Program, customer extraction device, and customer extraction method

Abstract

The purpose of the present disclosure is to provide a program capable of appropriately visualizing potential needs of customers. The program according to the present disclosure causes a computer to execute: a product feature acquisition step for acquiring product feature information indicating features of a target product; a customer attribute acquisition step for acquiring customer attribute information including customer preferences for each of a plurality of customers; a customer extraction step for extracting a target customer assumed to be interested in the target product from the plurality of customers on the basis of the product feature information and the customer attribute information; and a size specification step for classifying the target customer into groups corresponding to respective customer segments and specifying the size of each group.

Description

Program, customer extraction apparatus, and customer extraction method

[0001] The present disclosure relates to a program, a customer extraction apparatus, and a customer extraction method.

[0002] Technologies for collecting information on customer preferences and using the same for marketing are known. As a related technology, Patent Document 1 discloses a sales support system capable of extracting prospective customers for cross-selling, up-selling, and product replacement based on product properties and customer purchase histories.

[0003] Japanese Unexamined Patent Publication No. 2002-334201

[0004] Examples of information on customer preferences include the history of products purchased by customers and purchase frequencies. Such information indicates, for example, when and what products a customer purchased, and can be used for analysis of purchase trends and optimization of targeted advertising. However, Patent Document 1 does not specifically disclose what kind of data should be analyzed in what manner to appropriately visualize customers' latent needs.

[0005] An example of the object of the present disclosure is to provide a program, a customer extraction apparatus, and a customer extraction method capable of appropriately visualizing customers' latent needs in view of the above-described problem.

[0006] A program according to one aspect of the present disclosure causes a computer to execute: a product feature acquisition step of acquiring product feature information indicating features of a target product; a customer attribute acquisition step of acquiring customer attribute information including customer preferences for each of a plurality of customers; a customer extraction step of extracting target customers who are assumed to be interested in the target product from the plurality of customers based on the product feature information and the customer attribute information; and a size identification step of classifying the target customers into groups each corresponding to a customer segment and identifying the size of each group.

[0007] A customer extraction device relating to one aspect of this disclosure includes: a product feature acquisition unit that acquires product feature information indicating the characteristics of a target product; a customer attribute acquisition unit that acquires customer attribute information, including customer preferences, for each of a plurality of customers; a customer extraction unit that extracts target customers who are expected to be interested in the target product from the plurality of customers based on the product feature information and the customer attribute information; and a scale identification unit that classifies the target customers into groups according to customer segment and identifies the size of each group.

[0008] A customer extraction method relating to one aspect of this disclosure includes: a product feature acquisition step of acquiring product feature information indicating the characteristics of the target product; a customer attribute acquisition step of acquiring customer attribute information, including customer preferences, for each of a plurality of customers; a customer extraction step of extracting target customers who are expected to be interested in the target product from the plurality of customers based on the product feature information and the customer attribute information; and a scale identification step of classifying the target customers into groups according to customer segment and identifying the size of each group.

[0009] One example of the effects of the program, customer extraction device, and customer extraction method described herein is the ability to appropriately visualize customers' latent needs.

[0010] Figure 1 is a block diagram showing the configuration of the customer extraction device. Figure 2 is a flowchart showing the processing flow performed by the customer extraction device. Figure 3 is a block diagram showing the configuration of the information provision system. Figure 4 is a block diagram showing the configuration of the customer extraction device. Figure 5 shows an example of an input screen displayed on the display unit. Figure 6 shows an example of customer attribute information. Figure 7 shows an example of an analysis results screen showing the analysis results in the scale determination unit. Figure 8 shows an example of an analysis results details screen showing detailed analysis results. Figure 9 shows an example of a cluster addition screen for adding a unique customer cluster. Figure 10 is a block diagram showing the configuration of the communication terminal. Figure 11 is a block diagram showing the configuration of the customer information management device. Figure 12 is a flowchart showing the processing flow performed by the customer extraction device. Figure 13 is a block diagram illustrating the hardware configuration of a computer that implements the customer extraction device, etc.

[0011] Embodiments of the present disclosure will be described in detail below with reference to the drawings. In each drawing, the same or corresponding elements are denoted by the same reference numerals. For clarity of explanation, redundant explanations will be omitted where necessary.

[0012] <Embodiment 1> (Customer Extraction Device 100) Figure 1 is a block diagram showing the configuration of the customer extraction device 100 according to the present disclosure. The customer extraction device 100 comprises a product feature acquisition unit 101, a customer attribute acquisition unit 102, a customer extraction unit 103, and a scale determination unit 104.

[0013] The product feature acquisition unit 101 acquires product feature information that indicates the characteristics of the target product. The customer attribute acquisition unit 102 acquires customer attribute information, including customer preferences, for each of multiple customers. The customer extraction unit 103 extracts target customers who are expected to be interested in the product from the multiple customers based on the product feature information and customer attribute information. The scale determination unit 104 classifies the target customers into groups according to customer segment and determines the scale of each group.

[0014] The customer extraction device 100 includes a processor, memory, and storage device (not shown). The storage device stores a computer program on which the processing described herein is implemented. The processor can load the computer program from the storage device into memory and execute the computer program. In this way, the processor realizes the functions of the product feature acquisition unit 101, the customer attribute acquisition unit 102, the customer extraction unit 103, and the scale determination unit 104.

[0015] Alternatively, the product feature acquisition unit 101, the customer attribute acquisition unit 102, the customer extraction unit 103, and the scale determination unit 104 may each be implemented with dedicated hardware. Furthermore, some or all of each component may be implemented by general-purpose or dedicated circuits, processors, etc., or combinations thereof. These may be configured by a single chip or by multiple chips connected via a bus. Some or all of each component may be implemented by a combination of the aforementioned circuits, etc., and programs.

[0016] (Processing of the customer extraction device 100) The processing performed by the customer extraction device 100 will be explained with reference to Figure 2. Figure 2 is a flowchart showing the flow of processing performed by the customer extraction device 100.

[0017] First, the product feature acquisition unit 101 acquires product feature information (S1). Next, the customer attribute acquisition unit 102 acquires customer attribute information for each of the multiple customers (S2). Subsequently, the customer extraction unit 103 extracts target customers from the multiple customers (S3). Next, the scale identification unit 104 classifies the target customers into groups according to customer segment (S4). The scale identification unit 104 also identifies the scale of each group (S5).

