Customer analysis system and method

Abstract

To generate a model resulting from integrating a model relating to people's values and life styles in overall life and a model relating to customer's needs for products or services.SOLUTION: A customer analysis system 10 generates a third model by integrating a first model and a second model by their parts indicating the same or similar values of products or services, wherein the first model includes general characteristics such as people's values and life styles in overall life which do not depend on demographic characteristics such as age and gender, and defines correlations between the general characteristics and the values of products or services, and the second model represents customer's needs for products or services and defines correlations between customer's values of products or services and contents about products or services to be selected by the customer. This third model allows a bond with the customer to be strengthened while providing a product or a service satisfying the customer.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to the technical field of analyzing customers using models. [Background technology]

[0002] Conventionally, companies that provide products and services have modeled customer needs based on survey data collected from multiple customers, and have used this model to provide products and services that are suited to each customer.

[0003] For example, Patent Document 1 discloses a technology for quantitatively evaluating a customer's purchasing preference type and designing a purchasing preference type that has a high degree of agreement with the actual product purchase history. Patent Document 2 discloses a technology for statistically processing characteristic data of multiple customers who have similar characteristics using a clustering technique. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent No. 6411396 [Patent Document 2] Patent No. 5649756 Summary of the Invention [Problem to be solved by the invention]

[0005] In the future, businesses will be expected to not only provide products and services suited to customers, but also to deepen ties with customers (in other words, to form long-term relationships with customers) by conducting continuous business at various touch points, such as providing additional services after the provision of the products or services. In this case, in order to deepen ties with customers, it is considered necessary not only to use a simple model of customer needs as described above, but also to communicate based on essential (general) values ​​and lifestyles in people's lives in general. This is because, for example, it will be possible to estimate the essential characteristics of a customer and provide customer service that resonates with the customer.

[0006] However, conventional models of customer needs cannot estimate values ​​and lifestyles in a person's life in general. In order to build a model for estimating such values ​​and lifestyles, it is necessary to obtain big data, which does not directly benefit the business of the company. Therefore, it can be said that building such a model is economically unreasonable and unrealistic for a company. Conventionally, a technology has been proposed that finds commonalities between multiple models related to multiple products and services and builds a model that integrates multiple models based on the commonalities (see Patent Document 2, for example). However, this technology simply integrates multiple models, and does not include values ​​and lifestyles in a person's life in general. In addition, a technology that uses common questions including demographic information and values ​​to link different industries and perform modeling is also conceivable, but in this case, data from values ​​and lifestyles to customer needs are obtained from the same sample within the same industry, and there is a problem with economic rationality when building a model. At the same time, modeling using common questions is insufficient for communication based purely on values, such as connecting models using demographic information that does not depend on values, such as age and gender.

[0007] The present invention has been made to solve the above-mentioned problems, and aims to provide a customer analysis system and method that can create a model that accurately integrates a model related to values ​​and lifestyles in people's lives in general with a model related to customer needs for products or services. [Means for solving the problem]

[0008] In order to achieve the above-mentioned object, the present invention provides a customer analysis system for analyzing customers using a model, comprising: a first model acquisition means for acquiring a first model indicating a person's general characteristics, including at least values ​​and lifestyle in general life; a second model acquisition means for acquiring a second model indicating the customer's needs for a product or service; and a third model creation means for creating a third model by integrating the first and second models, wherein the general characteristics defining the first model are values ​​and lifestyle expressed independently of demographic characteristics indicating at least age and gender, the first model includes a correlation between the general characteristics and the values ​​regarding the product or service, the second model includes a correlation between the customer's values ​​regarding the product or service and content related to the product or service selected by the customer, and the third model creation means creates the third model by integrating the first and second models using parts of the first and second models that indicate the same or similar values ​​regarding the product or service.

[0009] According to the present invention configured in this manner, the customer analysis system includes general characteristics such as values ​​and lifestyle in a person's life in general, and among these general characteristics, a first model (higher model) which defines the correlation between pure values ​​and lifestyle that are not dependent on demographic information and values ​​regarding products or services, and a second model (lower model) which indicates the customer's needs for products or services and defines the correlation between the customer's values ​​regarding products or services and the details of the products or services selected by the customer, is integrated with parts which show the same or similar values ​​regarding products or services to create a third model. This makes it possible to create a third model that precisely integrates the first and second models, and by using this third model to provide products or services, it is possible to deepen ties with customers while providing products and services that are suitable for them. In other words, by clearly understanding the true nature of customers based on their genuine values ​​and lifestyles, and by communicating appropriately based on values ​​that are not dependent on age or gender, it is possible to gain sympathy from customers and effectively deepen ties with them, in other words, to form relationships with customers that will continue for a long time.

[0010] In the present invention, preferably, the first model includes a correlation between general characteristics and values ​​toward products or services for each of a plurality of product or service classifications, the second model acquisition means acquires a plurality of second models indicating customer needs for each of the plurality of product or service classifications, and the third model creation means creates a third model that integrates the first model and the plurality of second models in accordance with the plurality of product or service classifications. The third model created in this way makes it possible to understand the relationship between products or services in different classifications. Therefore, by using this third model, it becomes possible to develop effective collaborative products or tie-up products through collaboration between different industries, for example. It also makes it possible to develop products or services that utilize information from other industries and acquire new customers.

