TWI674544B - Intelligent product marketing method and system - Google Patents

Abstract

In a smart product marketing method and system, a modeling server uses the preference score estimation model and the preference probability assessment model to estimate each customer's information based on historical financial transaction data of multiple customers stored by a data server. Multiple preference scores and multiple preference probabilities corresponding to various financial products, respectively, to generate the preference score results and preference probability results for these customers. The marketing server unit generates a product recommendation result including multiple recommended financial products selected from multiple financial products for each customer according to the preference probability result, and will indicate the customer when an electronic device used by the customer is connected to it The product information of these recommended financial products is transmitted to the electronic device so that the electronic device displays the product information.

Description

Intelligent product marketing method and system

The present invention relates to the marketing of financial products, and more particularly to a method and system for marketing smart products of financial products.

In existing financial institutions such as banks, there are often a number of financial consultants who are specialized in the marketing of financial products. Traditionally, financial consultants usually market financial products to a customer base with more than 3 million bank assets. In the marketing process, product recommendations and sales are often targeted at familiar customers in the customer base. However, for unfamiliar customers, due to the lack of relevant customer analysis data, they may not be able to accurately understand the financial products that the customers want. Considerable labor and time costs, but still fail to improve the success rate of marketing.

Therefore, there is still much room for improvement in traditional financial product marketing methods.

Therefore, an object of the present invention is to provide a smart product marketing method and system for financial products, which can overcome the shortcomings of conventional techniques.

Therefore, an aspect of the present invention provides a smart product marketing method. The smart product marketing method is implemented by a smart product marketing system. The product marketing system includes a data server, a modeling server, and a sales server unit. The data server stores a plurality of customer reference data corresponding to multiple customers, and each customer reference data includes information related to these customers. One of them corresponds to the customer's historical financial transaction data. The smart product marketing method includes the following steps: (A) By using the modeling server, according to the historical financial transaction data of each customer reference data stored by the data server, a method related to transaction frequency, transaction amount, and The predetermined preference score estimation model of the transaction balance estimates a plurality of preference scores corresponding to a variety of financial products for each client to generate a preference score result including the preference scores of each client; (B) using the The modeling server uses a predetermined preference probability evaluation model according to the preference score result to evaluate the preference probability of each customer corresponding to these financial products, so as to generate a preference including each customer. Probability preference result; (C) by the marketing servo unit, according to the preference probability result, a product recommendation result including a variety of recommended financial products for each customer is generated, wherein each customer's recommended financial products are Selected from these financial products; and (D) when an electronic device used by one of these customers is connected to the marketing servo unit, The marketing by the servo unit, indicating that corresponding to the client such kind of financial recommended for product information transmitted to the electronic device, the electronic device so that the product information displayed thereon.

Therefore, another aspect of the present invention provides an intelligent product marketing system. The intelligent product marketing system includes a data server, a modeling server, and a sales server unit.

The data server stores a plurality of customer reference data corresponding to multiple customers, and each customer reference data includes historical financial transaction data related to one of the corresponding customers of the customers.

The modeling server is connected to the data server to receive the customer reference data from the data server, and includes a preference score estimation module and a preference probability estimation module. The preference score estimation module uses a predetermined preference score estimation model related to transaction frequency, transaction amount, and transaction balance based on the historical financial transaction data of each customer's reference data to estimate that each of the customers corresponds to a variety of Preference scores for financial products to generate a preference score result that includes each customer's preference scores. The preference probability evaluation module is electrically connected to the preference score estimation module to receive the preference score result, and according to the preference score result, a predetermined preference probability assessment model is used to evaluate a plurality of each customer corresponding to the preference score. Preferences of various financial products to generate a preference probability result that includes each customer's preferences.

The marketing servo unit includes an access server, which is connected to the modeling server to receive the preference probability result, and generates a product recommendation result including multiple recommended financial products for each customer according to the preference probability result. ,its These recommended financial products for each of the customers are selected from these financial products. When an electronic device used by one of the customers is connected to the channel server, the channel server transmits product information indicating the customers' recommended financial products to the electronic device, so that the electronic device will The product information is displayed on it.

The effect of the present invention is that the marketing servo unit can automatically and appropriately obtain a variety of recommended financial products corresponding to each customer according to the preference probability result generated by the modeling server, and can timely use an electronic device to The product information of these recommended financial products is displayed to customers, so that the purpose of marketing multiple products can be achieved quickly with the lowest marketing cost.