[0018] As described above, the customer extraction device 100 related to this disclosure acquires product characteristic information and customer attribute information, and based on this information, extracts target customers who are expected to be interested in the product from among multiple customers. The customer extraction device 100 also classifies the target customers into groups according to customer segment and identifies the size of each group. With this configuration, the customer extraction device 100 can appropriately visualize the latent needs of customers.

[0019] <Embodiment 2> Next, Embodiment 2 will be described. Embodiment 2 is a specific example of Embodiment 1 described above. Figure 3 is a block diagram showing the configuration of the information provision system 1 according to this disclosure. The information provision system 1 is a system capable of providing information to a user. The user is a user of the information provision system 1. The user may be a person who uses the communication terminal 20.

[0020] The information provision system 1 comprises a customer extraction device 10, a communication terminal 20, and a customer information management device 30. The customer extraction device 10, the communication terminal 20, and the customer information management device 30 can communicate with each other via a network N. The network N includes, for example, an internet connection or a wireless communication network, but the type of communication is not limited to these. The customer extraction device 10, the communication terminal 20, and the customer information management device 30 each have a communication unit (not shown in Figure 3) and communicate with each other via the network N.

[0021] (Customer Extraction Device 10) The customer extraction device 10 is a device that can extract customers that meet the conditions specified by the user and determine their scale. For example, the customer extraction device 10 obtains the conditions desired by the user from a communication terminal 20 and executes predetermined processing. The customer extraction device 10 may be composed of a server device or the like.

[0022] Figure 4 is a block diagram showing the configuration of the customer extraction device 10. The customer extraction device 10 is an example of the customer extraction device 100 described above. The customer extraction device 10 includes a product feature acquisition unit 11, a customer attribute acquisition unit 12, a customer extraction unit 13, a scale determination unit 14, and a communication unit 15.

[0023] The product feature acquisition unit 11 is an example of the product feature acquisition unit 101 described above. The product feature acquisition unit 11 acquires product feature information that indicates the characteristics of the target product. The target product is a product that is the target of marketing. The target product indicates a product that will serve as the criterion for customer selection. The genre of the target product is arbitrary. Target products may include, for example, food, beverages, toothpaste, cosmetics, accessories, books, or home appliances. The target product can be any product for which sales promotion is being considered. Furthermore, the target product is not limited to goods but also includes services. For example, it may include services in various fields such as streaming services and wealth management.

[0024] Product feature information is information that describes the characteristics of the product in question. This information may include, for example, the nature, category, intended use, ingredients, or price range of the product. Product feature information is used to analyze customer interest and to formulate appropriate marketing strategies.

[0025] For example, the product feature acquisition unit 11 acquires product feature information via the display screen of the communication terminal 20. The communication terminal 20 is equipped with a display unit 22 that displays various display screens for receiving information in the information provision system 1.

[0026] The product feature acquisition unit 11 can acquire product feature information by receiving input on the display screen shown on the display unit 22. For example, the product feature acquisition unit 11 may acquire product feature information by receiving text input from a predetermined input field on the screen and performing natural language processing on the input text. The product feature acquisition unit 11 acquires product feature information by using natural language processing to extract product features from the input text.

[0027] Product feature information may include, for example, product feature tags that indicate the characteristics of the product. Product feature tags may include, but are not limited to, information indicating the product's category, brand, function, use, ingredients, or price range. Various types of information indicating the characteristics of a product may be included as product feature tags. Product feature information for a single product may include multiple product feature tags. Product feature information may also be obtained based on whether or not a word in the input text matches a predefined product feature tag. Alternatively, product feature tags similar to a word may be identified based on the feature quantity of that word.

[0028] Furthermore, the product feature acquisition unit 11 may accept the designation of an information source containing information about the target product and acquire product feature information by extracting the features of the target product contained in the designated information source. The information source containing information about the target product is a data source that provides detailed information about the target product. The information source may be, for example, a web page or database that describes the details of the target product. The information source may include the manufacturer's official website, the product page on an e-commerce site, or a product catalog. The information source may be designated by entering a URL, or by specifying a specific database or file. In addition, the process for designating the information source may be accepted when the display screen shown on the display unit 22 accepts input. Alternatively, the designation may be made before or after the input is accepted on the display screen shown on the display unit 22. Furthermore, the display screen shown on the display unit 22 may perform processing to identify the information source based on the input information.

[0029] Figure 5 shows an example of input screen A displayed on the display unit 22. Input screen A is a screen for users to input the necessary information in order to generate a customer cluster from product or service information.

[0030] As shown in Figure 5, input screen A includes a name field a1, a product features input field a2, a URL field a3, and an analysis start button a4. The name field a1 is for entering the name of the product or service. The product features input field a2 is for entering the product's features. The URL field a3 is for entering the URL of the webpage containing information about the target product. The analysis start button a4 is a button for starting the analysis of customer clusters.

[0031] Furthermore, input screen A may be provided with a means for selecting the number of customer clusters to generate or the size of the customer clusters. As an example, Figure 5 displays three radio buttons for each item. By selecting one of the radio buttons associated with each item, the conditions for analyzing customer clusters or outputting the customer cluster analysis results can be set.

[0032] The "number of customer clusters to generate" is information used to control how many customer clusters are created. By setting this information, the number of clusters to be output by the scale determination unit 14 can be set. Naturally, the product feature acquisition unit 11, customer attribute acquisition unit 12, and customer extraction unit 13 may also refer to this information and change the processing content. As an example, the scale determination unit 14 classifies customers into the specified number of clusters.

[0033] Furthermore, "customer cluster size" is information used to specify the size of the customer clusters to be output. By setting this information, the cluster size output by the size determination unit 14 can be adjusted. Naturally, the product feature acquisition unit 11, customer attribute acquisition unit 12, and customer extraction unit 13 may also refer to this information and change their processing content. As an example, the size determination unit 14 outputs size information only for customer clusters of the selected size.

[0034] Furthermore, an icon indicating that the system recommends the option can be displayed next to the recommended radio button. The location where such an icon is displayed may be dynamically changed depending on, for example, the environment in which the system is running or the congestion of the communication line. More specifically, the position where the icon is displayed is controlled based on the relationship between the computer's resources and the resources used by the customer extraction device 10. If the customer extraction device 10 uses or is expected to use X% of the computer's memory, the icon is displayed in association with the option that has the lowest load among the options.