[0011] In the present invention, preferably, the first model is formed in a hierarchical structure so as to represent a direct or indirect causal relationship between a plurality of items indicating the content of general characteristics and a plurality of items indicating the content of values ​​regarding a product or service, the second model is formed in a hierarchical structure so as to represent a direct or indirect causal relationship between a plurality of items indicating the content of values ​​regarding a product or service and a plurality of items indicating specific content regarding the product or service, and the third model creation means creates the third model by combining the bottom part of the hierarchical structure in the first model with the top part of the hierarchical structure in the second model. According to the present invention configured in this manner, a more accurate third model can be created by combining the lowest-level part of a first model formed in a hierarchical structure with the highest-level part of a second model formed in a hierarchical structure.

[0012] In the present invention, preferably, the system further comprises a first model creation means for creating a first model based on any one or more of the questionnaire results, the psychological test results, the brain response test results, and the physiological response test results, and the first model acquisition means acquires the first model created by the first model creation means. According to the present invention thus configured, it is possible to create a highly accurate first model, particularly by using test results of brain responses and physiological responses in addition. Effect of the Invention

[0014] According to the customer analysis system and method of the present invention, it is possible to create a model that accurately integrates a model relating to values ​​and lifestyles in a person's life in general with a model relating to the customer's needs for products or services. By using this model, it is possible to deepen bonds with customers while providing them with products and services that are suitable for them. [Brief description of the drawings]

[0015] [Figure 1] 1 is a block diagram showing a schematic configuration of a customer analysis system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is an explanatory diagram of the basic concepts of first to third models in an embodiment of the present invention. [Diagram 3] FIG. 11 is an explanatory diagram of an example of a method of using a third model in the embodiment of the present invention. [Figure 4] FIG. 11 is an explanatory diagram of another example of a method of using the third model in the embodiment of the present invention. [Diagram 5] 3 illustrates an example of a first model according to an embodiment of the present invention. [Figure 6] 13 illustrates an example of a second model according to an embodiment of the present invention. [Figure 7] 4 is a flowchart showing a process of creating a first model according to an embodiment of the present invention. [Figure 8] 13 is a flowchart showing an example of a process for creating second and third models according to an embodiment of the present invention. [Figure 9] 13 is a flowchart showing another example of the process of creating the second and third models according to the embodiment of the present invention. [Figure 10] 13 is a flowchart showing yet another example of the process of creating the second and third models according to the embodiment of the present invention. [Figure 11] 13 shows an example of a third model created according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a customer analysis system and method according to an embodiment of the present invention will be described with reference to the accompanying drawings.

[0017] [System Configuration] Fig. 1 is a block diagram showing a schematic configuration of a customer analysis system according to the present embodiment. As shown in Fig. 1, a customer analysis system 10 mainly includes a first information processing device 1 and a plurality of second information processing devices 5 connected to the first information processing device 1 via a communication line (Internet line) 3 as a network.

[0018] The first and second information processing devices 1, 5 each mainly include a processor 1a, 5a such as a CPU that executes various processes, memory (ROM, RAM, hard disk, etc.) 1b, 5b that stores programs to be executed by the processor 1a, 5a and various data necessary for executing the programs, communication devices 1c, 5c for communicating via a communication line 3, input devices 1d, 5d such as a mouse, keyboard, or touch panel for inputting various information to the first and second information processing devices 1, 5, and output devices 1e, 5e such as a display or speaker for outputting various information from the first and second information processing devices 1, 5.

[0019] The second information processing device 5 is a computer device used in various companies. Basically, the multiple second information processing devices 5 are used by different types of companies. Specifically, each of the multiple second information processing devices 5 creates a second model (details will be described later) that indicates customer needs for products or services provided by each company based on the results of, for example, questionnaires or Internet surveys conducted by each company.

[0020] On the other hand, the first information processing device 1 is a computer device equivalent to a server device (center device) that collects various information transmitted from a plurality of second information processing devices 5 and the like. Specifically, the first information processing device 1 performs a process of creating a first model (details of which will be described later) that indicates a person's general characteristics, including at least values ​​and lifestyles in general life, based on the results of questionnaires, psychological tests, brain reactions, physiological reactions, and the like collected from a plurality of people. In addition, the first information processing device 1 acquires information indicating the values ​​of customers regarding products or services provided by each company from the plurality of second information processing devices 5, and creates the first model using this information as well. In this case, for example, a company that uses the first information processing device 1 (i.e., a company that creates the first model) will purchase information from each company that uses the second information processing device 5.

[0021] Then, the second information processing device 5 acquires the first model created by the first information processing device 1 as described above, and performs a process of creating a third model (described in detail later) by integrating the first model with the second model. In this case, for example, each company using the second information processing device 5 will purchase the first model from the company using the first information processing device 1 (i.e., the company creating the first model).

