OF THE INVENTION

[0015] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described in detail with reference to the drawings.

[0016] <System configuration> 1 is a diagram showing an example of the configuration of a sales support system according to this embodiment. The sales support system 100 in FIG. 1 includes a financial institution system 10, a sales promotion support system 20, a management server 30, a processing DB 40, and a sales terminal 50, which are connected via a network 70.

[0017] The financial institution system 10 is a variety of systems and databases (DBs) owned by financial institutions such as banks. The financial institution system includes, for example, a core system for managing various deposits such as ordinary deposits and fixed term deposits, a sales management system for managing sales records of financial products, a front compliance system for managing and proposing financial products according to the asset management needs of each customer, a history system (CRM) for managing negotiation history, and various DBs associated with each system. Note that types of financial products include, for example, investment trusts, free loans, card loans, personal car loans, and education loans.

[0018] The sales promotion support system 20 is a system for supporting and assisting the sales activities of sales representatives at financial institutions. For example, it manages the customers of sales representatives and manages sales progress such as daily sales reports. The system also allows sales representatives to share information and exchange opinions, and report information to superiors and exchange instructions.

[0019] The management server 30 is a server device that extracts prospective customers who are predicted to be highly likely to purchase financial products (e.g., investment trust products) along with their priorities (in order of highest likelihood of purchase) based on the processed data in the processed DB 40. The management server 30 also creates (generates) a sales destination priority list to be used by sales representatives based on the priorities.

[0020] Furthermore, when the results of sales activities carried out by sales representatives based on the sales client priority list are fed back to the sales promotion support system 20, the management server 30 compares the sales client priority list (prediction list) with the fed back results to verify the accuracy of the prediction and reflects it in creating the next sales client priority list.

[0021] The processed DB 40 is a DB that stores processed data obtained by sequentially acquiring raw data (called this because it is unprocessed) from the DB of the financial institution system 10 operated by the financial institution and processing the acquired raw data into a format that can be processed by the management server 30. Since machine learning methods often cannot handle raw input data such as text and time-series data as is, such unstructured data is converted into vectors through a conversion process. Note that although the formats of raw data and processed data differ, the content and meaning of the data itself are the same.

[0022] The sales terminal 50 is, for example, a personal computer (PC), smartphone, or tablet terminal, and is a user terminal used by a sales representative at a sales office or on the go. The sales representative uses the sales terminal 50 to access the management server 10, retrieve and display the sales target priority list. As described above, the sales target priority list is a list of potential customers predicted to be highly likely to purchase a specified financial product (e.g., an investment trust product), arranged in order of priority. By approaching customers (targets) on the sales target priority list in order of priority, the sales representative can conduct efficient sales activities even within a limited time. Note that a predetermined application program or web browser, etc., for retrieving and displaying the sales target priority list from the management server 30 is pre-installed on the sales terminal 50.

[0023] The network 70 is a communication network that may be wired or wireless, and may include, for example, the Internet, a public line network, or Wi-Fi (registered trademark).

[0024] <Hardware configuration> 2 is a diagram illustrating an example of the hardware configuration of a management server according to this embodiment. As shown in FIG. 2, the management server 30 includes a CPU (Central Processing Unit) 31, a ROM (Read Only Memory) 32, a RAM (Random Access Memory) 33, an HDD (Hard Disk Drive) 34, and a communication device 35.

[0025] The CPU 31 executes various programs and performs arithmetic processing. The ROM 32 stores programs required at startup, etc. The RAM 33 is a work area for temporarily storing processing by the CPU 11 and storing data. The HDD 34 stores various data and programs. The communication device 35 communicates with other devices via the network 70.

[0026] <Software configuration> 3 is a diagram showing an example of the software configuration of the management server according to this embodiment. The management server 30 has, as its main functional units, a data acquisition unit 301, a feature calculation unit 302, a relevance determination unit 303, an extraction unit 304, a sales destination list creation unit 305, and an output unit 306.

[0027] The data acquisition unit 301 acquires customer information and transaction data of customers at financial institutions from the processed DB 40 (or the sales promotion support system 20).

[0028] The feature amount calculation unit 302 calculates the feature amount (for example, a feature vector) of the transaction data from the pattern of the transaction data of the customer.

[0029] The relevance determination unit 303 determines whether or not there is a relevance between a customer's transaction data at the time when the customer previously purchased a financial product and the customer's current transaction data based on the degree of relevance between the two. For example, one method for determining the relevance between transaction data can be distance determination using feature vectorization. Specifically, the unit calculates the distance (corresponding to the degree of relevance) between a feature vector calculated from the pattern of the customer's transaction data at the time when the customer previously purchased a financial product and a feature vector calculated from the pattern of the customer's current (current) transaction data, and determines whether or not there is a relevance between the two transaction data based on whether the distance is close.

[0030] The extraction unit 304 extracts customers whose transaction data at the time of past purchase of financial products is determined by the association determination unit 303 to be associated with the customer's current transaction data, along with information indicating priority, as potential sales targets.

[0031] The sales destination list creation unit 305 creates sales destination list information in which the customers extracted by the extraction unit 304 are arranged in descending order of priority.

[0032] The output unit 306 outputs the sales destination list information to the sales terminal 50.

[0033] Each functional unit is realized by a computer program executed on hardware resources such as the CPU, ROM, and RAM of the computer that constitutes the management server 30. These functional units may be referred to as "means," "module," "unit," or "circuit." Each DB may be located on the HDD 34 of the management server 30 or on an external storage device on the network 70. Each functional unit of the management server 30 may not only be realized by a single server device, but may also be realized as a system consisting of multiple devices with distributed functions.

[0034] (Example of data items in a financial institution system database) Next, an example of data items in the DB (or processed DB 40) of the financial institution system 10 will be shown. However, it goes without saying that the data items shown in Fig. 4 are merely an example, and other data items may also be present.

[0035] FIG. 4A shows an example of customer information held by a financial institution system 10 (for example, a core system that manages various types of deposits such as ordinary deposits and fixed term deposits). This is customer attribute information for each customer, and is registered mainly based on applications from customers when opening an account, etc. Depending on the type of customer, there are information for individual customers and information for corporate customers.