[0035] Note that X is used as an example threshold to indicate high memory usage, and this can be arbitrarily set and changed by a computer or customer extraction device 10, etc. In Figure 5, the "Number of customer clusters to generate" item displays three options: 10, 15, and 20, and an icon is displayed indicating the recommended option for the 10 clusters, which is the smallest number to generate.

[0036] The relationship between computer resources and the resources used by the customer extraction device 10 may be based on indicators such as CPU (Central Processing Unit) usage, GPU (Graphics Processing Unit) usage, job queue dwell time, or predicted processing time, or a combination of these. Furthermore, when controlling the display position based on the congestion status of the communication line, indicators such as RTT (Round-Trip Time), throughput, retransmission rate, queue length, number of timeouts, or a combination of these may be used. More specifically, when these values ​​indicate that the communication line is congested or the communication speed is slow, icons are associated with and displayed for options that can further reduce the amount of processing or information (options that result in a relatively smaller number of outputs or reduce the number of outputs).

[0037] In this way, by enabling selection to control the amount of information processed or output when analyzing customer clusters or outputting results based on those analyses, it is possible to perform the desired customer analysis while reducing the risk of excessive impact on computer operation. Furthermore, in addition to introducing such selection methods, displaying recommended icons reduces the amount of operation required to set favorable conditions.

[0038] The product feature acquisition unit 11 may acquire product feature information from the input content in the product feature input field a2. Alternatively, the product feature acquisition unit 11 may access the page at the URL entered in the URL field a3 and acquire product feature information from the content of that page. In this case, the product feature acquisition unit 11 performs natural language processing on the text data on the page and acquires product feature information by extracting words related to the product's features. For example, the product feature acquisition unit 11 analyzes headings, product descriptions, specification tables, or tag information on the page to identify features such as product name, brand, category, ingredients, or intended use.

[0039] Furthermore, the product feature acquisition unit 11 may acquire information on the target product and use natural language processing or statistical methods to calculate the importance and relevance of each feature as a product feature score. The product feature score is an indicator that shows the characteristics of the product.

[0040] For example, the product feature acquisition unit 11 performs natural language processing to extract features such as "skincare" and "moisturizing" from the product description and tag information. The product feature acquisition unit 11 scores the importance of these features based on past data. The past data may include historical data such as what kind of purchase trends products with similar features have shown and which features customers are more likely to respond to.

[0041] Furthermore, the product feature acquisition unit 11 may calculate the popularity of products having specific features in the market by using a statistical method, and assign an appropriate product feature score to each feature. The product feature score can be used as data for vectorizing product features. The product feature score is used for similarity calculation by being compared with the customer attribute score of the target customer in the customer extraction unit 13. Note that when vectorizing product features, words representing product features may be vectorized.

[0042] Furthermore, the product feature acquisition unit 11 may further acquire planned sales area information indicating a planned sales area for the target product. For example, the product feature acquisition unit 11 can acquire the planned sales area information from a product information database, official information of a manufacturer, data provided by a seller, or the like.

[0043] Returning to FIG. 4. The customer attribute acquisition unit 12 is an example of the customer attribute acquisition unit 102 described above. The customer attribute acquisition unit 12 acquires customer attribute information including customer preferences for each of a plurality of customers. The customer attribute information is information indicating customer characteristics and behavior patterns. The customer attribute information includes customer preferences and interests. For example, the customer attribute information may include the customer's age group, gender, or categories of products or services that the customer is interested in.

[0044] The plurality of customers are customers whose information is registered in the information providing system 1. The plurality of customers may be a plurality of persons belonging to a specific group. For example, the plurality of customers may be registered users of a specific EC site, or users having a usage history of a specific service. Without being limited thereto, the plurality of customers only need to be a population that is an acquisition target of customer attribute information.

[0045] The customer attribute acquisition unit 12 may acquire customer attribute information identified based on the purchase history information of each of multiple customers. The purchase history information shows a record of products that a customer has purchased in the past, and is information that associates the purchase date and time, the name of the purchased product, the purchase price, and the store where the purchase was made. The purchase history information may be stored in the customer information management device 30. The customer attribute acquisition unit 12 may acquire customer attribute information by acquiring the purchase history information from the customer information management device 30 and identifying customer attribute information for each customer based on the purchase history information.

[0046] Figure 6 shows an example of customer attribute information. As shown in Figure 6, customer attribute information includes, for example, customer ID, age group, amount spent, number of slips, or customer attribute tags. The customer ID is information used to identify each customer. The age group is information that indicates the age range of the customer. The age group is divided into a certain range (for example, 20s). Note that age may be included as customer attribute information instead of age group.

[0047] The spending amount is information that shows the total amount of goods a customer has purchased in the past. The number of slips is information that shows the number of purchase transactions a customer has made in the past. The number of slips is related to the frequency of purchases. The number of slips is typically an integer of 1 or more. For example, each time a customer makes a purchase, the number of slips is increased by 1. Therefore, customers who make many purchase transactions will have a higher number of slips. Customer attribute tags are information that shows the customer's preferences. Examples of customer attribute tags include "beauty," "health," "luxury-oriented," and "outdoor." The scope of products that make up the customer attribute information may include all past purchase history, or it may include purchase history within a specified period from the present.

[0048] The customer attribute acquisition unit 12 may acquire customer attribute information including a customer attribute score weighted for each feature of a purchased product included in purchase history information. The customer attribute score is an indicator indicating customer preference. The customer attribute score may be a numerical value calculated based on purchase history information. For example, in the example of FIG. 6, as the customer attribute information of the customer with customer ID "001", customer attribute tags such as "lotion_10, brand_35, ..." are included. The customer attribute score is the numerical part of the customer attribute tag. For example, in the example of this customer, the customer attribute score of "lotion" is 10, and the customer attribute score of "brand" is 35.

[0049] The customer attribute score may be set such that the higher the purchase frequency of the product, the larger the value. For example, the customer attribute score may be set such that when a customer purchases lotion once, the customer attribute score of "lotion" is "1", and when the customer purchases lotion additionally, the customer attribute score becomes "2".