[0022] The first information processing device 1 (particularly the processor 1a) corresponds to an example of the "first model creation means" and the "first model acquisition means" in the present invention, and the second information processing device 5 (particularly the processor 5a) corresponds to an example of the "second model acquisition means" and the "third model creation means" in the present invention.

[0023] [Basic concept] Next, the basic concept of this embodiment will be described with reference to Figures 2 to 4. First, Figure 2 is an explanatory diagram of the basic concepts of first to third models in this embodiment. Figure 2 shows a simplified conceptual diagram of each model.

[0024] As shown in Fig. 2, the first model is formed in a hierarchical structure including an upper layer and a lower layer (note that each of these upper and lower layers is not limited to being composed of only one layer). Specifically, in the first model, the upper layer includes a plurality of nodes A to C indicating general characteristics such as pure values ​​and lifestyles in a person's life in general, independent of demographic characteristics such as age and sex, and the lower layer includes a plurality of nodes a to f indicating values ​​for a product or service. That is, the first model is a Bayesian network that indicates the relative relationship between the plurality of nodes A to C indicating general characteristics of a person and the plurality of nodes a to f indicating values ​​for a product or service (particularly, the causal relationship between the events of nodes A to C and the events of nodes a to f).

[0025] According to such a first model, for example, a person having a general characteristic A is estimated to have values ​​b and d for products or services. In the example of the first model shown in Fig. 2, the nodes a to c indicated by reference character A01 and the nodes d to f indicated by reference character A02 belong to different product or service categories. For example, the nodes a to c indicate events related to the field of cars, specifically, the contents of a person's values ​​related to cars, and the nodes d to f indicate events related to the field of food, specifically, the contents of a person's values ​​related to food.

[0026] Next, the second model is formed in a hierarchical structure including an upper layer showing the customer's values ​​regarding products or services, and a lower layer showing the specific details regarding the products or services selected by the customer (note that each of these upper and lower layers is not limited to being composed of only one layer). In Fig. 2, two models, second models A and B, are shown, and the second models A and B correspond to different classifications of products or services.

[0027] The second model A is a Bayesian network that represents the relative relationship between a plurality of nodes a-c (see symbol A03) that indicate the values ​​of a customer regarding a product or service, and a plurality of nodes I-III that indicate specific details regarding the product or service (particularly the causal relationship between the events of nodes I-III and the events of nodes a-c). According to this second model A, for example, a customer who has a value of a regarding a product or service is estimated to select a detail III regarding the product or service. For example, nodes a-c indicate the contents of the customer's values ​​regarding cars, and nodes I-III indicate the types of cars selected by the customer.

[0028] On the other hand, the second model B is a Bayesian network that represents the relative relationship between a plurality of nodes d-f (see symbol A04) that indicate the customer's values ​​regarding a product or service, and a plurality of nodes IV-VI that indicate specific details regarding the product or service (particularly the causal relationship between the events of nodes IV-VI and the events of nodes d-f). According to this second model B, for example, a customer who has a value d regarding a product or service is estimated to select a content V regarding the product or service. For example, nodes d-f indicate the contents of the customer's values ​​regarding food, and nodes IV-VI indicate the types of food and drink selected by the customer.

[0029] 2, nodes a-c (see reference symbol A01) indicating values ​​regarding a product or service in the first model are common to nodes a-c (see reference symbol A03) indicating values ​​regarding a product or service in the second model A. Also, nodes d-f (see reference symbol A02) indicating values ​​regarding a product or service in the first model are common to nodes d-f (see reference symbol A04) indicating values ​​regarding a product or service in the second model B.

[0030] Therefore, in this embodiment, the first model and the second model A are combined via the lower layer portion indicated by reference character A01 in the first model and the upper layer portion indicated by reference character A03 in the second model A, and the first model and the second model B are combined via the lower layer portion indicated by reference character A02 in the first model and the upper layer portion indicated by reference character A04 in the second model B, thereby creating a third model that integrates the first model and the second models A and B. Note that the nodes a to f included in the portions indicated by reference characters A01 and A02 in the first model and the nodes a to f included in the portions indicated by reference characters A03 and A04 in the second models A and B are nodes for integrating the first model and the second models A and B.

[0031] According to this third model, for example, a customer having a general characteristic A can be presumed to have a value of b toward a product or service and to select a product or service content I in a certain product or service category, and to have a value of d toward a product or service and to select a product or service content V in another product or service category.

[0032] Next, an example of a method of using the third model will be described with reference to Fig. 3. Fig. 3 shows a third model in which a first model indicated by reference symbol A11 and a second model indicated by reference symbol A12 are integrated at a node portion indicated by reference symbol A13. The first model is a Bayesian network representing a causal relationship between a plurality of nodes included in a layer (upper layer) indicating general characteristics such as general values ​​and lifestyles, and a plurality of nodes included in a layer (lower layer) indicating values ​​regarding cars. The second model is a Bayesian network representing a causal relationship between a plurality of nodes included in a layer (upper layer) indicating values ​​regarding cars, and a plurality of nodes included in a layer (lower layer) indicating features of cars selected by customers when purchasing a car (important points when purchasing a car). In these Bayesian networks of the first and second models, the causal relationships with the highest and second highest likelihoods are represented by thick and thin arrows, respectively.

