CUSTOMER ANALYSIS DEVICE

Abstract

A customer analysis device acquires replies from a user, who receives a financial service, to one or more defined first type questions and calculates an action indicator based on one or more replies in order to survey an action desire for the financial service. Next, the customer analysis device acquires replies from the user, who receives the financial service, to a plurality of defined second type questions and calculates a plurality types of emotion indicators based on the plurality of replies in order to survey a customer emotion for the financial service. The customer analysis device calculates correlation coefficients of the action indicator and each of the plurality types of emotion indicators, and graphically displays an influence of each of the plurality of emotion indicators on the action indicator based on the calculated correlation coefficients.

Description

BACKGROUND

Technical Field

The present invention relates to customer analysis, and more particularly, to a customer analysis technique for realizing customer-oriented business operations.

Related Art

In recent years, importance of customer experience has increased. The customer experience is not only a material, financial, objective, and rational value of goods and services, but also a subjective and emotional value that appeals to customer emotion for all customer experiences such as from promotion before purchase of goods or services to support after purchase of the goods or services.

Bernd H. Schmitt, a professor at Columbia University classifies emotional values into five types: sense (sensual), feel (emotional), think (intellectual), act (action), and relate (lifestyle). When a customer is stimulated by five senses due to deliciousness, good touch, and the like (sense) and stimulated by internal feelings such as cool and cute (feel), is required for intellectual desire or self-development (think), experiences a different life than before (act), and has a sense of belonging or sharing such as participating in activity (relate), the customer recognizes his/her emotional value, that is, the customer experience.

It has been known that the customer experience has a strong correlation with the profitability of a company, and advanced companies are focusing on enhancing the customer experience (see JP 2006-268405 A).

SUMMARY

However, there is no established method for measuring customer experience. For example, there has been proposed a method for performing a five-step evaluation that asks a customer “Are you satisfied with services of our company?”. However, companies that have been highly evaluated by this method do not necessarily gain high profitability. A more rational measurement method based on customer experience is required.

The present invention is an invention completed on the basis of the above problem recognition, and a main object of the present invention is to provide a technique for a method for setting an action indicator highly correlated with profitability and improving the action indicator.

A customer analysis device in an aspect of the present invention includes an action indicator calculation unit that obtains replies from a user, who receives a predetermined service, to one or more defined first type questions and calculates an action indicator based on one or more replies in order to survey an action desire for the service, an emotion indicator calculation unit that obtains replies from the user, who receives the service, to a plurality of defined second type questions and calculates a plurality types of emotion indicators based on the plurality of replies in order to survey a customer emotion for a service, a correlation calculation unit that calculates correlation coefficients for the action indicator and each of the plurality types of emotion indicators, and an influence display unit that graphically displays an influence of each of the plurality of emotion indicators on the action indicator based on the calculated correlation coefficient.

According to the present invention, it is easy to improve the customer experience.

BRIEF DESCRIPTION OF DRAWINGS

FIG. 1 is a schematic diagram for describing a relationship among a business indicator, an action indicator, and an emotion indicator;

FIG. 2 is a schematic diagram for describing the action indicator;

FIG. 3 is a diagram showing a survey result for a strength of correlation between a business intention and each of a recommendation intention and CX;

FIG. 4 is a data structure diagram of the emotion indicator;

FIG. 5 is a functional block diagram of a customer analysis device;

FIG. 6 is a flowchart showing a calculation process of a psychology coefficient of correlation;

FIG. 7 is a screen diagram of an expectation value screen;

FIG. 8 is a screen diagram of an evaluation value screen;

FIG. 9 is a screen diagram of an emotion indicator screen;

FIG. 10 is a screen diagram of a psychology correlation screen;

FIG. 11 is a screen diagram of a customer psychology analysis screen related to an Internet bank; and

FIG. 12 is a screen diagram of a customer psychology analysis screen related to a store type bank.

DETAILED DESCRIPTION

FIG. 1 is a schematic diagram for describing a relationship among a business indicator, an action indicator, and an emotion indicator.

The business indicator is an indicator indicating profitability of service provided by a company. As the business indicator, for example, sales, an operating profit ratio, a repeat rate, and the like may be considered. In the present embodiment, the customer experience is represented by two types of indicators, “action indicator” and “emotion indicator”. Note that in the present embodiment, the “service” is not limited to intangible services such as meals and lodging, but also includes a purchase of tangible items such as products.