[0050] The present invention is not limited to this, and the customer attribute score may be weighted for each feature. For example, when a customer purchases lotion once, the customer attribute score of "lotion" is "1", but when the customer purchases lotion additionally, the customer attribute score is "3", and when purchasing further, it may be set to "5", and so on. In this way, by changing the increment of the value according to the number of purchases, the strength of interest in a specific feature can be appropriately reflected. In addition, the strength of interest may be reflected according to the quantity of purchased products or the period until repurchase. More specifically, the larger the quantity of purchased products, the higher score may be set. Further, the shorter the period until purchasing the same or similar product as the previously purchased product, the higher score may be set.

[0051] Returning to Figure 4, let's continue the explanation of the customer attribute acquisition unit 12. The customer attribute acquisition unit 12 may further acquire extended customer attribute information, which is an extension of the existing customer attribute information, for each of the multiple customers. Extended customer attribute information represents information that is a higher-level concept of the existing customer attribute information. For example, as in the example above, if the customer attribute information based on purchase history is "lotion," then category information such as "cosmetics" or "skincare," which are higher-level concepts of lotion, can be included as extended customer attribute information.

[0052] The customer attribute acquisition unit 12 further acquires extended customer attribute information, which expands upon the existing customer attribute information. This allows the system to acquire information that reflects customer preferences more broadly, rather than just the information included in purchase history. As a result, the customer extraction unit 13 can accurately extract target customers based on their interests.

[0053] The customer attribute acquisition unit 12 may accept text input to add a new group different from the group classified by the size identification unit 14, and acquire customer attribute information corresponding to the new group by performing natural language processing on the input text. Details on adding a new group will be described later.

[0054] The customer extraction unit 13 is an example of the customer extraction unit 103 described above. Based on product feature information and customer attribute information, the customer extraction unit 13 extracts target customers from multiple customers who are expected to be interested in the target product. These target customers can become the target customers when advertising the product. For example, the customer extraction unit 13 extracts target customers from registered users of an e-commerce site.

[0055] Specifically, the customer extraction unit 13 extracts target customers based on the similarity between customer attribute information and product feature information. For example, the customer extraction unit 13 may calculate the similarity between a vectorized version of customer attribute information and a vectorized version of product feature information, and extract target customers based on that similarity.

[0056] As shown in the example in Figure 6, customer attribute information can be represented using numerical data, which are customer attribute scores, such as "Lotion_10, Brand_35, ...". The customer extraction unit 13 handles this customer attribute information as a multidimensional vector. Similarly, product feature information is quantified for each feature, such as "Lotion," "Skincare," and "Moisturizing." One example of a quantification method is vectorization. This allows each product feature to be represented as a vector. Hereafter, vectorized customer attribute information will be referred to as "customer attribute vectors," and vectorized product feature information will be referred to as "product feature vectors."

[0057] The customer extraction unit 13 calculates the similarity between the customer attribute vector and the product feature vector. The customer extraction unit 13 extracts customers with high similarity as target customers. The customer extraction unit 13 may also determine whether the similarity is high or not using a predetermined threshold.

[0058] For example, the customer extraction unit 13 can quantify the degree of interest using cosine similarity. This method indicates a high degree of similarity when the angle between two vectors is close to zero. The customer extraction unit 13 extracts customers whose similarity exceeds a threshold as target customers. This allows the customer extraction unit 13 to accurately extract customers who are expected to be interested in a particular product, based on purchase history information and customer attribute information.

[0059] The customer extraction unit 13 may extract target customers based on product characteristic information and extended customer attribute information. For example, if the customer attribute information is "lotion" as described above, the extended customer attribute information can include not only "lotion" but also category information such as "cosmetics" or "skincare," which are higher-level concepts of lotion. This allows the customer extraction unit 13 to extract target customers using the extended customer attribute information. In this case, the customer extraction unit 13 can extract not only customers whose customer attribute information includes "lotion," but also customers whose customer attribute information includes "cosmetics" or "skincare," etc., as target customers. This allows the customer extraction unit 13 to extract target customers with high accuracy based on customer interests.

[0060] The customer extraction unit 13 may extract multiple target customers who are expected to be interested in the target product in the planned sales area. For example, the customer extraction unit 13 may extract target customers from among customers residing in the planned sales area based on the similarity between the customer attribute vector and the product feature vector. In this way, it is possible to efficiently identify customers who are likely to be interested in the target product while taking into account the characteristics of the planned sales area.

[0061] The scale determination unit 14 is an example of the scale determination unit 104 described above. The scale determination unit 14 analyzes the scale of the target customers extracted by the customer extraction unit 13. Specifically, the scale determination unit 14 classifies the target customers into groups according to customer segment and determines the scale of each group. The groups are, for example, customer clusters generated by clustering processing. The scale determination unit 14 classifies multiple customers into multiple customer clusters by executing clustering processing and determines the scale of each of the multiple customer clusters.

[0062] The following are three possible methods for the scale identification unit 14 to classify target customers into customer clusters:

[0063] (1) Method for generating customer clusters based on product characteristic information The scale identification unit 14 generates multiple customer clusters based on product characteristic information and classifies target customers into these customer clusters. For example, the scale identification unit 14 generates clusters such as "people who like beauty," "people who are interested in health," "people who prefer luxury," and "people who like the outdoors," and classifies target customers into these customer clusters. As a result, the scale identification unit 14 can classify appropriate target customers into each customer cluster based on product characteristic information.

[0064] (2) Method for generating customer clusters based on customer attribute information common to a predetermined number of target customers The scale identification unit 14 may extract a predetermined number of target customers and generate customer clusters with similar purchase tendencies by analyzing the purchase history information and customer attribute information of the extracted target customers. For example, the scale identification unit 14 may extract customers who have purchased products of a specific category more than a predetermined number of times in the past year (for example, 10,000 people) and generate customer clusters based on that category (for example, 10 clusters). For example, the scale identification unit 14 may generate customer clusters such as "people who frequently purchase cosmetics" and "people who prefer a specific brand of home appliances" and classify the target customers into these customer clusters. In this way, the scale identification unit 14 can extract customer segments that are likely to be interested in specific products or services and classify the target customers based on customer attribute information common to those customer segments.

[0065] (3) Method using pre-registered customer clusters The scale identification unit 14 may extract only target customers that fit into pre-registered customer clusters and exclude other customers. As pre-registered customer clusters, customer clusters registered using default conditions may be used, or customer clusters generated by conditions uniquely set by the user, as described later, may be used. For example, the scale identification unit 14 may extract only target customers that fit into a customer cluster uniquely set by the user, such as "health-conscious people who like organic products," and exclude other customers. This allows the scale identification unit 14 to perform classification according to a pre-set marketing strategy, for example.