[0033] Consider providing a communication service using the above model when a salesperson sells a car to a customer. In conventional communication services, for example, if a salesperson finds out that a customer places importance on the size of the trunk, he or she may simply recommend a car with a large trunk or an equipment option for the trunk. In this case, it is not possible to effectively deepen the bond with the customer.

[0034] In contrast, in a communication service using the third model as in this embodiment, for example, when a salesperson finds that a certain customer places importance on the size of the luggage compartment, the salesperson presumes that the customer has the value that a car is a means of transportation and has the general characteristic of being busy every day. In this case, based on the general characteristic of being busy every day, the salesperson can recommend a seat that is less tiring during travel or audio equipment to relieve stress during travel. Alternatively, the salesperson presumes that the customer has the value that a car is a space for family get-togethers and has the general characteristic of valuing family. In this case, based on the general characteristic of valuing family, the salesperson can recommend optional equipment that can be used for family camping. According to such a communication service using the third model of this embodiment, the true nature of the customer can be clearly understood based on pure values ​​and lifestyle, and appropriate communication can be performed based on pure values ​​and lifestyle that are not dependent on demographic characteristics such as age and gender. As a result, a deeper understanding of the customer can be obtained, and sympathy from the customer can be obtained, effectively deepening the bond with the customer.

[0035] Next, another example of a method of using the third model will be described with reference to Fig. 4. Fig. 4 shows a third model in which a first model indicated by reference symbol A21 and two second models indicated by reference symbols A22 and A23 are integrated at a node indicated by reference symbol A24. The first model is a Bayesian network that indicates the causal relationship between a plurality of nodes included in a layer (upper layer) indicating general characteristics such as general values ​​and lifestyles, and a plurality of nodes included in a layer (lower layer) indicating values ​​regarding food.

[0036] The second model, denoted by reference symbol A22, is a Bayesian network that represents the causal relationships between multiple nodes included in the layer (upper layer) that indicates values ​​regarding food and multiple nodes included in the layer (lower layer) that indicates foods selected by customers, particularly their favorite alcoholic beverages. The second model, denoted by reference symbol A23, is a Bayesian network that represents the causal relationships between multiple nodes included in the layer (upper layer) that indicates values ​​regarding food and multiple nodes included in the layer (lower layer) that indicates foods selected by customers, particularly their favorite dishes. In these first and second model Bayesian networks, the causal relationships with the highest and second highest likelihoods are represented by thick and thin arrows, respectively.

[0037] Consider providing a communication service using the above model when a salesperson sells wine to a customer. In conventional communication services, when a salesperson finds out that a certain customer likes wine, he or she may simply communicate with the customer to sell wine to that customer. In this case, the salesperson cannot effectively deepen the bond with the customer.

[0038] In contrast, in a communication service using the third model as in this embodiment, for example, when a salesperson finds out that a certain customer likes wine, the salesperson presumes that the customer has a value that food is a time to enjoy delicious food, and has the general characteristic of being free and relaxed. In addition, from the general characteristic of being free and relaxed, the salesperson presumes that the customer also has a value that food is a time and space to enjoy alone, and that the customer also likes sake and Japanese food. In this case, the salesperson can recommend sake and Japanese food in addition to recommending wine. Alternatively, when a salesperson finds out that a certain customer likes wine, the salesperson presumes that the customer also likes French food because the customer has a value that food is a time to enjoy delicious food, and can recommend French food (e.g. cheese) in addition to recommending wine.

[0039] The communication service using the third model according to this embodiment also allows for deeper understanding of customers, and thus allows for empathy from customers, effectively deepening ties with customers. In addition, the third model shown in FIG. 4 integrates the second models of different fields, alcohol and food, so that the relationship between products in different fields can be understood (for example, it can be seen that a customer who likes wine also likes Japanese sake, Japanese food, and French cuisine). Therefore, by using this third model, it becomes possible to develop effective collaborative products and tie-up products through cross-industry collaboration.

[0040] [Specific processing details] Next, specific processing contents of the customer analysis system 10 (specifically, the first information processing device 1 and the second information processing device 5) in this embodiment will be described.

[0041] First, the specific contents of the first model according to the present embodiment will be described with reference to Fig. 5. As shown in Fig. 5, the first model is roughly divided into a general life level layer consisting of a plurality of layers defined by items related to general life, and a specific area level layer consisting of a plurality of layers prepared separately for a plurality of product or service categories and defined by items corresponding to each category.

[0042] The general life level layer includes, from top to bottom, a personal attribute layer, a values ​​and personality traits layer, and a lifestyle layer. The personal attribute layer is a layer that includes nodes that indicate gender, age, education and academic background, occupation, family structure, place of residence, and place of origin (i.e., demography), life stage, etc. The values ​​and personality traits layer is a layer that includes nodes that indicate a person's values, worldview, personality (one example is a sensitive type, a cooperative type, a self-paced type, or a curious type). The lifestyle layer is a layer that includes nodes that indicate a person's lifestyle (one example is a type that values ​​connections with others, a type that values ​​a life that suits oneself, or a type that values ​​a safe life).