[0036] FIG. 4B shows an example of balance information held by financial institution system 10 (for example, a core system that manages various types of deposits such as ordinary deposits and fixed term deposits). This is balance information for each account held by each customer, such as ordinary deposit accounts, current accounts, fixed term deposit accounts, foreign currency deposit accounts, and investment trust accounts. Not only the current (latest) balance, but also all past balances are recorded.

[0037] FIG. 4C shows an example of transaction details information held by a financial institution system 10 (for example, a core system that manages various types of deposits such as ordinary deposits and fixed-term deposits). This is detailed information on each transaction that has occurred for each customer, such as ordinary deposit transaction details, credit card usage details, current account transaction details, and exchange transaction details. All transactions from the past to the present are recorded.

[0038] 4D shows an example of investment trust information held by the financial institution system 10 (a sales management system that manages sales performance of financial products such as investment trusts). This information includes customer investment trust account information, deposit balance information, dividend redemption details information, profit and loss information, and other information related to the investment trust for each customer.

[0039] Figure 4E shows an example of asset management policy information held by the financial institution system 10 (a front-end compliance system for managing and proposing financial products according to the asset management needs of each customer). This is the policy and asset management needs for each customer's financial asset management, and is registered mainly based on the customer's application when opening an investment trust account, etc. Sales representatives will propose financial products and asset management that match the policy needs in accordance with the policy.

[0040] FIG. 4F shows an example of negotiation history information held by the financial institution system 10 (a history system that manages negotiation history). The history of negotiations, proposals, and direct mail sent by sales representatives and financial institution counter staff is recorded for each customer. This allows the entire history of past negotiations between customers and financial institutions to be managed and made viewable.

[0041] Here, customer information and asset management policy information are relatively static and unlikely to change. On the other hand, balance information, transaction details information, investment trust information, and negotiation history information are relatively dynamic and prone to change, and are "transaction data." Generally, transaction data is one type of data handled by corporate information systems, and refers to data that records details of events that occur in the course of business.

[0042] The transaction data according to this embodiment refers to data that records details of events that occur when a customer uses financial services, particularly at a financial institution, such as deposit balance information, transaction details information, and investment trust information. Furthermore, the transaction data according to this embodiment is data that changes relatively frequently (updates frequently) on a daily basis, compared to information such as customer information, which does not change very frequently after registration. However, because the update frequency, when viewed on an individual customer basis, depends in part on the frequency with which the customer uses financial services, the specific update frequency is not important.

[0043] <Information Processing> Next, various information processes executed by the management server 30 will be described in detail. (Data update process) The management server 30 sequentially acquires raw data from the DB of the financial institution system 10 operated by the financial institution, processes and converts the acquired raw data into a format that can be calculated by the management server 30, and updates the processed data in the processed DB 40. The update cycle can be set arbitrarily depending on the operation, for example, every second, every few seconds to tens of seconds, every minute, every few minutes to tens of minutes, every hour, every few hours, every day, or every specified day, but the shorter the update interval, the better, and it is desirable to update in real time along with updates to the raw data in the DB of the financial institution system 10. As a result, the processed DB 40 updates data equivalent to that in the DB of the financial institution system 10 at the above update intervals, although the raw data and processed data have different formats.

[0044] (Pretreatment) In this embodiment, an example will be described in which prospective customers who are predicted to have a high probability of purchasing investment trust products, as an example of financial products, are extracted.

[0045] First, the management server 30 classifies target customers into groups based on whether they have previously purchased investment trusts, based on customer transaction data (especially investment trust information). The customers to be classified are existing customers of the financial institution who hold at least a regular savings account or the like. Customer Group A: Customers who have previously purchased investment trusts and currently hold them (targets for additional investment trust purchases), and customers who have previously purchased investment trusts but do not currently hold them (targets for repurchasing investment trusts). Customer Group B: Customers who only have a regular deposit account, have no investment trust purchase experience, and currently do not own any investment trusts (target for new investment trust purchases)

[0046] (Prospective customer extraction process for investment trusts) Next, a process for determining and extracting potential customers for the investment trust will be described. The extracted customers are listed in a sales target priority list.

[0047] In the process of extracting potential customers for investment trusts, for customers classified into customer group A who have previously purchased investment trusts, a feature vector P indicating the features of the customer's transaction data from past purchases and a feature vector N indicating the features of the customer's current transaction data are calculated, and based on the distance between feature vector P and feature vector N, a determination is made as to whether or not the customer is currently likely to purchase investment trusts, and to what extent.

[0048] On the other hand, for customers classified in customer group B who have no experience of purchasing investment trusts in the past, a feature vector P' indicating the feature amounts of transaction data from past purchases by customers classified in customer group A and a feature vector N indicating the feature amounts of the customer's current transaction data are calculated, and the presence or absence and degree of likelihood of the customer purchasing investment trusts is determined based on the distance between feature vector P' and feature vector N. This will be explained in detail below.

[0049] 5 is a flowchart showing the prospective customer extraction process according to this embodiment. The process is performed by the management server 30.

[0050] S1: The management server 30 selects one candidate customer from the candidate customers to be extracted. The candidate customer is a specified customer segment. The customer segment that becomes the candidate customer can be selected and set as follows: For example, it can be all customers of the financial institution (e.g., customers who hold accounts), or if it is a grouping of sales branches, it can be all customers of that branch (e.g., customers who hold accounts), or if it is a grouping of sales representatives, it can be all customers handled by that sales representative. Alternatively, it can be a grouping of customers who hold accounts but have never purchased investment trusts (equivalent to the above-mentioned customer group B), or customers who have purchased investment trusts and hold accounts (equivalent to the above-mentioned customer group A). ​​It is also possible to set a specific number of customers among all customers of the financial institution. A determination is made individually for each candidate customer as to whether they are a prospective customer. Candidate customers can also be arbitrarily determined by inputting detailed conditions (customer attribute information such as place of residence, age, occupation, and income).

[0051] S2: The management server 30 determines whether the candidate customer acquired in S1 has any experience purchasing investment trusts in the past. If the candidate customer acquired in S1 has any experience purchasing investment trusts in the past, i.e., is classified into customer group A, the process proceeds to S3. On the other hand, if the candidate customer acquired in S1 has no experience purchasing investment trusts in the past, i.e., is classified into customer group B, the process proceeds to S21.

[0052] S3: The management server 30 obtains from the processed DB 40 transaction data from the time when the candidate client classified into client group A previously purchased an investment trust (or for a predetermined period prior to the purchase, which corresponds to the time immediately before the purchase).