[0066] The above classification method is just one example; the scale identification unit 14 may classify target customers using methods other than those described above. The scale identification unit 14 identifies the size of each of the multiple classified customer clusters.

[0067] For example, the size determination unit 14 performs clustering based on the customer attribute vectors of the target customers and determines the size of each group. The size of each group may be represented by the number of customers belonging to the group, or by the sum of scores based on each customer's purchase history. The size determination unit 14 may start the analysis and determine the group classification and size when the analysis start button a4 on the input screen A shown in Figure 5 is pressed.

[0068] The scale determination unit 14 may also accept the user's specification of the number of customer clusters to be generated and the size of the customer clusters via the input screen A shown in Figure 5. The scale determination unit 14 may also accept the user's specification of age (age group) and gender via the input screen A.

[0069] Figure 7 shows an example of the analysis results screen B, which displays the analysis results in the scale determination unit 14. The analysis results screen B includes a product characteristics information field b1, an analysis results field b2, a cluster addition button b3, and a policy planning button b4.

[0070] The product features information section b1 displays information such as the name, features, sales performance, brand information, official URL, target age, and target gender of the product being analyzed. The analysis results section b2 displays a list of the analysis results from the scale identification section 14. The analysis results section b2 includes information on customer attributes, spending amount, number of uses, and number of people for each customer cluster. The cluster addition button b3 is a button for the user to define a new customer cluster and add it to the analysis target. The strategy planning button b4 is a button for performing operations to plan marketing measures and sales strategies for the displayed customer clusters.

[0071] The scale identification unit 14 can display the analysis results column b2 in any display manner. For example, the scale identification unit 14 may sort the customer clusters in descending order of usage amount or number of people and display the analysis results column b2.

[0072] Furthermore, the size determination unit 14 may estimate the customer's occupation and lifestyle based on product purchase history information, member registration information for e-commerce sites, survey data, web browsing history, or location information, and classify the customer clusters taking the estimation results into consideration.

[0073] For example, the scale identification unit 14 analyzes purchase trends for desks and PC peripherals for working from home based on purchase history information and identifies remote workers. The scale identification unit 14 may also identify target customers who spend a long time at home based on location information. The scale identification unit 14 integrates this information and performs clustering using machine learning to classify customers into customer clusters such as "remote workers who work in a home office."

[0074] The scale identification unit 14 may highlight customer clusters that are particularly noteworthy among the classified customer clusters. In the example in Figure 7, the scale identification unit 14 indicates that the customer cluster of "foodies who frequently go out to eat and drink" is an important customer cluster based on the analysis results by shading it. The scale identification unit 14 may also determine whether a customer cluster is important or not based on, for example, market size or purchasing trends.

[0075] The scale determination unit 14 may delete customer clusters that the user deems unnecessary, upon receiving user input. Furthermore, as described later, the scale determination unit 14 may add customer clusters that the user wishes to add.

[0076] Figure 8 shows an example of the detailed analysis results screen C, which displays detailed analysis results. The detailed analysis results screen C may also be accessed by selecting a customer cluster in the analysis results field b2 of the analysis results screen B shown in Figure 7. Here, an example is shown that displays the details of the customer cluster that was shaded in Figure 7.

[0077] The detailed analysis results screen C includes a customer cluster characteristics field c1, a customer cluster potential needs field c2, a customer cluster attribute list field c3, and a close button c4. The customer cluster characteristics field c1 displays the characteristics of the customer cluster. In this example, detailed analysis results for the customer cluster "foodies who frequently go out to eat and drink" are displayed.

[0078] The customer cluster potential needs column c2 displays the potential needs of the customer cluster. For example, the scale identification unit 14 may analyze customer purchase history information and estimate the customer's potential needs. For example, if customers who frequently eat out purchase bad breath care products, the scale identification unit 14 may estimate that post-meal bad breath countermeasures are a potential need. The scale identification unit 14 may perform this estimation not only based on purchase history information, but also based on, for example, the contents of a questionnaire.

[0079] The customer cluster attribute list field c3 displays a list of customer attribute information included in the customer cluster. In this example, the customer attribute tags included in the customer cluster "Foodies who frequently eat out or go to drinking parties" are displayed in descending order of frequency.

[0080] The close button c4 is used to close the detailed analysis results screen C. The user presses the close button c4 to return to the original screen. As a result, the scale identification unit 14 closes the detailed analysis results screen C and displays the analysis results screen B again.

[0081] The scale determination unit 14 may further determine the scale of the new group based on the customer attribute information corresponding to the new group obtained by the customer attribute acquisition unit 12. Here, the scale determination unit 14 further determines the scale of the new customer cluster based on the customer attribute information corresponding to the new customer cluster. The new customer cluster is a customer cluster generated by conditions uniquely set by the user.

[0082] Figure 9 shows an example of the cluster addition screen D for adding a custom customer cluster. The scale determination unit 14 may display the cluster addition screen D in response to the cluster addition button b3 shown in Figure 7 being pressed. The cluster addition screen D includes a customer cluster field d1, a customer profile characteristics field d2, a customer cluster addition button d3, a custom customer cluster list field d4, and an analysis start button d5.

[0083] The Customer Cluster field d1 is where the user enters the name of the new customer cluster to be added. The Customer Profile Characteristics field d2 is where the user enters the preferences and behavioral characteristics of the customer cluster to be added, and information about purchasing trends and lifestyle can be entered. The Add Customer Cluster button d3 is a button to register the entered customer cluster and add it to the cluster list. The Custom Customer Cluster List field d4 is a field that displays customer clusters that the user has registered in the past. The Start Analysis button d5 is a button that starts the analysis of customer clusters, similar to the Start Analysis button a4 mentioned above.

[0084] The customer attribute acquisition unit 12 accepts text input into the customer cluster field d1 and the customer profile characteristics field d2, and performs natural language processing on the input text to acquire customer attribute information corresponding to the new group. The scale determination unit 14 automatically extracts relevant data based on the name and characteristics of the input customer cluster, and constructs a new customer cluster based on the extracted data. The scale determination unit 14 adds the constructed customer cluster to the existing customer cluster. The scale determination unit 14 further determines the scale of the new customer cluster based on the customer attribute information corresponding to the new customer cluster.