[0043] For example, lifestyles are analyzed using the AIO approach. This AIO approach attempts to capture the lifestyle characteristics of consumers in three aspects: Activities (what they spend their time on), Interests (what they are interested in), and Opinions (what they think about various events such as politics and social issues). By asking questions about these three aspects and demographic characteristics, the lifestyle of consumers in general or in specific areas of life or product categories is measured. In this general life level layer, nodes such as "introverts: bothered by human relationships, tired from work, single," "extroverts: like to socialize, sensitive," "parenting demographic: married, child-centered, female tendencies," and "meritocracy demographic: workaholics, male tendencies" are formed.

[0044] On the other hand, the specific domain level layer includes a hierarchical structure defined for each individual domain (specific domain) according to a predetermined industry classification, product / service classification, lifestyle classification, etc. In the example shown in FIG. 5, the specific domain level layer includes those related to the mobility domain, food domain, and living domain. These domains each include, from top to bottom, a value / lifestyle layer and an emotional benefit layer. The value / lifestyle layer includes nodes indicating values, lifestyles, behavioral characteristics, etc. for products or services in each domain. In one example, the value / lifestyle layer in the food domain includes nodes such as high-quality food, communication through food, healthy food, particular home cooking, not good at cooking, particular eating out, and easy eating out.

[0045] The emotional benefit layer is defined by the emotional benefits that indicate the positive emotions (sense of security, luxury, enjoyment, coolness, fulfillment, etc.) that users get from acquiring and using a product or service. In one example, the emotional benefit layer includes nodes such as extraordinary / self-improvement, relaxation / fulfillment, satisfying curiosity, and interaction / empathy with others. Furthermore, the nodes of the emotional benefit layer are located at the bottom layer of the first model and serve as nodes for integrating the first model with the second model (shown by hatching in Figure 5).

[0046] In order to create the first model including such a general life level layer and a specific area level layer, information obtained from one or more of the following is used: a questionnaire survey (Internet survey, interview survey, etc.) that does not limit the target area, a psychological test with an academic background (personality trait test, YG personality test, Big Five, etc.), a brain response test result (fMRI, EEG, etc.) that is a set of stimuli-responses, and a physiological response test result (heart rate, sweating, vascular stiffness, pupil diameter, eye movement, etc.) that is a set of stimuli-responses. In other words, the first model is created by statistically analyzing the information obtained in this way, for example, using clustering, factor analysis, Bayesian network, etc.

[0047] Next, the specific contents of the second model according to this embodiment will be described with reference to FIG. 6. FIG. 6(A) illustrates the second model related to the mobility domain, and FIG. 6(B) illustrates the second model related to the food domain. As shown in FIG. 6(A), the second model related to the mobility domain includes an emotional benefit layer desired from mobility, a lifestyle layer of car use, a functional benefit layer desired from a car (in other words, the important points desired from a car), a brand layer to be selected, and a car layer to be purchased and owned. Also, as shown in FIG. 6(B), the second model related to the food domain includes an emotional benefit layer desired from food, a lifestyle layer of bread eating, a functional benefit layer desired from a bread shop (in other words, the important points desired from a bread shop), a self-expression benefit layer desired when purchasing bread, a brand layer to be used, and a bread layer to be purchased.

[0048] Emotional benefits are as described above, while functional benefits refer to the positive effects on users that are brought about by the features of a product or service (convenient, fast, cheap, simple, tasty, light, durable, etc.). Self-expression benefits refer to the forms of self-expression and self-realization that are made possible by acquiring and using a product or service (being able to be yourself, having confidence in yourself, getting closer to your ideals, etc.). Emotional benefits and self-expression benefits are related to values, while functional benefits are related to what customers value about a specific product or service. Functional benefits, emotional benefits, and self-expression benefits are based on Aaker's three-category benefit classification.

[0049] As an example, in the second model for the food domain shown in Figure 6(B), the emotional benefit layer desired from food includes nodes such as extraordinary / self-improvement, relaxation / fulfillment, satisfying curiosity, and interaction / sympathy with others. The lifestyle layer for bread eating includes nodes such as orientation to specialty bakeries and orientation to major distribution manufacturers. The functional benefit layer desired from bread / stores includes nodes such as soft and gentle taste, flavor and chewiness of ingredients, moist texture, emphasis on ingredients, emphasis on deliciousness, emphasis on neatness of the store, and emphasis on handmade manufacturing methods at the store. The self-expression benefit desired when purchasing bread includes nodes such as recognition from others and emphasis on uniqueness.

[0050] In the second model for the mobility domain, the node of the emotional benefit layer desired from mobility is located at the top of the second model and serves as a junction for integrating the second model with the first model (see Fig. 5) (shown by hatching in Fig. 6(A)). In the second model for the food domain, the node of the emotional benefit layer desired from food is located at the top of the second model and serves as a junction for integrating the second model with the first model (see Fig. 5) (shown by hatching in Fig. 6(B)).