100‧‧‧ Intelligent Product Marketing System

1‧‧‧Data Server

2‧‧‧ modeling server

21‧‧‧Preference score estimation module

22‧‧‧Preference probability evaluation module

23‧‧‧Grouping Module

3‧‧‧ Marketing Servo Unit

31‧‧‧passage server

32‧‧‧Select Server

33‧‧‧ Marketing Server

4‧‧‧ electronic device

5‧‧‧ end of use

6‧‧‧Customer

200‧‧‧Communication Network

S1-S10‧‧‧step

S31-S34‧‧‧step

Other features and effects of the present invention will be clearly presented in the embodiment with reference to the drawings, in which: FIG. 1 is a block diagram illustrating an embodiment of the intelligent product marketing system of the present invention; FIG. 2 is a flow chart FIG. Illustrates how the embodiment executes a smart product marketing program; and FIG. 3 is a flowchart illustrating how the marketing server of the embodiment obtains a selection result.

Referring to FIG. 1, an embodiment of a smart product marketing system 100 according to the present invention includes It includes a data server 1, a modeling server 2, and a sales server unit 3.

The data server 1 stores a plurality of customer reference data respectively corresponding to a plurality of customers. In this embodiment, each customer reference data includes, for example, customer basic data, attribute data, exclusion condition data, and historical financial transaction data related to a corresponding customer among these customers, wherein the customer basic data includes, for example, the corresponding Customer's identity information, holding account information and contact information. The attribute data includes information such as age, gender, occupation, education, marital status, holding account, transaction behavior, etc. Historical financial transaction data includes, for example, all deposit transactions , Foreign currency fixed deposit transactions, stock transactions, fund transactions, and / or insurance transactions, etc., the exclusion condition data includes, for example, information related to at least one of customer-customized conditions, customer credit conditions, and regulatory conditions. For example, the customer's self-defined conditions include, for example, an agreement not to accept joint marketing and / or a statement that they do not accept advertising marketing, etc. The customer's credit conditions include, for example, a blacklisted customer on the marketing list, a refusal to a current account, a rebate customer, a deposit Frozen customers, court false seizure customers, debt negotiation customers, and / or debt settlement customers, etc. The regulations include, for example, age restrictions on purchasers of different financial products. It is worth noting that the content of each customer reference material is not limited to the content listed above, but in application, the content of the data can be increased or decreased according to actual needs.

The modeling server 2 is connected to the data server 1 to receive the client reference data from the data server 1. In this embodiment, the modeling server 2 includes a preference score estimation module 21 and an electrical connection to the preference score estimation module 21 Preference probability module 22, and a clustering module 23.

In this embodiment, the marketing servo unit 3 includes, for example, a path server 31 connected to the modeling server 2, a selection server 32 connected to the modeling server 2, and a selection server 32 connected to the selection server 32. Marketing server 33.

In this embodiment, the path server 31, the selection server 32, and the marketing server 33 can be connected to the data server 1 via a communication network 200 such as the Internet.

The following describes how the smart product marketing system 100 executes a smart product marketing process with reference to FIGS. 1 and 2. The smart product marketing process includes the following steps.

First, in step S1, the preference score estimation module 21 uses a predetermined preference score estimation model related to transaction frequency, transaction amount and transaction balance to estimate each A plurality of customers respectively correspond to the preference scores of various financial products to generate a preference score result including the preference scores of each client. In this embodiment, these types of financial products include, but are not limited to, seven types of financial products such as foreign exchange live deposits, fixed foreign exchange deposits, funds, insurance, gold passbooks, credit cards, and credits.

In step S2, the preference probability evaluation module 22 receives the preference score result from the preference score estimation module 21, and uses, for example, a predetermined preference probability assessment model to evaluate the number of customers based on the preference score result Corresponding to These seven financial products have a preference probability to produce a preference probability result that includes each customer's preference probability.

In step S3, the path server 31 receives the preference probability result of 2 from the modeling server, and generates a product recommendation result including multiple recommended financial products for each customer according to the preference probability result. It is worth noting that the recommended financial products of each client are selected from those who have a relatively high probability of preference among the seven financial products. For example, the seven financial products correspond to the top three Three types of financial products with a high probability of preference are used as such recommended financial products, but not limited thereto.

In step S4, in this embodiment, when an electronic device 4 such as a mobile phone or a personal computer used by one of the customers 6 is connected to the path server 31 via the communication network 200, for example, the path servo The device 31 transmits the product information of the recommended financial products of the customer 6 to the electronic device 4 so that the electronic device 4 displays the product information on it, thereby marketing the product to the customer 6 The purpose of recommending financial products. However, in other embodiments, the electronic device 4 may also be, for example, a terminal set in the banking institution, and the path server 31 detects the operation of the client 6 in the terminal, and uses the product. The data is transmitted to the terminal so that the terminal displays the product information on it. In addition, the bank financial consultant can also use the channel server 31 to recommend the product generated by each customer to achieve the goal of personalized financial services for the customer, and jointly achieve the purpose of marketing the product.