The action indicator indicates strength of an action intention (for example, purchase desire) for services. The emotion indicator indicates how much a user who actually receives a service acts on what emotion the user is feeling. It is considered that when the user positively strengthens the action intention for the service, in other words, when the action indicator becomes high, the business indicator indicating the profitability is also improved. In addition, when the user has a positive emotion for the service, in other words, when the emotion indicator is large, the action indicator is considered to be high. The emotion indicator of the user who feels customer experience is considered to be high. A psychological model shown in FIG. 1 is a model obtained by modeling a correlation of three factors such as emotion that urges action and action that leads to profit.

A customer analysis device in the present embodiment analyzes a customer emotion in order to improve profitability (business indicator), and proposes what service is to be improved based on the analysis result.

In the following, a description of a financial service provided by a financial institution such as a bank or a securities company will be given.

However, there is a net promoter score (NPS) as an indicator for searching for the customer experience. The NPS was proposed in 2003 by Frederick F. Reichheld of Bain & Company, American consulting firm. First, the NPS asks a user to reply to questions such as “How likely are you to recommend this company (product, service, and brand) to a close friend or colleague?” within a range of 0 to 10 points. The NPS regards 9 points or more as “promoters”, 8 or 7 points as “passives”, and 6 points or less as “detractors”, and a value obtained by subtracting a ratio of detractors from a ratio of promoters in a user group to be surveyed as an indicator value. The NPS is recognized as having certain effectiveness, which is currently a de facto standard.

However, the NPS also has its weaknesses. In particular, a service that the NPS considers as a nuisance is a financial service. In the case of using the NPS to search for the customer experience in the financial service, it has been found that NPS tends to be very low. According to a survey conducted by the present applicant, about 80% of the respondents reply “I think so” or “if anything, I think so” to a question such as “Do you think financial institutions are not something to recommend to people?” (the number of samples: 12,612).

When we ask the reason to users who have thought (hereinafter, referred to as “noise thought”) that interferes with the analyzing of customer psychology as described above, most of the users replies “(choice of financial institution) is their own responsibility”, “no one is responsible when something happened”, “people have different expectations and values”, “I do not want to talk about money with others”, and the like.

The NPS does not work well for services that are prone to noise thought such as financial services “which should not be recommended to people”. Therefore, in the present embodiment, the following action indicator is proposed in order to compensate for the drawbacks of the NPS.

FIG. 2 is a schematic diagram for describing the action indicator.

In the present embodiment, the action indicator is decomposed into three elements, “recommendation intention”, “continuation intention”, and “purchase intention”.

As in the NPS, for the recommendation intention, we ask a user to evaluate a service on a scale of 10 points for the question “Can you recommend a service to a close friend or colleague?”. Hereinafter, an evaluation value of the recommendation intention is referred to as “recommendation value”. The recommendation values for each user are obtained.

The noise thought in recommendation intention is “(this service) is not something we recommend to others”. Since the recommendation intention includes such noise thought, the recommendation value tends to be low (reducing bias) as described above.

The continuation intention indicates user's desire to continue a service. For the continuation intention, we ask a user to evaluate a service on a scale of 10 points for the question “How likely are you to continue using our service in the future?”. For example, 10 points may be set for 100%, 9 points may be set for 90% or more and less than 100%, and 0 points may be set for less than 10%. Hereinafter, the evaluation value of the continuation intention is referred to as “continuation value”. The continuation values for each user are obtained.

Even when a user has noise thought about the recommendation intention, if you are satisfied with the service, it is considered to continue the service. Therefore, the continuation value complementarily functions for the reducing bias of the recommendation value.

On the other hand, the noise thought in the continuation intention is “I do not like the service but it is troublesome to change the service”. The continuation intention has such noise thought (inertial continuation), and therefore the continuation value tends to be high (increasing bias).

The purchase intention indicates a new use intention of a service by a user. For the purchase intention, we ask a user to evaluate a service on a scale of 10 points for the question “How likely are we to buy, sell, trade, and manage risk products in the future?”. For example, 10 points may be set for 100%, 9 points may be set for 90% or more and less than 100%, and 0 points may be set for less than 10%. Hereinafter, an evaluation value of the purchase intention is referred to as “purchase value”. The purchase values for each user are obtained.

For the continuation intention, even when the user has the noise thought, it is considered that if you don't like the service, you are less likely to buy new financial products. For this reason, the purchase value complementarily functions for the increasing bias of the continuation value.

The noise thought in the purchase intention is “I no longer want to buy anything”. Due to this noise thought, the purchase value tends to be low (decreasing bias).

Even when a user has the noise thought about the purchase intention, it is considered that if you are satisfied with the service, you are more likely to recommend the service to others. Therefore, the recommendation value complementarily functions for the reducing bias of the purchase value.