[0085] For example, when adding a customer cluster such as "health-conscious people who like organic products," the customer attribute acquisition unit 12 extracts related words such as "organic," "health-conscious," "additive-free," and "natural food." The customer extraction unit 13 automatically extracts customers who are assumed to be interested in these related words. In this process, the customer extraction unit 13 automatically recognizes synonyms and related words by performing natural language processing.

[0086] Therefore, even if a user tries to add a customer cluster using different expressions, such as "natural-oriented" or "additive-free foods," the customer extraction unit 13 appropriately classifies customers who are expected to be interested in these related words based on customer attribute tags. As a result, the customer extraction device 10 can create customer clusters flexibly and efficiently compared to methods that require the collection of large amounts of data and the setting of flags to create new clusters.

[0087] The scale determination unit 14, having determined the scale of the customer cluster, outputs display screens as shown in Figures 7 to 9 in any manner. For example, the scale determination unit 14 transmits data for displaying these display screens to the communication terminal 20 via the communication unit 15. As a result, the scale determination unit 14 displays these display screens on the display unit 22 of the communication terminal 20.

[0088] Returning to Figure 4, the communication unit 15 is a communication interface for wired or wireless communication. The communication unit 15 communicates with the communication terminal 20 and the customer information management device 30.

[0089] (Communication Terminal 20) Figure 10 is a block diagram showing the configuration of the communication terminal 20. The communication terminal 20 is a terminal device used by the user. The communication terminal 20 comprises an input unit 21, a display unit 22, and a communication unit 23. The communication terminal 20 may be, for example, a PC (Personal Computer), a smartphone, a tablet terminal, or a mobile phone.

[0090] The input unit 21 is an input device that receives user input. The input unit 21 may be, for example, a mouse, keyboard, touch panel, smartphone, tablet, or voice input device.

[0091] The display unit 22 is a display device that displays various types of information. The display unit 22 is, for example, a display. The display unit 22 may also be a touch panel that has the functions of the input unit 21. Although not shown in the figures, the communication terminal 20 may also be equipped with an audio output unit such as a speaker and output information by voice.

[0092] The communication unit 23 is a communication interface for communication via wired or wireless means. The communication unit 23 communicates with the customer extraction device 10 and the customer information management device 30.

[0093] (Customer Information Management Device 30) Figure 11 is a block diagram showing the configuration of the customer information management device 30. The customer information management device 30 is a device for managing customer information. The customer information management device 30 may be managed by, for example, a business that sells goods or services or a business that provides online payment services. The customer information management device 30 may manage customer information provided by multiple businesses. The customer information management device 30 includes a communication unit 31 and a storage unit 32.

[0094] The communication unit 31 is a communication interface for communication via wired or wireless means. The communication unit 31 communicates with the customer extraction device 10 and the communication terminal 20.

[0095] The storage unit 32 stores various types of data and programs. At least a portion of the storage unit 32 is composed of non-volatile memory so that data is retained even when the power to the customer information management device 30 is turned off. The storage unit 32 may store various types of data, including the analysis results described above. For example, the storage unit 32 may store the purchase history information described above.

[0096] The configuration of the information provision system 1 has been described above. The configuration of the information provision system 1 described above is merely an example and can be modified as appropriate. For example, if some or all of the components of the information provision system 1 are implemented by multiple information processing devices or circuits, these multiple information processing devices or circuits may be centrally located or distributed. For example, the information processing devices or circuits may be implemented in a form in which each is connected via a communication network, such as a client-server system or a cloud computing system. Also, for example, the functions of the customer extraction device 10 may be provided in SaaS (Software as a Service) format.

[0097] (Processing of the customer extraction device 10) Next, the processing performed by the customer extraction device 10 will be explained with reference to Figure 12. Figure 12 is a flowchart showing the flow of processing performed by the customer extraction device 10.

[0098] First, the product feature acquisition unit 11 acquires product feature information (S11). For example, the product feature acquisition unit 11 receives input from a user operating the communication terminal 20 via an input screen A as shown in Figure 5.

[0099] Next, the customer attribute acquisition unit 12 acquires customer attribute information for each of the multiple customers (S12). For example, the customer attribute acquisition unit 12 acquires purchase history information of a specific credit card member from the customer information management device 30 and acquires customer attribute information.

[0100] Next, the customer extraction unit 13 extracts target customers from multiple customers (S13). Subsequently, the scale determination unit 14 classifies the target customers into customer clusters according to customer segment (S14). The scale determination unit 14 also determines the size of each customer cluster (S15). The scale determination unit 14 transmits the analysis results to the communication terminal 20 via the communication unit 15 (S16). The scale determination unit 14 displays the analysis results using a screen such as the analysis results screen B shown in Figure 7 or the detailed analysis results screen C shown in Figure 8.

[0101] The scale identification unit 14 may also select a destination to display the analysis results according to the type of customer segment identified. For example, if a customer cluster of "remote workers working from home offices" is extracted, the scale identification unit 14 will select a department that handles products for remote workers as the destination.

[0102] Alternatively, the scale identification unit 14 may acquire sales data from stores and select stores with high sales performance targeting remote workers as destinations. The scale identification unit 14 may also select destinations based on the size of the amount spent and the number of people in the customer cluster. Destinations may include, for example, the telephone number or email address of a company, store, or individual. Destination information may be pre-registered as a destination database in, for example, the storage unit (not shown) of the customer extraction device 10. By selecting destinations in this way, the scale identification unit 14 can select an appropriate department to provide the analysis results to, even if the departments responsible for each sales strategy are separated.

[0103] Furthermore, when providing analysis results via email, the scale identification unit 14 may include the analysis results and information about the entered products in the email body. For example, the scale identification unit 14 may include information such as customer attributes, spending amount, number of uses, and number of people for each customer cluster in the email body as part of the analysis results. The scale identification unit 14 may also include information about the entered products in the email body, such as the name of the target product, its characteristics, sales performance, brand information, official URL, target age, and target gender. In this way, users can easily grasp the analysis results and product information.

[0104] Next, the scale determination unit 14 determines whether or not to add a new customer cluster (S17). For example, if the cluster addition button b3 is pressed on the analysis results screen B shown in Figure 7, the scale determination unit 14 determines to add a new customer cluster.