[0053] The transaction data to be acquired is not limited to one type of transaction data, but at least two or more types of transaction data as of the purchase date. Regarding the type of transaction data to be acquired, it is possible to acquire all types of transaction data, but it is preferable to selectively acquire appropriate types of transaction data. The type of transaction data to be selected depends on the specific financial product in question, but it can be selected based on, for example, an analysis of the transaction data history of customers who purchased the target financial product, or the results of sales activities (i.e., whether or not customers listed on the sales priority list ultimately purchased investment trusts), and this selection can also be changed or revised.

[0054] Furthermore, if multiple investment trust purchase dates and times are specified, transaction data (transaction data values) for all purchase dates and times are acquired. Alternatively, if multiple investment trust purchase dates and times are specified, transaction data for any one date and time, such as the most recent date and time, may be acquired.

[0055] S4: The management server 30 calculates a feature vector P from the patterns of the multiple transaction data (combinations of multiple transaction data (values)) acquired in S3. The calculated feature vector P indicates the features of the transaction data situation of the candidate customer at the time when the candidate customer previously purchased an investment trust. If the dates and times of multiple investment trust purchases have been identified and transaction data for multiple purchase dates and times has been acquired, the feature vector P (P1, P2, P3, etc.) for each purchase date and time can be calculated.

[0056] Figure 6 shows the pattern distribution of each transaction data at the date and time of past investment trust purchases. In the figure, A, B, C, D, and E each represent a different type of transaction data item. The customer in question has a history of purchasing investment trusts in the past at times when each transaction data value formed the pattern distribution shown in Figure 6. Therefore, feature vector P is a numerical representation of the situation in which the customer's demand for investment trusts is predicted to increase, based on the pattern distribution (feature group) of past transaction data held by the financial institution.

[0057] S5: The management server 30 then obtains the current transaction data for the candidate customer from the processed DB 40. The current transaction data is the transaction data that was last updated on the processed DB 40. The transaction data to be obtained is of the same type as the transaction data obtained in S3, so that the types of transaction data to be compared match.

[0058] S6: The management server 30 calculates a feature vector N from the pattern of the transaction data (combination of transaction data (values)) acquired in S5. The calculated feature vector N indicates the feature amount of the transaction data status of the candidate customer at the current time.

[0059] S7: The management server 30 executes a process for determining the association between the feature vector P and the feature vector N. For customers classified into customer group A who have previously purchased investment trusts, the management server 30 calculates the distance (corresponding to the degree of association) between the feature vector P of the transaction data of the customer at the time of the previous purchase and the feature vector N of the customer's current transaction data, and determines whether there is an association based on the calculated distance.

[0060] Specifically, first, the management server 30 calculates the distance between the feature vector P and the feature vector N. Note that the distance between the feature vectors can be calculated using a conventional method, but the closer the distance between the calculated feature vectors, the higher (larger) the degree of association, and the farther the distance, the lower (smaller) the degree of association.

[0061] Next, it is determined whether the calculated distance between the feature vectors is smaller than a reference value (whether they are closer than the reference value). If it is smaller than the reference value, the distance between feature vector P and feature vector N is close, and it is determined that feature vector P and feature vector N are related. Note that this reference value is a value used as a criterion for determining whether feature vectors are related in the relatedness determination process, but it is not necessarily a fixed value, and can be a dynamic value calculated relatively.

[0062] Then, if feature vector P and feature vector N are related, the candidate customer is determined to be likely to purchase the investment trust, and the distance between the feature vectors previously calculated is also obtained as the degree of association. In this embodiment, the smaller (closer) the distance value between feature vector P and feature vector N is, that is, the higher the degree of association between feature vector P and feature vector N, the higher the degree of likelihood / expectation of the customer to purchase the investment trust.

[0063] Figure 7 shows a diagram explaining the relationship between the pattern distributions of each transaction data. Even for the same customer (e.g., Yamada Hanako), each transaction data changes from day to day. If the pattern distribution (corresponding to feature vector N) of the daily transaction data of a customer (e.g., Yamada Hanako) classified into customer group A is related to the pattern distribution (corresponding to feature vector P) of transaction data from when that customer (e.g., Yamada Hanako) previously purchased an investment trust, then that time is a time when that customer (e.g., Yamada Hanako) is likely to purchase an investment trust.

[0064] In the figure, for example, the pattern distribution of customer Yamada Hanako's transaction data on January 7, 2019 (two weeks ago) and January 14, 2019 (one week ago) is not related to the pattern distribution of each transaction data at the date and time of customer Yamada Hanako's past investment trust purchases. However, as each transaction data changes over time, the pattern distribution of customer Yamada Hanako's transaction data on January 21, 2019 (present) is related to the pattern distribution of each transaction data at the date and time of customer Yamada Hanako's past investment trust purchases. In other words, as of January 21, 2019 (present), the pattern distribution of the current transaction data indicates that customer Yamada Hanako is in the same situation as when she actually purchased investment trusts in the past, meaning that now is an appropriate time to sell investment trust products to customer Yamada Hanako, as she is likely to purchase investment trusts again.

[0065] Here, because each transaction data changes from moment to moment, if the transaction data were to change significantly thereafter, the pattern distribution of customer Yamada Hanako's transaction data would no longer be related to the pattern distribution of each transaction data at the date and time of her past investment trust purchases. In such a case, customer Yamada Hanako is no longer considered to be in a situation where she is likely to purchase investment trusts, and it can be said that the appropriate timing to sell investment trust products to customer Yamada Hanako has been missed.

[0066] In addition, when the dates and times of multiple investment trust purchases have been identified and transaction data for multiple purchase dates and times has been obtained, an association determination process is performed between the feature vector P (P1, P2, P3, etc.) for each purchase date and time and the feature vector N, and if the distance between any one of the feature vectors P and the feature vector N is smaller than a reference value, it can be determined that they are associated. Furthermore, if it is determined that multiple feature vectors P are associated, it can be said that there is an even higher possibility that the investment trust will be purchased again, so the distance obtained as the degree of association can be changed to a larger value as the number of feature vectors P determined to be associated increases. Alternatively, the distance obtained as the degree of association can be the maximum value of the multiple distances obtained when determining the association between the feature vector P (P1, P2, P3, etc.) and the feature vector N, or the average value of the multiple distances obtained.