[0051] Such a second model is created based on information obtained from Internet surveys (with subject screening), store questionnaires, store interviews, etc., physiological responses of customers (heart rate, sweating, vascular stiffness, pupil diameter, eye movement, etc.) obtained while a salesperson is serving them, information obtained from store staff, customer purchase information obtained from the store, etc. In other words, the second model is created by statistically analyzing the information obtained in this way, for example, using clustering, factor analysis, Bayesian networks, etc.

[0052] Next, a process for creating a first model according to this embodiment will be described with reference to Fig. 7. Fig. 7 is a flowchart showing a process for creating a first model according to this embodiment. The process shown in this flowchart is executed by the first information processing device 1 (more specifically, the processor 1a).

[0053] In the flowchart shown in FIG. 7, steps S101 to S104 are performed to create a database structure for a first model, and steps S105 to S are performed to create the first model based on this database structure.

[0054] First, in step S101, the first information processing device 1 roughly divides the hierarchy of the first model into a general life level layer and a specific area level layer. Next, in step S102, the first information processing device 1 sets a type label in the database for the general life level layer, and also sets a type label in the database for the specific area level layer. For example, the first information processing device 1 sets a personal attribute label, a value / personality trait label, a lifestyle label, etc. for the database for the general life level layer. In addition, the first information processing device 1 sets a value / lifestyle label, an emotional benefit label, etc. for each product or service classification for the database for the specific area level layer.

[0055] Next, in step S103, the first information processing device 1 associates various data acquired for creating the first model with the type label set in step S102. Basically, the first information processing device 1 appropriately labels the acquired data. Note that the contents of a questionnaire survey (Internet survey, interview survey, etc.) conducted for creating the first model may be set so that data suitable for the set label is acquired. Next, in step S104, the first information processing device 1 determines a label to be applied to a node (node ​​node) that will be a node in the first model. For example, the first information processing device 1 determines a label of emotional benefit as the node node.

[0056] Next, in step S105, the first information processing device 1 creates nodes by compressing the dimensions of data within the labels set as described above. Next, in step S106, the first information processing device 1 models the relationships (correlation, for example, causal relationships) between the nodes thus created using a predetermined statistical method or analysis program. Next, in step S107, the first information processing device 1 forms a hierarchical structure of a general life level layer and a specific area level layer for the first model. Note that a further hierarchical structure may be set between the labels in these layers. Next, in step S108, the first information processing device 1 sets the hierarchical structure of the specific area level layer so that the nodal node is located at the bottom layer. In another example, the first information processing device 1 may set the label of the node located at the bottom layer in the hierarchical structure of the specific area level layer (i.e., the node located at the end in the causal analysis) to the nodal point.

[0057] Next, the creation process of the second and third models according to this embodiment will be described with reference to Fig. 8. Fig. 8 is a flowchart showing an example of the creation process of the second and third models according to this embodiment. The process shown in this flowchart is executed by the second information processing device 5 (more specifically, the processor 5a).

[0058] 8, steps S201 to S203 are for creating a database structure for the second model, steps S204 to S205 are for acquiring data for creating the second model, steps S206 to S208 are for creating the second model, and step S209 is for integrating the second model with the first model. This flowchart is also applicable to cases where research for creating the second model is performed after acquiring the information on the first model.

[0059] First, in step S201, the second information processing device 5 sets the label of the node node based on the information of the first model. For example, the second information processing device 5 determines the label of emotional benefit as the node node. Next, in step S202, the second information processing device 5 sets the type label of the database for the second model. For example, the second information processing device 5 sets the label of emotional benefit, the label of functional benefit, the label of self-expression benefit, and the label indicating the specific content of the product or service selected by the customer. Next, in step S203, the second information processing device 5 associates various data acquired for creating the second model with the type label set in step S202. Basically, the second information processing device 5 appropriately labels the acquired data. Note that the content of the questionnaire survey (Internet survey, interview survey, etc.) conducted for creating the second model may be set so that data suitable for the set label is acquired.

[0060] Next, in step S204, the second information processing device 5 acquires the identity information of the node node in order to directly inquire which of the nodes of the first model the node corresponds to. Next, in step S205, the second information processing device 5 acquires data corresponding to the set various labels.

[0061] Next, in step S206, the second information processing device 5 directly uses the pass / fail information as the node of the node label. Next, in step S207, the second information processing device 5 creates a node by dimensionally compressing the data within the label set as described above. Next, in step S208, the second information processing device 5 models the relationship (correlation, for example, causal relationship) between the nodes thus created using a predetermined statistical method or analysis program. Note that a further hierarchical structure may be set between each label.

[0062] Next, in step S209, the second information processing device 5 creates a third model by directly connecting the second model created as described above to the first model at the node of the node label.

[0063] 9 is a flowchart showing another example of the process for creating the second and third models according to the present embodiment. The process shown in this flowchart is also executed by the second information processing device 5 (more specifically, the processor 5a).