On the other hand, in step S5 following step S2, the grouping module 23 uses, for example, a predetermined grouping model to divide the customers into a plurality of customer groups with different tendencies, based on the reference customer data. In order to produce a grouping result indicating the customers contained in each customer group. In this embodiment, the customer groups may include, for example, an elite customer group, a silver customer group, and a potential customer group, but not limited thereto. The clients of this elite customer group tend to, for example, have the relatively highest assets, relatively high socioeconomic status, act as corporate leaders, have close dealings with the banking institution, and have purchased relatively many financial products, but not here limit. The customers of this silver-haired customer group tend to have, for example, relatively second-to-highest assets, relatively oldest women, relatively conservative trading behavior, purchase of fixed deposit or savings (capital-protected) financial products, etc., but not limited to this. Clients of this potential customer group, for example, tend to own the third-highest assets and prefer to buy funds and women on a regular basis, but not so much. Then, the modeling server 2 transmits the preference probability result and the clustering result to the selection server 32.

In step S6, the selection server 32 receives the grouping result and the preference probability result from the modeling server 2, and according to a target marketing product belonging to one of the seven financial products, from the A target customer group is selected from the customer groups of the grouping result. It is worth noting that there is a relatively high correlation between the tendency of the target customer group and the target marketing product. The selection server 32 regards the customers included in the target customer group as a plurality of candidate customers, and extracts a plurality of results from the preference probability and respectively corresponds to the candidate customers for the target marketing product. The probability of preference of a financial product is used as a plurality of candidate preferences.

In step S7, the selection server 32 calculates the exclusion condition data related to each of the candidate customers according to the candidate preference probability, a probability threshold value related to the distribution of the candidate preference probability, and Wait for the candidate customers to select multiple target customers to generate a selection result that includes the customer list information of the target customers. Hereinafter, a procedure performed by the selection server 32 to obtain the selection result is described with reference to FIGS. 1 and 3. The procedure includes the following steps.

In step S71, the selection server 32 selects a plurality of candidate customers from the candidate customers according to the candidate preference probability and the probability threshold, wherein the candidate preference probability corresponding to the selected candidate customers is greater than The probability threshold. For example, if the candidate preference probabilities are approximately evenly distributed between 0 and 100%, the probability threshold may be determined by the number of target customers to be marketed. However, if the candidate preference probabilities are not evenly distributed, they are roughly distributed in When there is a specific probability range, the probability threshold may be appropriately selected as the probability threshold according to the number of target customers to be marketed, but is not limited thereto.

In step S72, the selection server 32 transmits an exclusion condition request related to the selected candidate customers to the data server 1 via the communication network 200.

Therefore, when the data server 1 receives the request for the exclusion condition from the selection server 32, the data server 1 includes a list of the selected candidate customers. The exclusion condition replies of the exclusion condition data and contact information of each of the households are transmitted to the selection server 32 via the communication network 200.

In step S73, the selection server 32, after receiving the exclusion condition response from the data server 1, responds to the exclusion condition data of each selected candidate customer included according to the exclusion condition, and determines the selected one. Whether candidate customers must be excluded. For example, if the exclusion condition data of a selected candidate customer indicates that the current account is rejected, the selected candidate customer must be excluded.

In step S74, the selection server 32 selects the selected candidate customers as those that do not need to be excluded as the target customers, and generates the selection result according to the target customers. It is worth noting that, in this embodiment, the selection result includes not only the customer list information of the target customers, but also the contact information of the target customers. The selection server 32 transmits the selection result to the marketing server 33.

In step S8, after receiving the selection result from the selection server 32, the marketing server 33 transmits a product message related to the target marketing product to a plurality of target customers according to the selection result. The designated use end 5 so that the use end 5 displays the product information thereon, thereby achieving the purpose of marketing these target marketing products to these target customers (not shown). In addition, the bank financial consultant may also use the selection result received by the marketing server 33 to achieve the goal of personalizing financial services to the customer and jointly achieve the purpose of marketing these various target marketing products.

Therefore, these customers can easily obtain appropriate product information of recommended financial products and / or product information of target marketing products through the above-mentioned marketing methods.