[0067] S8: If it is determined that the feature vector P and the feature vector N are related as a result of the relatedness determination process, proceed to S9. If it is not determined that they are related, proceed to S10.

[0068] S9: The candidate customer is extracted as a prospective customer who is predicted to have a high possibility of purchasing an investment trust at the current time. Also, the distance between the feature vectors calculated earlier is obtained as the degree of association.

[0069] S10: Determine whether there are any unprocessed candidate customers. If there are any unprocessed candidate customers, proceed to S1 again. If this process has been executed for all candidate customers, this process ends.

[0070] S21: On the other hand, if in S2 the candidate customer has no experience of purchasing investment trusts in the past, the management server 30 determines customers (referred to as similar customers) who have similar customer information (customer attribute information) or are related to the candidate customer from among customers (customer group A) who have experience of purchasing investment trusts in the past. This is because customers who have similar customer attributes, such as age, gender, occupation, and annual income, are likely to purchase investment trusts at similar times under similar transaction data pattern conditions.

[0071] S22: The management server 30 acquires transaction data from the processed DB 40 when the similar customer determined in S21 previously purchased an investment trust (or a predetermined period before the purchase corresponding to the purchase). The acquired transaction data is not only one type of transaction data, but at least two or more types of transaction data as of a specific date.

[0072] If there are multiple similar customers, the multiple similar customers may be determined in S21, and all transaction data from when the multiple similar customers previously purchased investment trusts may be acquired in S22.

[0073] S23: The management server 30 calculates a feature vector P' from the pattern of the transaction data (combination of transaction data (values)) acquired in S22. The calculated feature vector P' indicates the feature amount of the transaction data situation of the similar customer at the time when the similar customer previously purchased the investment trust.

[0074] Furthermore, if transaction data of the multiple similar customers when they previously purchased investment trusts is acquired in S22, multiple feature vectors P' (P'1, P'2, P'3...) may be calculated.

[0075] S24: The management server 30 acquires current transaction data for the candidate customer who is now classified into customer group B from the processed DB 40. The current transaction data is the transaction data that was last updated on the processed DB 40. The acquired transaction data is also of the same type as the transaction data acquired in S22, corresponding to the transaction data.

[0076] S25: The management server 30 calculates a feature vector N from the pattern of the transaction data (combination of transaction data (values)) acquired in S24. The calculated feature vector N indicates the feature amount of the transaction data status of the candidate customer at the current time.

[0077] S26: The management server 30 executes a process for determining the association between the feature vector P' and the feature vector N. For customers classified into customer group B who have no experience of purchasing investment trusts in the past, the management server 30 calculates the distance between the feature vector P' of the transaction data of past purchases of customers classified into customer group A and the feature vector N of the customer's current transaction data, and if the calculated distance is smaller than a reference value, it is determined that the distance between the feature vector P' and the feature vector N is close, and therefore the feature vector P' and the feature vector N are associated.

[0078] Figure 8 shows a diagram explaining the relationship between the pattern distributions of each transaction data. Even for the same customer (e.g., Ichiro Suzuki), each transaction data changes from day to day over time. If it is determined that the pattern distribution (corresponding to feature vector N) of the daily transaction data of a customer (e.g., Ichiro Suzuki) classified into customer group B is related to the pattern distribution (corresponding to feature vector P') of transaction data from a past purchase of an investment trust by a customer (e.g., Hanako Yamada) who belongs to customer group A and has similar or related customer attributes, then this is the time when the customer (e.g., Ichiro Suzuki) is likely to purchase an investment trust.

[0079] In the figure, for example, on January 7, 2019 (two weeks ago) and January 14, 2019 (one week ago), the pattern distributions of transaction data for customer Ichiro Suzuki, who has no prior investment trust purchase experience, are not related to the pattern distributions of transaction data for customer Hanako Yamada, who has prior investment trust purchase experience, at the date and time of her previous investment trust purchase. Meanwhile, the pattern distribution of transaction data for customer Ichiro Suzuki on January 21, 2019 (present) is related to the pattern distribution of transaction data for customer Hanako Yamada, who has prior investment trust purchase experience. In other words, as of January 21, 2019 (present), the pattern distribution of customer Ichiro Suzuki's current transaction data indicates that customer Hanako Yamada, who has similar or related customer attributes, is in the same circumstances as when she actually purchased investment trusts in the past. This means that now, when there is a high possibility that she will purchase investment trusts, is an appropriate time to promote investment trust products to customer Ichiro Suzuki.

[0080] If the result of the association determination process shows that the feature vector P' and the feature vector N are associated (S8: YES), the candidate customer is extracted as a prospective customer predicted to have a high probability of purchasing an investment trust at the current time, along with the distance between the feature vectors previously calculated as the association degree (S9). It is also determined whether there are any unprocessed candidate customers (S10), and if there are any unprocessed candidate customers, the process proceeds to S1 again. When this process has been performed for all candidate customers, the process ends.

[0081] Conventionally, the relationship between a target customer and their needs has been determined based on experience or knowledge, or on one or more data values ​​(for example, the amount of data value, or whether it matches or does not match a predetermined value, etc.), but in this embodiment, the determination is made based on the mutual patterns (mutual combinations) or relative patterns (relative combinations) of various transaction data values.

[0082] For example, even if one transaction data value (such as a fixed-term deposit account balance) and another transaction data value (such as a savings account balance) among multiple transaction data are both trending within a low range, there may be a situation in which both transaction data values ​​change to moderately high values ​​that are not yet high, which may increase a customer's desire to purchase investment trusts. If one focuses only on the magnitude of one of the transaction data values, this increase in the desire to purchase investment trusts will be overlooked.

[0083] That is, this embodiment makes it possible to extract target customers who are predicted to have high needs, even though this has been difficult to do using conventional methods. Furthermore, the timing at which financial products are likely to be purchased is calculated and quantified for each individual customer based on their past and present transaction data, which changes daily. This allows sales representatives to improve sales efficiency by proposing and selling financial products at the appropriate timing specific to each individual customer.