[0064] In the flowchart shown in Fig. 9, in steps S301 to S303, a process of creating a database structure for the second model is performed, in steps S304 to S305, data for creating the second model is acquired, in steps S306 to S309, a process of creating the second model is performed, and in steps S310 to S311, a process of integrating the second model with the first model is performed. This flowchart is also applied when an investigation for creating the second model is performed after acquiring the information of the first model. In addition, this flowchart is configured to create a third model by connecting a node node of the first model with one or more node nodes of the second model similar to the node node of the first model when a node node identical to the node node of the first model is not created in the second model.

[0065] The processing of steps S301 to S303 is similar to the processing of steps S201 to S203 in Fig. 8, and therefore description thereof will be omitted. In the following step S304, the second information processing device 5 acquires data (e.g., questionnaire question data) used for creating the node node of the first model. That is, the second information processing device 5 acquires the same data used for creating the node node of the first model for creating the second model. Next, in step S305, the second information processing device 5 acquires data corresponding to the set type-specific labels.

[0066] Next, in step S306, the second information processing device 5 creates nodes by dimensionally compressing data within the labels. Next, in step S307, the second information processing device 5 sets the number of node nodes of the second model to be equal to or greater than the number of node nodes of the first model. Next, in step S308, the second information processing device 5 models the relationships (correlation, for example, causal relationships) between the nodes using a predetermined statistical method or analysis program. Note that a further hierarchical structure may be set between each label. Next, in step S309, the second information processing device 5 sets the hierarchical structure of the second model so that the node nodes are located at the top layer.

[0067] Next, in step S310, the second information processing device 5 evaluates the similarity between the node nodes of the second model set as described above and the node nodes of the first model. For example, the second information processing device 5 compares the degree of coincidence of the contents of the answers by the customers. Next, in step S311, the second information processing device 5 creates a third model by connecting the second model to the first model through the node of the second model that is closest to the node node of the first model in the node nodes of the second model based on the evaluation result of step S310.

[0068] 10 is a flowchart showing yet another example of the process for creating the second and third models according to the present embodiment. The process shown in this flowchart is also executed by the second information processing device 5 (more specifically, the processor 5a).

[0069] In the flowchart shown in Fig. 10, step S401 is a process of creating a database structure for the second model, step S402 is a process of specifying data items with node labels, steps S403 to S406 are a process of creating the second model, and steps S407 to S408 are a process of integrating the second model with the first model. This flowchart is also applied when the results of a survey for creating the second model have already been obtained before obtaining the information on the first model. In this case, an item expressing the nature of the node is specified in the data used for the second model, and a standard for determining identity is created (including word matching and synonyms obtained by text mining).

[0070] First, in step S401, the second information processing device 5 sets a type label of the database for the second model. For example, the second information processing device 5 sets a label of emotional benefit, a label of functional benefit, a label of self-expression benefit, a label indicating the specific content of the product or service selected by the customer, and the like. Next, in step S402, the second information processing device 5 specifies a question corresponding to the question data constituting the node of the first model. In this case, the second information processing device 5 performs a judgment including the degree of agreement of words and synonyms by text mining. The processing of steps S403 to S408 after this step S402 is the same as the processing of steps S306 to S311 in FIG. 9, and therefore the description thereof will be omitted.

[0071] Next, FIG. 11 shows an example of the first to third models (Bayesian networks) actually created by this embodiment. FIG. 11 shows a third model in which the first and second models are integrated as a whole. Specifically, the symbol A31 indicates the first model, the symbol A32 indicates the second model related to the mobility domain (particularly cars), and the symbol A33 indicates the second model related to the food domain (note that the second model A33 includes models related to bread and alcohol). Furthermore, the symbol A34 indicates a node between the first model A31 and the second model A32, and the symbol A35 indicates a node between the first model A31 and the second model A33. Note that the countless arrows in FIG. 11 indicate connections with high likelihood (occurrence probability) between the nodes.

[0072] [Action and Effects] Next, the operation and effects of the customer analysis system 10 according to this embodiment will be described.

[0073] According to this embodiment, the customer analysis system 10 creates a third model by integrating a first model, which includes general characteristics such as pure values ​​and lifestyles in a person's life in general, which are not dependent on demographic characteristics such as age and gender, and defines a correlation between the general characteristics and the values ​​for a product or service, and a second model, which indicates the needs of a customer for a product or service, and defines a correlation between the customer's values ​​for a product or service and the contents of the product or service selected by the customer, by a part that indicates the same or similar values ​​for the product or service. This makes it possible to create a third model that accurately integrates the first model and the second model, and by providing a product or service using this third model, it is possible to provide a product or service suitable for the customer while deepening the bond with the customer. That is, according to this embodiment, the true nature of the customer can be clearly understood based on the pure values ​​and lifestyle, and appropriate communication based on the pure values ​​that are not dependent on demographic characteristics such as age and gender can be performed. As a result, a deeper understanding of the customer can be obtained, and the bond with the customer can be effectively deepened by gaining sympathy from the customer, in other words, a long-term relationship with the customer can be formed.