In step S9, when any one of the channel server 31 and the marketing server 33 receives a marketing result related to any customer, the marketing result is transmitted to the data server via the communication network 200. 1. In this embodiment, the marketing result includes, but is not limited to, data related to, for example, the purchase of a customer, the sale of a product, whether the sale was successful, the customer response to the sale process, and / or the reason for the failure of the sale.

Finally, in step S10, when the data server 1 receives the marketing result from the marketing server unit, it updates the stored customer reference data according to the marketing result, as a future transaction behavior for the customer For analytical purposes.

In summary, the marketing servo unit 3 can automatically and appropriately obtain a variety of recommended financial products corresponding to each customer according to the preference probability result generated by the modeling server 2, and can timely use an electronic device 4 Display the product information of these kinds of recommended financial products to customers, so that the purpose of marketing multiple products can be achieved quickly with the lowest marketing cost. In addition, since the modeling server 2 also effectively groups these customers, the marketing server unit 3 can also select the target customer group from the customer groups based on the target marketing product and further select a target marketing product suitable for marketing Target customers, and can also automatically provide product information to target customers by displaying them on the end 5 designated by the target customer, which will help increase the probability of successful product sales. On the other hand, the bank financial adviser is In the case of the customer's product recommendation results, it can indeed achieve the goal of personalizing financial services to the customer and jointly achieve the purpose of product marketing. Therefore, the purpose of the invention can be achieved.

However, the above are only examples of the present invention. When the scope of implementation of the present invention cannot be limited by this, any simple equivalent changes and modifications made according to the scope of the patent application and the contents of the patent specification of the present invention are still Within the scope of the invention patent.

Claims (10)