[0084] (Supplementary Note 1) Figure 9 shows a diagram explaining the relationship between the pattern distribution of indices based on each transaction data. Transaction data A to E shown in Figures 7 and 8 do not necessarily have a one-to-one correspondence with each individual transaction data. For example, let A to E be the indices listed below. Each index is index data that includes one or more transaction data, customer information, and / or asset management policy information. A: "Risk tolerance" - Index data based on investment trust information, asset management policy information, etc. B: "Trading experience": Indicator data based on trading history, investment trust information, etc. C: "Large deposits" - Index data based on current deposit balances, etc. D: "Investment Capacity": Indicator data based on current deposit balances, liquidity deposit balances, investment trust balances, etc. E: "Investment Timing" - Indicator data based on customer information, regular deposit and withdrawal transactions (salary, bonus payment dates, etc.)

[0085] For each customer, transaction data changes as the customer uses financial services, and the index data also changes. If the pattern distribution of the customer's daily index data (corresponding to feature vector N) is related to the pattern distribution of the index data (corresponding to feature vector P and feature vector P') when the customer previously purchased an investment trust, then that is the timing when the customer is likely to purchase an investment trust.

[0086] This makes it possible to create more effective and realistic indicators based on current transaction data, customer information, and / or asset management policy information, from the perspective of extracting customers who are more likely to purchase, and to determine the likelihood of a customer purchasing an investment trust based on the mutual pattern distribution of the indicators.

[0087] (Supplementary Note 2) For example, if a customer has a history of purchasing an investment trust when their current savings account balance was 3 million yen in the past, and their current savings account balance is around 3 million yen, the current and past transaction data will match, and they will be determined to be related.

[0088] On the other hand, if the current balance of a savings account is around 5 million yen, the current and past transaction data may be deemed unrelated. However, if a customer has previously purchased an investment trust when their balance was 3 million yen, they are likely to also purchase an investment trust when their balance is 5 million yen. In such a case, rather than determining that the larger balance in the savings account is unrelated, it is possible to determine that the current balance, which is higher than the balance at the time of the previous purchase, is relevant.

[0089] Specifically, if there is predetermined transaction data of this type, the transaction data at the time of the previous investment trust purchase is obtained, and if the current transaction data is greater than the transaction data at the time of the previous investment trust purchase, the current transaction data is corrected so that it matches or approximates the value of the transaction data at the time of the previous investment trust purchase. Then, a feature vector can be calculated that includes this type of transaction data as one transaction data.

[0090] Note that the opposite is also true when this type of predetermined transaction data is of a type where the smaller the value, such as risk tolerance, the more desirable it is. That is, if the current transaction data is smaller than the transaction data at the time of the previous investment trust purchase, the current transaction data is corrected so that it matches or approximates the value of the transaction data at the time of the previous investment trust purchase. Then, this type of transaction data can be included as one transaction data to calculate a feature vector.

[0091] (Sales destination priority list creation process) When the management server 30 receives a request to acquire or display a sales destination priority list from the sales terminal 50 of the sales representative, the management server 30 creates a sales destination priority list in which the candidate customers extracted by the potential customer extraction process (S9) are arranged in order of priority (in order of highest acquired relevance / in order of shortest acquired distance). In addition, the management server responds with the created sales destination priority list based on the acquisition request from the sales terminal 50.

[0092] 10 shows an example of a customer priority list according to this embodiment. A sales representative obtains and displays a customer priority list screen from the management server 10 on a sales terminal 50. The customer priority list 51 includes, for example, a sales branch selection field 52, a sales representative selection field 53, a financial product selection field 54, a customer number specification field 55, and a list 56.

[0093] The sales branch selection field 52 is a field for selecting a sales branch. Customers who have accounts at the selected sales branch become target candidate customers on the sales priority list 51.

[0094] The sales representative selection column 53 is a column for selecting a sales representative. The customers for which the selected sales representative is in charge become target candidate customers in the sales destination priority list 51.

[0095] The financial product selection field 54 is a field for selecting the type of financial product that will be used as a key for list search. Customers who are likely to purchase the selected financial product will become target candidate customers on the sales priority list 51.

[0096] The customer number specification field 55 is a field for specifying and selecting the number of customers (number of cases) to be displayed in the list 56. The specified number of customers, starting with customers with the highest priority and score, are displayed in the list 56. A sales representative specifies, for example, the number of customers that can be approached on that day.

[0097] List 56 is a sales target list in which customers who have accounts at the selected branch and are handled by the selected sales representative are selected through the prospective customer extraction process for the selected financial product, and are arranged in order of priority. List 56 also has items such as priority 56a, score 56b, customer group 56c, and customer information 56d.

[0098] The priority 56a indicates the order of priority in which sales should be made. The higher the relevance obtained in S9 (the closer the distance obtained), the higher the likelihood of a customer purchasing the investment trust, so the priority 56a corresponds to the obtained relevance (distance). In other words, on the list 56, customers are arranged in descending order of the relevance obtained together with customers extracted by the potential customer extraction process (in descending order of distance). The higher the priority 56a, the higher the likelihood of a customer purchasing the selected financial product, so sales representatives can make efficient sales by approaching customers in order from the highest priority 56a.

[0099] The score 56b is the value of the relevance (distance) itself or a corrected value thereof, and allows a salesperson to grasp the degree of likelihood of purchase for each customer in terms of a specific numerical value.

[0100] The customer group 56c indicates whether or not the customer has experience purchasing the financial product 54. For example, in accordance with the above, customers who have purchased investment trusts in the past and actually still hold them (targets for additional investment trust purchases) are designated as A1, customers who have purchased investment trusts in the past but do not currently hold them (targets for repurchasing investment trusts) are designated as A2, and customers who have not purchased investment trusts in the past and do not currently hold them (targets for new investment trust purchases) are designated as B. This information is useful for sales representatives when proposing and selling financial products.

[0101] The customer information 56d is information about the customer including the customer's customer number, name, age, etc. The salesperson refers to the customer information 56d on the list 56 when approaching the customer.

[0102] When creating the sales destination priority list, it is also possible to exclude customers who are already in sales or have already been sold to from the sales destination priority list 51 (list 56) based on the past negotiation history of the sales promotion support system 20. This is to avoid duplicate sales.