[0105] If it is determined that a new customer cluster should be added (YES in S17), the customer attribute acquisition unit 12 acquires customer attribute information corresponding to the customer cluster to be added (S18). For example, the customer attribute acquisition unit 12 may acquire customer attribute information using a screen such as the cluster addition screen D shown in Figure 9. After that, the process returns to step S13 and repeats the subsequent processing. If it is determined that a new customer cluster should not be added (NO in S17), the process ends.

[0106] As explained above, according to the information provision system 1 related to this disclosure, the customer extraction device 10 acquires product characteristic information and customer attribute information, and based on this information, extracts target customers who are expected to be interested in the product from among multiple customers. The customer extraction device 10 also classifies the target customers into customer clusters according to customer segment and identifies the size of each customer cluster. With this configuration, the customer extraction device 10 can appropriately visualize the potential needs of customers.

[0107] <Example of Hardware Configuration> Each functional component of the customer extraction device 100, customer extraction device 10, communication terminal 20, and customer information management device 30 (hereinafter referred to as "customer extraction device 100, etc.") may be implemented by hardware that realizes each functional component (e.g., hardwired electronic circuits, etc.), or by a combination of hardware and software (e.g., a combination of electronic circuits and programs that control them, etc.). The case in which each functional component of the customer extraction device 100, etc. is implemented by a combination of hardware and software will be further explained below.

[0108] Figure 13 is a block diagram illustrating the hardware configuration of a computer 900 that implements the customer extraction device 100, etc. The computer 900 may be a dedicated computer designed to implement the customer extraction device 100, etc., or it may be a general-purpose computer. The computer 900 may also be a portable computer such as a smartphone or tablet terminal.

[0109] For example, by installing a predetermined application on the computer 900, the various functions of the customer extraction device 100 and other devices are realized on the computer 900. The above application consists of a program for realizing the functional components of the customer extraction device 100 and other devices.

[0110] The computer 900 includes a bus 902, a processor 904, a memory 906, a storage device 908, an input / output interface 910, and a network interface 912. The bus 902 is a data transmission path for the processor 904, memory 906, storage device 908, input / output interface 910, and network interface 912 to send and receive data to and from each other. However, the method of connecting the processor 904 and the other components to each other is not limited to bus connection.

[0111] The processor 904 is a variety of processors such as a CPU, GPU, FPGA (Field-Programmable Gate Array), or quantum processor (quantum computer control chip). The memory 906 is a main memory device implemented using RAM (Random Access Memory), etc. The storage device 908 is an auxiliary storage device implemented using a hard disk, SSD (Solid State Drive), memory card, or ROM (Read Only Memory), etc.

[0112] The input / output interface 910 is an interface for connecting the computer 900 with input / output devices. For example, input devices such as keyboards and output devices such as display devices are connected to the input / output interface 910.

[0113] The network interface 912 is an interface for connecting the computer 900 to a network. This network may be a LAN (Local Area Network) or a WAN (Wide Area Network).

[0114] The storage device 908 stores programs that implement each functional component of the customer extraction device 100 (programs that implement the aforementioned applications). The processor 904 reads these programs into the memory 906 and executes them to implement each functional component of the customer extraction device 100.

[0115] Each processor executes one or more programs containing a set of instructions for causing a computer to perform the algorithms described with reference to the drawings. These programs, when loaded into a computer, contain a set of instructions (or software code) for causing the computer to perform one or more functions described in the embodiments. The programs may be stored in various types of non-transitory computer-readable medium or tangible storage medium. Examples, but not limited to, include non-transitory computer-readable medium or tangible storage medium, such as RAM, ROM, flash memory, SSD or other memory technologies, CD-ROM, DVD (Digital Versatile Disc), Blu-ray® disc or other optical disc storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices. The programs may also be transmitted over various types of transient computer-readable medium or communication medium. Examples, but not limited to, include transient computer-readable medium or communication medium, such as electrical, optical, acoustic or other forms of propagating signals.

[0116] Although the present disclosure has been described above with reference to embodiments, the present disclosure is not limited to the embodiments described above. Various modifications to the structure and details of the present disclosure are possible, as can be understood by those skilled in the art within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0117] Each drawing is merely illustrative to illustrate one or more embodiments. Each drawing may be associated with one or more other embodiments, rather than being associated with only one specific embodiment. As those skilled in the art will understand, various features or steps described with reference to any one drawing can be combined with features or steps shown in one or more other drawings to create embodiments that are not explicitly illustrated or described. Not all features or steps shown in any one drawing to illustrate an exemplary embodiment are necessarily required, and some features or steps may be omitted. The order of steps shown in any of the drawings may be changed as appropriate.

[0118] Some or all of the above embodiments may also be described as follows, but are not limited to the following: (Note 1) A program that causes a computer to perform the following steps: a product feature acquisition step of acquiring product feature information that indicates the characteristics of a target product; a customer attribute acquisition step of acquiring customer attribute information, including customer preferences, for each of a plurality of customers; a customer extraction step of extracting target customers who are expected to be interested in the target product from the plurality of customers based on the product feature information and the customer attribute information; and a scale identification step of classifying the target customers into groups according to customer segment and identifying the size of each group. (Note 2) The program according to Note 1, wherein the product feature acquisition step accepts the specification of an information source containing information on the target product, and acquires the product feature information by extracting the characteristics of the target product included in the specified information source. (Note 3) The program according to Note 1 or 2, wherein the product feature acquisition step accepts text input from a predetermined input field, and acquires the product feature information by performing natural language processing on the input text. (Note 4) The program according to any one of Notes 1 to 3, wherein in the customer attribute acquisition step, the program acquires the customer attribute information identified based on the purchase history information of each of the multiple customers. (Note 5) The program according to Note 4, wherein in the customer attribute acquisition step, the program acquires the customer attribute information including customer attribute scores weighted according to the characteristics of the purchased items included in the purchase history information, and in the customer extraction step, the program extracts the target customers based on the similarity between the customer attribute information and the product characteristic information. (Note 6) The program according to any one of Notes 1 to 5, wherein in the customer attribute acquisition step, the program acquires extended customer attribute information which is an extension of the customer attribute information, for each of the multiple customers, and in the customer extraction step, the program extracts the target customers based on the product characteristic information and the extended customer attribute information.(Note 7) The program according to any one of Notes 1 to 6, wherein in the customer attribute acquisition step, the program accepts text input for adding a new group different from the group, and performs natural language processing on the input text to acquire customer attribute information corresponding to the new group, and in the size determination step, it further determines the size of the new group based on the acquired customer attribute information corresponding to the new group. (Note 8) The program according to any one of Notes 1 to 7, wherein in the product feature acquisition step, it further acquires sales area information indicating the planned sales area of ​​the target product, and in the customer extraction step, it extracts a plurality of target customers who are expected to be interested in the target product in the planned sales area. (Note 9) The program according to any one of Notes 1 to 8, wherein in the size determination step, it classifies the plurality of customers into a plurality of customer clusters by performing clustering processing, and determines the size of each of the plurality of customer clusters. (Note 10) A customer extraction device comprising: a product feature acquisition unit that acquires product feature information indicating the characteristics of a target product; a customer attribute acquisition unit that acquires customer attribute information including customer preferences for each of a plurality of customers; a customer extraction unit that extracts target customers from the plurality of customers who are expected to be interested in the target product based on the product feature information and the customer attribute information; and a scale identification unit that classifies the target customers into groups according to customer segment and identifies the size of each group. (Note 11) A customer extraction method comprising: a product feature acquisition step that acquires product feature information indicating the characteristics of a target product; a customer attribute acquisition step that acquires customer attribute information including customer preferences for each of a plurality of customers; a customer extraction step that extracts target customers from the plurality of customers who are expected to be interested in the target product based on the product feature information and the customer attribute information; and a scale identification step that classifies the target customers into groups according to customer segment and identifies the size of each group.