A smart product marketing method is implemented by a smart product marketing system. The smart product marketing system includes a data server, a modeling server, and a sales server unit. The data server stores multiple data corresponding to multiple data servers. Customer reference materials for each customer. Each customer reference material contains historical financial transaction data and attribute data related to one of these customers. The smart product marketing method includes the following steps: (A) using the modeling server According to the historical financial transaction data of each customer reference data stored by the data server, a predetermined preference score estimation model related to transaction frequency, transaction amount, and transaction balance is used to estimate a plurality of separate correspondences for each customer Generate preference score results including preference scores of each customer from various financial product preference scores; (B) using the modeling server to evaluate the model using a predetermined preference probability based on the preference score results, Evaluate the preference probability of each customer corresponding to these financial products to generate a Preference probability results including the preference probability of each customer; (C) With the marketing servo unit, according to the preference probability result, a product recommendation result including multiple recommended financial products for each customer is generated, where each customer These recommended financial products are selected from these financial products; (D) When an electronic device used by one of these customers is connected to the marketing servo unit, the marketing servo unit will indicate the corresponding The product information of the recommended financial products from the customer is transmitted to the electronic Device so that the electronic device displays the product information thereon; (E) by means of the modeling server, based on the customer reference data stored by the data server, using a predetermined clustering model, The customer is divided into a plurality of customer groups with different tendencies, to produce a grouping result indicating the customers included in each customer group; (F) by means of the marketing servo unit, according to one of these types of financial products A target marketing product for a financial product, a target customer group is selected from the customer groups of the segmentation result, wherein the target customer group's tendency has a relatively high correlation with the target marketing product, and the target customer is The customers included in the group are regarded as multiple candidate customers; (G) by the marketing servo unit, multiple results are extracted from the preference probability results and respectively correspond to the type of the candidate customers for the target marketing product. The preference probability of a financial product is taken as a plurality of candidate preference probability, and according to the candidate preference probability, a probability threshold related to the distribution of the candidate preference probability, and the candidate probability For each of the customers' exclusion conditions, multiple target customers are selected from the candidate customers to generate a selection result that includes the customer list information of the target customers; and (H) by the marketing server unit, according to the As a result of the selection, a product message related to the target marketing product is transmitted to a plurality of users designated by the target customers, so that the users display the product message thereon. The smart product marketing method according to claim 1, wherein, in step (C), for each customer, the preference probability corresponding to these types of recommended financial products is relatively high. The smart product marketing method according to claim 1, wherein in step (G), the exclusion condition data related to each candidate customer includes at least one of a customer-customized condition, a customer credit condition, and a regulatory condition data of. According to the smart product marketing method described in claim 1 or 3, each customer reference data stored by the data server further includes exclusion condition data related to the corresponding customer, wherein step (G) includes the marketing server unit. The following sub-steps are performed: (G1) According to the candidate preference probability and the probability threshold, a plurality of candidate customers are selected from the candidate customers, and the candidate preference probability corresponding to the selected candidate customers is greater than The probability threshold; (G2) sends a request for exclusion conditions related to the selected candidate customers to the data server; (G3) upon receiving a data server that includes the selected candidate customers After replying to the exclusion condition of one of the exclusion condition data, determine whether the selected candidate customer must be excluded according to the exclusion condition data of each selected candidate customer received; and (G4) the selected candidate customer The post-selected customers are determined as those who do not need to be excluded as the target customers, and the selection result is generated according to the target customers. According to the smart product marketing method described in claim 1, after step (H), the method further includes the following steps: (I) The marketing server unit transmits a marketing result related to any of these customers to The data server; and (J) With the data server, when receiving the marketing result from the marketing servo unit, update the stored customer reference data according to the received marketing result. An intelligent product marketing system includes: a data server storing a plurality of customer reference data corresponding to multiple customers, each customer reference data including historical financial transaction data and attributes related to one of these corresponding customers Data; a modeling server connected to the data server to receive the customer reference data from the data server, and including a preference score estimation module, based on the historical financial transactions of each customer reference data Data, using a predetermined preference score estimation model related to transaction frequency, transaction amount, and transaction balance to estimate the preference scores of each customer corresponding to multiple financial products to generate a preference for each customer The score preference score result, a preference probability evaluation module, is electrically connected to the preference score estimation module to receive the preference score result, and according to the preference score result, a predetermined preference probability assessment model is used to evaluate each customer's Multiple preference probabilities corresponding to these financial products, respectively, to generate a The customer ’s preference probability results, and a clustering module. Based on the customer reference data stored by the data server, a predetermined clustering model is used to divide the customers into multiple ones with different tendencies. Customer groups to generate a grouping result indicating the customers included in each customer group; and marketing servo units, including A channel server connected to the modeling server to receive the preference probability result, and generate a product recommendation result including multiple recommended financial products for each customer according to the preference probability result, wherein The recommended financial products are selected from these financial products, and when an electronic device used by one of the customers is connected to the channel server, the product information indicating the recommended financial products of the customer is transmitted to The electronic device so that the electronic device displays the product information thereon, a selection server connected to the modeling server to receive the clustering result and the preference probability result, and according to one of the financial products belonging to A target marketing product of a financial product, a target customer group is selected from the customer groups of the grouping result, and there is a relatively high correlation between the target customer group's tendency and the target marketing product. The selection server The customers included in the target customer group are regarded as a plurality of candidate customers, and a plurality of results are extracted from the preference probability result and respectively correspond to The candidate customer's preference probability for the financial product to which the target marketing product belongs is used as a plurality of candidate preference probability, and according to the candidate preference probability, a probability threshold value related to the distribution of the candidate preference probability, and Exclusion data for each of these candidate customers, and multiple target customers are selected from the candidate customers to generate a selection result containing the customer list information of the target customers, and a sales server connected to the selection server The device is used to receive the selection result, and according to the selection result, send a product message related to the target marketing product to a plurality of agents respectively designated by the target customers. Users so that they can display the product message on them. The intelligent product marketing system according to claim 6, wherein, for each customer, the preference probability corresponding to the recommended financial products is relatively high. The smart product marketing system according to claim 6, wherein the exclusion condition data related to each candidate customer includes data related to at least one of a customer's custom condition, a customer credit condition, and a regulatory condition. The intelligent product marketing system according to claim 6 or 8, wherein: each customer reference data stored by the data server further includes exclusion condition data related to the corresponding customer; the selection server is connected to the data server, And according to the candidate preference probability and the probability threshold, a plurality of candidate customers are selected from the candidate customers, and the candidate preference probability corresponding to the selected candidate customers is greater than the probability threshold, and a correlation is transmitted. When the exclusion conditions of the selected candidate customers are requested to the data server, the data server, upon receiving the exclusion condition requests from the selection server, responds to the exclusion condition request and includes a The exclusion condition reply of the exclusion condition data of each of the candidate customers is transmitted to the selection server; and when the selection server receives the exclusion condition reply from the data server, the selection server is based on each selection received The candidate condition information of the candidate customers determines whether the selected candidate customers must be excluded, and the selected candidate is selected. Households which decided not to have to be excluded as those who target customers, and in accordance with these target customers, resulting in the selection result. The intelligent product marketing system according to claim 6, wherein: the channel server and the marketing server are also connected to the data server, and any one of the channel server and the marketing server will be related to the customers The marketing result of any one is transmitted to the data server; and the data server, upon receiving the marketing result from any of the channel server and the marketing server, based on the marketing result received To update these stored customer references.