[0103] Furthermore, in list 56, if there is a large difference in the scores 56b between a customer and the next-highest priority customer (e.g., a score of 0.1 or more), indicator 56e is displayed, for example, using an arrow, line, or color, to distinguish between the customer and the next-highest priority customer. For example, in the case of indicator 56e in FIG. 10, the difference in the scores between priority 6 and priority 7 is large. Therefore, especially when business hours are limited, by approaching only customers with priorities 1 to 6 according to indicator 56e, it is possible to optimize time-to-effectiveness compared to, for example, approaching customers with priorities 1 to 7. In this case, the customers with priorities 1 to 10 in list 56 are further divided into sales-priority groups, such as first priority group 1 to 6 and second priority group 7 to 10, according to indicator 56e.

[0104] As described above, according to the sales target priority list 51, it is possible to list customers who are appropriate to approach at the time when the sales target priority list 51 has been created. Sales representatives can improve sales efficiency by proposing and selling financial products based on the sales target priority list 51.

[0105] <Modification> Next, the sales product priority list will be explained. The sales customer priority list in Fig. 10 uses a financial product (e.g., investment trust) as a key, and lists customers who are likely to purchase the financial product, with the priority assigned to that product. The sales product priority list in this modification uses a customer (e.g., Yamada Hanako) as a key, and lists financial products (types of financial products) that the customer is likely to purchase from among multiple types of financial products, with the priority assigned to those products.

[0106] 11 shows an example of a sales product priority list (variation) according to this embodiment. The sales product priority list 61 has, for example, a customer number input field 62, a search button 63, customer information 64, and a list 65.

[0107] For example, consider a situation in which a customer (e.g., Yamada Hanako) comes to a customer service counter and a recommended financial product is proposed to the customer. The salesperson enters the customer number of the customer in the input field 62 on the sales terminal 50 and presses the search button 63, thereby displaying the customer information 64 and list 65 of the customer.

[0108] The list 65 is a sales product priority list in which financial products that are likely to be purchased by the selected customer are arranged in order of priority.

[0109] The priority 65a indicates the order of priority for sales. The priority 65a allows customers to be sorted in descending order of the degree of relevance (closest distance) obtained with each financial product. The higher the priority 65a, the higher the likelihood of purchasing the selected financial product. Therefore, sales representatives can efficiently propose products by approaching customers in descending order of priority 65a.

[0110] The score 65b is the value of the relevance (distance) itself or a corrected value thereof, and allows the salesperson to grasp the degree of likelihood of purchase for each financial product in a specific numerical value.

[0111] The financial product name 65c indicates the type and name of the specific financial product to be proposed (recommended) to the customer. It is also possible to display related information according to the specific financial product, such as "limit" and "latest purchase date." The sales representative makes product proposals to the customer while referring to the financial product name 65c on the list 65.

[0112] Figure 12 shows a diagram explaining the relationship between the pattern distributions of each transaction data. As mentioned above, each transaction data changes from day to day, even for the same customer. If the pattern distribution of a customer's (e.g., Hanako Yamada) daily transaction data is related to the pattern distribution of transaction data when the customer previously purchased each type of financial product (e.g., investment trust, car loan), then that time is the time when the customer (e.g., Hanako Yamada) is most likely to purchase that type of financial product.

[0113] Furthermore, for types of financial products (e.g., free loans, card loans, and education loans) that the customer (e.g., Yamada Hanako) has not purchased in the past, similar customers who have purchased the financial products in the past and have customer attributes similar or related to the customer (e.g., Yamada Hanako) in the past are determined, and the correlation with the pattern distribution of transaction data when the similar customers previously purchased each type of financial product (e.g., free loans, card loans, and education loans) is determined.

[0114] Specifically, first, with regard to investment trusts, on January 21, 2019 (as of now), when a customer (for example, Yamada Hanako) visits the customer service counter, the pattern distribution of transaction data for customer Yamada Hanako is not related to the pattern distribution of each transaction data at the date and time of customer Yamada Hanako's past investment trust purchases.

[0115] Next, regarding car loans, the pattern distribution of transaction data for customer Yamada Hanako as of January 21, 2019 (current time) is related to the pattern distribution of each transaction data at the date and time of customer Yamada Hanako's past car loan purchases.

[0116] Next, for free loans, since the customer (for example, Yamada Hanako) has never purchased a free loan in the past, similar customers (for example, Tanaka Taro) who have purchased free loans in the past and have customer attributes similar or related to the customer (for example, Yamada Hanako) are determined. As of January 21, 2019 (current time), the pattern distribution of transaction data for customer Yamada Hanako is related to the pattern distribution of each transaction data for customer Tanaka Taro at the date and time of past free loan purchases.

[0117] Next, for card loans, since the customer (e.g., Yamada Hanako) has never purchased a card loan in the past, similar customers (e.g., Suzuki Jiro) who have purchased card loans in the past and have customer attributes similar or related to the customer (e.g., Yamada Hanako) are determined. As of January 21, 2019 (current time), the pattern distribution of transaction data for customer Yamada Hanako is related to the pattern distribution of each transaction data for customer Suzuki Jiro at the date and time of past card loan purchases.

[0118] Next, for educational loans, since the customer (e.g., Yamada Hanako) has never purchased an educational loan in the past, similar customers (e.g., Sato Kazuko) who have purchased educational loans in the past and have customer attributes similar or related to the customer (e.g., Yamada Hanako) are determined. As of January 21, 2019 (current time), the pattern distribution of transaction data for customer Yamada Hanako is not related to the pattern distribution of each transaction data for customer Sato Kazuko at the date and time of her past educational loan purchases.

[0119] As described above, the association determination unit 303 of the management server 30 in this modified example determines, for each type of financial product, the association between the pattern distribution of the customer's transaction data and the pattern distribution of each transaction data at the date and time of past purchase of the financial product.

[0120] Next, the extraction unit 304 extracts the types of financial products to which the pattern distribution of the customer's transaction data is related by the relevance determination unit 303 as financial products to be promoted, along with a relevance indicating priority. Specifically, car loans, free loans, and card loans are extracted as financial products to be promoted, along with a relevance indicating the degree of relevance, such as the distance between the feature vectors described above. The sales destination list creation unit 305 then creates sales product list information in which each type of financial product extracted by the extraction unit 304 is arranged in descending order of priority (corresponding to the relevance) (FIG. 11).

[0121] Whether or not an investment trust has been purchased in the past can be determined by referring to, for example, the example of investment trust information in Figure 4D. However, it is also possible to determine whether or not an investment trust has been purchased in the past for financial products such as free loans, card loans, car loans, and education loans by referring to a financial institution system 10 (a sales management system that manages the sales performance of various financial products) not shown.