[0074] According to the present embodiment, the customer analysis system 10 creates a third model by integrating a first model, which defines the correlation between general characteristics and values ​​toward products or services for each of a plurality of product or service classifications, and a plurality of second models, which indicate customer needs for each of a plurality of product or service classifications. According to such a third model, the relationship between products or services in different classifications can be understood. Therefore, by using this third model, it becomes possible to develop effective collaborative products or tie-up products, for example, through collaboration between different industries. In addition, it becomes possible to develop products or services that utilize information from other industries and acquire new customers.

[0075] According to the present embodiment, the customer analysis system 10 uses a first model formed in a hierarchical structure so as to represent a direct or indirect causal relationship between a plurality of items (nodes) showing the contents of general characteristics and a plurality of items (nodes) showing the contents of values ​​regarding a product or service, and a second model formed in a hierarchical structure so as to represent a direct or indirect causal relationship between a plurality of items (nodes) showing the contents of values ​​regarding a product or service and a plurality of items (nodes) showing the specific contents regarding the product or service, and combines the bottom part of the hierarchical structure in the first model with the top part of the hierarchical structure in the second model. This makes it possible to more accurately integrate the first model and the second model.

[0076] Furthermore, according to this embodiment, the customer analysis system 10 creates a first model based on one or more of the results of a questionnaire, the results of a psychological test, the results of a brain response test, and the results of a physiological response test. This makes it possible to create a highly accurate first model, particularly by using the test results of the brain response and / or the physiological response in combination.

[0077] [Variations] In the above embodiment, the second information processing device 5 creates the second model, but in other examples, the first information processing device 1 may create the second model. In this case, the first information processing device 1 may acquire information required to create the second model from the second information processing device 5, and create the second model from the acquired information. Alternatively, the second information processing device 5 may create a model that is the basis of the second model, and the first information processing device 1 may edit the model acquired from the second information processing device 5 to create the second model. In such other examples, the first information processing device 1 (particularly the processor 1a) corresponds to an example of the "second model acquisition means" in the present invention.

[0078] In addition, in the above-described embodiment, the second information processing device 5 creates the third model, but in other examples, the first information processing device 1 may create the third model. In this case, the first information processing device 1 may create the third model by acquiring the second model from the second information processing device 5 and integrating the acquired second model into the first model. Alternatively, when the first information processing device 1 creates the second model as in the other examples described above, the third model may be created by integrating this second model into the first model. In such other examples, the first information processing device 1 (particularly the processor 1a) corresponds to an example of the "third model creation means" in the present invention.

[0079] In addition, in the configuration in which the first information processing device 1 creates the third model as described above, the customer analysis system 10 is configured without including the second information processing device 5, that is, the first information processing device 1 itself becomes the customer analysis system 10.

[0080] In the above embodiment, the Bayesian network formed in a hierarchical structure based on the causal relationship between nodes is shown as the first to third models, but the first to third models are not limited to using the Bayesian network. In another example, a model in which the relationship between nodes is defined by a correlation relationship that is not a causal relationship may be used. [Explanation of symbols]

[0081] 1. First information processing device 1a, 5a processors 1b, 5b memory 3. Communication lines 5. Second information processing device 10 Customer Analysis System

Claims

1. A customer analysis system that analyzes customers using a model, A first model acquisition means for acquiring a first model representing general characteristics of a person including at least values ​​and lifestyle in general life; A second model acquisition means for acquiring a second model indicative of customer needs for a product or service; a third model creating means for creating a third model by integrating the first model and the second model; having the general characteristics defining the first model are values ​​and lifestyles expressed independent of demographic characteristics indicative of at least age and gender; The first model includes a correlation between the general characteristics and a value perception toward a product or service; The second model includes a correlation between a customer's value regarding a product or service and the content regarding the product or service selected by the customer; the third model creation means creates the third model by integrating the first model and the second model based on parts of the first and second models that show the same or similar values ​​regarding products or services; A customer analysis system comprising:

2. The first model includes a correlation between the general characteristics and values ​​regarding the goods or services for each of a plurality of classes of goods or services; The second model acquisition means acquires a plurality of the second models indicative of customer needs for each of a plurality of product or service categories; the third model creation means creates the third model by integrating the first model and the plurality of second models according to the plurality of classifications of products or services. The customer analysis system according to claim 1 .

3. The first model is formed in a hierarchical structure so as to represent a direct or indirect causal relationship between a plurality of items indicating the contents of the general characteristics and a plurality of items indicating the contents of the values ​​regarding the product or service; The second model is formed in a hierarchical structure so as to represent a direct or indirect causal relationship between a plurality of items indicating the contents of values ​​regarding a product or service and a plurality of items indicating specific contents regarding the product or service; the third model creation means creates the third model by combining a part of a lowest layer of a hierarchical structure in the first model with a part of a highest layer of a hierarchical structure in the second model; The customer analysis system according to claim 1 or 2.

4. a first model creation means for creating the first model based on any one or more of a questionnaire result, a psychological test result, a brain response test result, and a physiological response test result; The first model acquisition means acquires the first model created by the first model creation means. The customer analysis system according to any one of claims 1 to 3.

Images