[0119] Some or all of the elements (e.g., configuration and function) described in Appendices 2 to 9 that are dependent on Appendice 1 may also be dependent on Appendices 10 and 11 in the same way as those described in Appendices 2 to 9. Some or all of the elements described in any appendice may be applied to various hardware, software, recording means, systems, and methods for recording software.

[0120] This application claims priority based on Japanese Patent Application No. 2025-049791, filed on 25 March 2025, and incorporates all of its disclosures herein.

[0121] 1 Information Provision System 10 Customer Extraction Device 11 Product Feature Acquisition Unit 12 Customer Attribute Acquisition Unit 13 Customer Extraction Unit 14 Scale Identification Unit 15 Communication Unit 20 Communication Terminal 21 Input Unit 22 Display Unit 23 Communication Unit 30 Customer Information Management Device 31 Communication Unit 32 Storage Unit 100 Customer Extraction Device 101 Product Feature Acquisition Unit 102 Customer Attribute Acquisition Unit 103 Customer Extraction Unit 104 Scale Identification Unit 900 Computer 902 Bus 904 Processor 906 Memory 908 Storage Device 910 Input / Output Interface 912 Network Interface A Input Screen a1 Name Field a2 Product Feature Input Field a3 URL Field a4 Start Analysis Button B Analysis Results Screen b1 Product Feature Information Field b2 Analysis Results Field b3 Add Cluster Button b4 Policy Planning Button C Analysis Results Details Screen c1 Customer Cluster Characteristics Section c2 Customer Cluster Potential Needs Section c3 Customer Cluster Attribute List Section c4 Close Button D Cluster Addition Screen d1 Customer Cluster Section d2 Customer Profile Characteristics Section d3 Add Customer Cluster Button d4 Unique Customer Cluster List Section d5 Start Analysis Button N Network

Claims

1. A program that causes a computer to perform the following steps:

1. A process to acquire product feature information that indicates the characteristics of the target product; 2. A process to acquire customer attribute information, including customer preferences, for each of multiple customers; 3. A process to extract target customers who are expected to be interested in the target product from the multiple customers based on the product feature information and the customer attribute information; and 4. A process to classify the target customers into groups according to customer segment and to determine the size of each group.

2. The program according to claim 1, wherein the process for obtaining the product characteristic information accepts the designation of an information source containing information of the target product, and obtains the product characteristic information by extracting the characteristics of the target product contained in the designated information source.

3. The program according to claim 1 or 2, which, in the process of acquiring the product characteristic information, accepts text input from a predetermined input field and acquires the product characteristic information by performing natural language processing on the input text.

4. The program according to claim 1 or 2, wherein the process for acquiring customer attribute information acquires customer attribute information identified based on the purchase history information of each of the plurality of customers.

5. The program according to claim 4, wherein the process for acquiring customer attribute information acquires customer attribute information including customer attribute scores weighted according to the characteristics of the purchased items included in the purchase history information, and the process for extracting target customers extracts target customers based on the similarity between the customer attribute information and the product characteristic information.

6. The program according to claim 1 or 2, wherein in the process of acquiring the customer attribute information, extended customer attribute information obtained by extending the customer attribute information is acquired for each of the plurality of customers, and in the process of extracting the target customers, the target customers are extracted based on the product characteristic information and the extended customer attribute information.

7. The program according to claim 1 or 2, wherein the process for acquiring customer attribute information accepts text input for adding a new group different from the group, and acquires customer attribute information corresponding to the new group by performing natural language processing on the input text, and the process for determining the size of each group further determines the size of the new group based on the acquired customer attribute information corresponding to the new group.

8. The program according to claim 1 or 2, wherein the process for obtaining the product characteristic information further obtains sales area information indicating the planned sales area for the target product, and the process for extracting target customers extracts a plurality of target customers who are expected to be interested in the target product in the planned sales area.

9. The program according to claim 1 or 2, wherein the process for determining the size of each group involves classifying the multiple customers into multiple customer clusters by performing a clustering process, and determining the size of each of the multiple customer clusters.

10. A customer extraction device comprising: a product feature acquisition means for acquiring product feature information indicating the characteristics of a target product; a customer attribute acquisition means for acquiring customer attribute information, including customer preferences, for each of a plurality of customers; a customer extraction means for extracting target customers who are expected to be interested in the target product from the plurality of customers based on the product feature information and the customer attribute information; and a scale identification means for classifying the target customers into groups according to customer segment and identifying the size of each group.

11. A customer extraction method that obtains product feature information indicating the characteristics of a target product, obtains customer attribute information including customer preferences for each of several customers, extracts target customers who are expected to be interested in the target product from the several customers based on the product feature information and the customer attribute information, classifies the target customers into groups according to customer segment, and identifies the size of each group.

Images