In the present embodiment, the recommendation value, the continuation value, and the purchase value are obtained based on the above-described three questions (hereinafter, referred to as “first type question”) for the recommendation intention, the continuation intention, and the purchase intention, and an average value of the recommendation value, the continuation value, and the purchase value are calculated as the action indicator. The action indicator is calculated for each user. Hereinafter, the action indicator thus calculated is referred to as “CX indicator” or simply “CX”. The present inventor hypothesized that in the CX indicator, by canceling the noise thought included in each of the recommendation intention, the continuation intention, and the purchase intention, it would be possible to better index the customer psychology.

FIG. 3 shows a survey result for the strength of correlation between the business intention and each of the recommendation intention and the CX.

The present inventor separately surveys a person (A: persons who prefer a bank) who deposits the most financial assets in the bank and a person (B: persons who prefer securities) who deposits the most financial assets in a securities company, as a depository of financial assets. In addition, those who have the noise thought about the recommendation intention are excluded from the survey target.

First, as the business indicator, five types, “(R1) deposit amount in a target financial institution”, “(R2) investment amount in risk product at a target financial institution”, “(R3) wallet share (a ratio of a deposit amount in a target financial institution to that in a total financial asset of a customer)”, “(R4) rate of change in a deposit amount in a target financial institution compared to 5 years ago”, and “(R5) rate of change in an investment amount in risk product at a target financial institution compared to five years ago” are defined.

As 862 persons who prefer a bank as a target, correlation coefficients of only the recommendation value (recommendation intention (NPS)) and each of the business indicators are obtained, and correlation coefficients of the NPS for “(R1) deposit amount” and “(R2) investment amount in risk product” are 0.7 or more, which shows a relatively strong correlation. On the other hand, the correlation coefficients of the recommendation value for “(R3) wallet share”, “(R4) rate of change in a deposit amount”, and “(R5) rate of change in an investment amount in risk product” fall below 0.7, which cannot indicate a very strong correlation.

On the other hand, as 2980 persons who prefer a bank as a target, correlation coefficients of the CX (average value of the recommendation value, the purchase value, and the continuation value) and each of the business indicators are obtained, and correlation coefficients of all four business indicators other than “(R1) deposit amount” are 0.7 or more. In addition, the correlation coefficient of “(R1) deposit amount” is 0.68, which slightly falls below the recommendation value. As far as we survey persons who prefer a bank as a target, it is confirmed that CX (three-valued average) has a strong positive correlation with various business indicators than only the recommendation value.

As 1758 persons who prefer securities as a target, correlation coefficients of the recommendation value and each of the business indicators are obtained, and the correlation coefficients of the NPS fall below 0.5 for all the five types of business indicators. For this reason, the NPS with only the recommendation value is not an appropriate indicator for searching for the customer experience of the persons who prefer securities.

On the other hand, as 6150 persons who prefer securities as a target, correlation coefficients of the CX (3-valued average) and each of the business indicators are obtained, and the correlation coefficients of all four business indicators other than the “(R3) wallet share” are 0.7 or more. In addition, the correlation coefficient of the “(R4) wallet share” is 0.67, which is at least a better result than only the recommendation value (NPS). For a person who prefers securities, it is confirmed that the CX has a clear advantage over the NPS.

As described above, despite analyzing a person who does not have the noise thought as the survey target, the CX as a whole has a stronger positive correlation with the business indicator than the NPS. In particular, it is confirmed that the CX is effective for persons who prefer securities. The hypothesis that the profitability can be improved by improving the CX as the action indicator has been confirmed, so it is necessary to review what measures should be taken to improve the CX as the next stage. In the present embodiment, as described with reference to FIG. 1, the emotion indicator that drives the action indicator (CX) is defined, and a measure for improving the action indicator is examined by analyzing the emotion indicator.

FIG. 4 is a data structure diagram of the emotion indicator.

The present applicant has collected and accumulated tens of thousands of questionnaires from a customer who receives a financial service. These questionnaires include free comments. The present inventor has analyzed a large number of free comments, and has concentrated various evaluations and requests regarding financial services on three axes of “reliability”, “convenience”, and “economic rationality”.

The reliability includes the following five types of evaluation points (hereinafter, referred to as “emotional points”).

(P1) Sympathy (Can servicer show sympathy for customer problem?)

(P2) Ability (Does servicer have expertise to respond to customer request?)

(P3) Personality (Do you think servicer is kind and works enthusiastically?)

(P4) Effectiveness of risk management (Is risk of financial service appropriately managed?)

(P5) Transparency (Is sincere response performed without lie or concealment?)

The convenience includes the following five types of emotion points.

(P6) Always (Is it possible to receive service always?)

(P7) Simple (Is it possible to receive desired service simply?)