[0122] It is also possible to use investment trusts as a key and create a list (investment trust product priority list) of specific investment trust product names that are likely to be purchased by the customer, with priority assigned to them, such as investment trust product A, investment trust product B, etc. In this case, for each investment trust product, it is possible to determine the correlation between the pattern distribution of the customer's transaction data and the pattern distribution of each transaction data at the date and time of past purchases of that investment trust product.

[0123] <Summary> As described above, the sales support system 100 according to this embodiment can accurately extract customers who are expected to have needs for a given financial product, and can support efficient sales activities at financial institutions.

[0124] Although the present invention has been described with reference to specific examples according to the preferred embodiments of the present invention, it is apparent that various modifications and changes can be made to these examples without departing from the broad spirit and scope of the present invention as defined in the appended claims. In other words, the details of the examples and the accompanying drawings should not be construed as limiting the present invention.

[0125] The management server 30 can be called a sales support device, a potential customer extraction device, a sales destination list creation device, etc., by focusing on the functionality of the server.

[0126] <About inventions extracted from this embodiment> The following describes the features of the inventions extracted from the above-described embodiments.

[0127] (Appendix A1) A sales support device for a financial institution, an acquisition means for acquiring customer transaction data at a financial institution; correlation determination means for determining a correlation between first transaction data of the customer at the time of past purchase of a financial product and second transaction data of the customer at the present time; extraction means for extracting customers whose first transaction data and second transaction data are determined to be related by the correlation determination means; an output means for outputting the customers extracted by the extraction means; A sales support device comprising:

[0128] (Appendix A2) A sales support device for a financial institution, an acquisition means for acquiring at least two or more different types of transaction data of customers at a financial institution; correlation determination means for determining a correlation between a pattern of first transaction data of the customer at the time of past purchase of financial products and a pattern of second transaction data of the customer at the present time; extraction means for extracting the customer for whom the correlation determination means has determined that the pattern of the first transaction data and the pattern of the second transaction data are correlated; an output means for outputting the customers extracted by the extraction means; A sales support device comprising:

[0129] (Appendix A3) when a plurality of customers are extracted by the extraction means, the output means outputs the customers in descending order of the degree of relevance determined by the relevance determination means; A sales support device according to appendix A1 or A2, characterized in that:

[0130] (Appendix A4) when a plurality of customers are extracted by the extraction means, the output means outputs to the terminal a list of the customers arranged in descending order of the degree of relevance determined by the relevance determination means; A sales support device according to appendix A1 or A2, characterized in that:

[0131] (Appendix A5) when outputting the customers in descending order of the degree of association, if the difference in the degree of association between a first customer and a second customer having the second highest degree of association after the first customer is greater than or equal to a predetermined value, the output means indicates a distinction between the first customer and the second customer; A sales support device according to appendix A3 or A4, characterized in that:

[0132] (Appendix A6) Regarding predetermined transaction data among the transaction data of the customer, when the second transaction data is larger than the first transaction data, the correlation determination means causes the second transaction data to match or approximate the first transaction data; A sales support device according to appendix A1 or A2, characterized in that:

[0133] (Appendix A7) Regarding predetermined transaction data among the transaction data of the customer, when the second transaction data is smaller than the first transaction data, the correlation determination means causes the second transaction data to match or approximate the first transaction data; A sales support device according to appendix A1 or A2, characterized in that:

[0134] (Appendix A8) If there are multiple times when the financial product was purchased in the past, the association determination means determines an association between each of the first transaction data at the plurality of past points in time and the second transaction data; the extraction means extracts customers for whom the correlation determination means has determined that the first transaction data and the second transaction data at at least one of the past points in time are correlated; A sales support device according to appendix A1, characterized in that

[0135] (Appendix A9) If there are multiple times when the financial product was purchased in the past, the association determination means determines an association between each of the patterns of the first transaction data at the plurality of past points in time and the pattern of the second transaction data; the extraction means extracts customers for whom the correlation determination means has determined that the second transaction data is correlated with at least one pattern of the first transaction data at the past time point; The sales support device according to appendix A2, characterized in that

[0136] (Appendix A10) The financial product is an investment trust product; A sales support device according to any one of appendices A1 to A9, characterized in that:

[0137] (Appendix B1) A sales support device for a financial institution, an acquisition means for acquiring customer transaction data at a financial institution; an association determination means for determining an association between first transaction data of a first client at the time when the first client previously purchased a financial product and second transaction data of a second client who has not previously purchased the financial product at the present time; extraction means for extracting the second customer whose first transaction data and second transaction data are determined to be related by the correlation determination means; an output means for outputting the second customer extracted by the extraction means; A sales support device comprising:

[0138] (Appendix B2) A sales support device for a financial institution, An acquisition means for acquiring at least two or more different types of transaction data of customers at a financial institution; correlation determination means for determining a correlation between a pattern of first transaction data of a first customer at the time of the past purchase of a financial product and a pattern of second transaction data of a second customer who has not previously purchased the financial product at the present time; extraction means for extracting the second customer whose pattern of the first transaction data and the pattern of the second transaction data are determined to be related by the correlation determination means; an output means for outputting the second customer extracted by the extraction means; A sales support device comprising:

[0139] (Appendix B3) The first customer is a customer whose customer information is related to the customer information of the second customer; A sales support device according to appendix B1 or B2, characterized in that:

[0140] (Appendix B4) The financial product is an investment trust product; A sales support device according to any one of appendices B1 to B3, characterized in that:

[0141] (Appendix C1) A sales support device for a financial institution, An acquisition means for acquiring transaction data of one customer at a financial institution; correlation determination means for determining, for each type of financial product, the correlation between first transaction data of the customer at the time of the previous purchase of the financial product and second transaction data of the customer at the current time; extraction means for extracting financial products of the type determined by the correlation determination means to be related to the first transaction data and the second transaction data; an output means for outputting the types of financial products extracted by the extraction means; A sales support device comprising:

[0142] (Appendix C2) A sales support device for a financial institution, An acquisition means for acquiring at least two or more different types of transaction data of one customer at a financial institution; correlation determination means for determining, for each type of financial product, the correlation between a pattern of first transaction data of the customer at the time of the previous purchase of the financial product and a pattern of second transaction data of the customer at the current time; extraction means for extracting financial products of a type determined by said correlation determination means to be correlated between said first transaction data pattern and said second transaction data pattern; an output means for outputting the types of financial products extracted by the extraction means; A sales support device comprising:

[0143] (Appendix D1) A sales support device for a financial institution, an acquisition means for acquiring transaction data of a second customer at the financial institution; For each type of financial product, first transaction data of the first customer at the time of purchasing the financial product in the past and the second customer at the present time who has not purchased the financial product in the past are correlation determination means for determining a correlation with the second transaction data; extraction means for extracting financial products of the type determined by the correlation determination means to be related to the first transaction data and the second transaction data; an output means for outputting the types of financial products extracted by the extraction means; A sales support device comprising:

[0144] (Appendix D2) A sales support device for a financial institution, an acquisition means for acquiring at least two or more different types of transaction data of one second customer at the financial institution; correlation determination means for determining, for each type of financial product, the correlation between a pattern of first transaction data of a first customer at the time of the previous purchase of the financial product and a pattern of second transaction data of the second customer at the current time who has not previously purchased the financial product; extraction means for extracting financial products of a type determined by said correlation determination means to be correlated between said first transaction data pattern and said second transaction data pattern; an output means for outputting the types of financial products extracted by the extraction means; A sales support device comprising:

[0145] (Appendix E1) A business contact list creation device for a financial institution, An acquisition means for acquiring customer information of a customer at a financial institution; extraction means for extracting a plurality of customers who will be sales targets from among the customers, together with information indicating priority; a creating means for creating a sales destination list information in which the plurality of customers are arranged in descending order of priority; an output means for outputting the sales destination list information to a sales terminal; A sales destination list creation device comprising:

[0146] (Appendix E2) The plurality of customers are customers whose transaction data at the time of past purchase of financial products is related to their current transaction data; A business destination list creation device according to appendix E1, characterized in that:

[0147] (Appendix E3) The plurality of customers are customers whose transaction data patterns at the time of past purchases of financial products are related to those at the present time; A business destination list creation device according to appendix E1, characterized in that:

[0148] (Appendix E4) the priority is a degree of the association; A business destination list creation device according to appendix E2 or 3, characterized in that:

[0149] (Appendix F1) A sales terminal at a financial institution, a transmitting means for transmitting a request to acquire business destination list information to the server device; receiving the sales destination list information in which a plurality of customers to be sales destinations are arranged in order of priority; receiving means for receiving the signal; A business terminal comprising:

[0150] (Appendix F2) The plurality of customers are customers whose transaction data at the time of past purchase of financial products is related to their current transaction data; 10. The business terminal of claim F1,

[0151] (Appendix F3) The plurality of customers are customers whose transaction data patterns at the time of past purchases of financial products are related to those at the present time; 10. The business terminal of claim F1,

[0152] (Appendix F4) the priority is a degree of the association; 3. The business terminal according to claim F2 or F3,

[0153] (Appendix G1) On the computer, a transmitting means for transmitting a request to acquire business destination list information to the server device; a receiving means for receiving the sales destination list information in which a plurality of customers who will be sales destinations are arranged in descending order of priority; a display means for displaying the business destination list information; A display program to make it function as a display program.

[0154] (Appendix G2) The plurality of customers are customers whose transaction data at the time of past purchase of financial products is related to their current transaction data; The display program according to appendix G1, characterized in that

[0155] (Appendix G3) The plurality of customers are customers whose transaction data patterns at the time of past purchases of financial products are related to those at the present time; The display program according to appendix G1, characterized in that

[0156] (Appendix G4) the priority is a degree of the association; The display program according to appendix G2 or 3, characterized in that

[0157] (Appendix H1) A server device of a financial institution, Acquire customer transaction data at financial institutions, Extracting a customer if the acquired customer transaction data is similar to the customer's transaction data at the time of the customer's past purchase of a financial product; A server device characterized by:

[0158] (Appendix H2) A server device of a financial institution, Acquire multiple types of customer transaction data from financial institutions, If the status of the acquired transaction data of multiple types of customers is similar to the status of transaction data of the customer at the time of purchasing financial products in the past, extracting the customer into a sales target list; A server device characterized by:

[0159] (Appendix H3) When a plurality of customers are extracted from the sales destination list, extracting the plurality of customers in descending order of the degree of similarity; The server device according to claim H1 or H2, [Explanation of symbols]

[0160] 10 Financial Institution System 20 Sales Promotion Support System 30 Management Server 40 Processing DB 50 Sales Terminals 70 Network 100 Sales Support System 301 Data Acquisition Department 302 Feature Calculation Unit 303 Related Judgment Department 304 Extraction part 305 Sales List Creation Department 306 Output section

Claims

1. an acquisition means for acquiring customer transaction data that is updated as the customer uses financial services; a relevance level acquisition means for acquiring a relevance level, which is a degree of likelihood that a candidate customer will purchase a financial product, using at least the feature values ​​of the transaction data of a customer who has previously purchased a financial product and the feature values ​​of the last updated transaction data of a candidate customer; a sales destination priority list creation means for listing the candidate customers whose relevance levels have been acquired by the relevance level acquisition means in order of priority based on the relevance levels; A sales support device comprising:

2. A classification means for classifying multiple target customers into groups based on whether or not they have experience purchasing financial products, based on customer transaction data; a relevance acquisition means for acquiring a relevance, which is the degree of likelihood of a candidate customer purchasing a financial product, by using machine learning between feature values ​​converted into vectors of transaction data of at least customers who have purchased financial products in the past, among the transaction data of customers classified by the classification means, and feature values ​​converted into vectors of transaction data of a candidate customer; a sales destination priority list creation means for listing the candidate customers whose relevance levels have been acquired by the relevance level acquisition means in order of priority based on the relevance levels; A sales support device having the above.

3. an acquisition means for acquiring customer transaction data that is updated as the customer uses financial services; a relevance acquisition means for acquiring a relevance, which is the degree of likelihood of a candidate customer purchasing a financial product, by using machine learning of at least the feature values ​​converted into vectors of the transaction data of customers who have purchased financial products in the past and the feature values ​​converted into vectors of the last updated transaction data of a candidate customer; a sales destination priority list creation means for listing the candidate customers whose relevance levels have been acquired by the relevance level acquisition means in order of priority based on the relevance levels; A sales support device comprising:

Images