(P8) Speedy (Is service executed quickly)

(P9) Easy to understand (Is proceeding and procedure for receiving service easy to understand)

(P10) Useful (Is service useful?)

The economic rationality includes the following three types of emotion points.

(P11) Cost (Are costs such as fees appropriate?)

(P12) Profit (Is it possible to obtain profit through a service)

(P13) Added value (Is it possible to feel added value?)

For each of the above 13 types of emotion points, a user can input a magnitude (hereinafter, referred to as “expectation value”) of expectation before receiving a service and a satisfaction level (hereinafter, referred to as “evaluation value”) after actually receiving the service based on a maximum of 10 points.

In the present embodiment, a value obtained by subtracting the expectation value from the evaluation value is calculated as the emotion indicator. For each user, 13 types of emotion indicators are calculated, corresponding to the 13 types of emotion points. Next, the correlation coefficients for the action indicators (CX) of each of the 13 types of emotion indicators are obtained (which will be described in detail below).

In order to analyze the emotion indicator, various questions (hereinafter, referred to as “second type question”) are prepared in advance for the user. For example, for “(P1) sympathy”, the question “Do you think servicer is making an effort to understand his/her own request?” is prepared. The emotion indicator for “(P1) sympathy” is calculated based on the user's reply (expectation value and evaluation value) to the question.

As other examples, for “(P8) speedy”, questions such as “Is it possible to trade without missing timing of instantaneous trading?”, for “(P11) cost”, questions such as “Do you think that fees are cheap?”, and the like are prepared. One second type question may be associated with one emotion point, or a plurality of second type questions may be associated with one emotion point. For example, for a certain emotion point PX, a weighted average expectation value and a weighted average evaluation value are calculated by performing a weighted average of reply values (expectation value and evaluation value) of each of the plurality of second type questions, and a weighted average evaluation value-weighted average expectation value may be calculated as the emotion indicator of the emotion point PX.

By calculating a difference value between the evaluation value and the expectation value as the emotion indicator, the emotion indicator indicates a size of a gap between expectation and reality. If the value is higher than expected, the emotion indicator becomes a positive value. On the other hand, when the value is lower than expected, the emotion indicator becomes a negative value. When there is a large gap between expectation and reality as a human psychological tendency, people are moved or dissatisfied. In the present embodiment, for the plurality of emotion points, the emotion indicator numerically indicates how much the customer's emotion is shaken.

FIG. 5 is a functional block diagram of a customer analysis device 100.

Each component of the customer analysis device 100 includes an arithmetic unit such as a central processing unit (CPU) and various coprocessors, storage devices such as a memory and a storage, hardware including a wired or wireless communication line connecting between the arithmetic unit and the storage device, and software stored in the storage device and supplying processing instructions to the arithmetic unit. A computer program may be configured by a device driver, an operating system, various application programs located in an upper layer thereof, and a library that provides a common function to these programs. Each block described below is not a configuration in a hardware unit but a block in a functional unit.

The customer analysis device 100 includes a user interface processing unit 110, a communication unit 114, a data processing unit 112, and a data storage unit 116.

The user interface processing unit 110 receives an operation from the user, and is in charge of processing related to the user interface such as image display or voice output. The communication unit 114 is in charge of communication processing with an external device via the Internet. The data storage unit 116 stores various data. The data processing unit 112 executes various processes based on data acquired by the user interface processing unit 110 and the communication unit 114, and data stored in the data storage unit 116. The data processing unit 112 also functions as an interface of the user interface processing unit 110, the communication unit 114, and the data storage unit 116

The user interface processing unit 110 includes an input unit 118 that receives an input from a user, and an output unit 120 that outputs various information such as image and voice to the user.

The output unit 120 includes an influence display unit 122 and an improvement notification unit 124. The influence display unit 122 displays the magnitude of the influence on the action indicators of each of the plurality of emotion indicators. Specifically, the influence display unit 122 displays a screen (described later) shown in FIGS. 7, 8, 9, 10, 11, and 12. The improvement notification unit 124 notifies an analyst of effective emotion points for improving the action indicator.

The communication unit 114 includes a transmission unit 132 that transmits data, and a reception unit 134 that receives data.

The data processing unit 112 includes an action indicator calculation unit 126, an emotion indicator calculation unit 128, and a correlation calculation unit 130. The action indicator calculation unit 126 calculates the action indicator based on the above-described first type question. Specifically, the action indicator calculation unit 126 collects the reply data of the first type question in a user group to be surveyed, calculates the recommendation value, the continuation value, and the purchase value by the above-described method, and calculates an average value of these three values as the action indicator. The action indicator calculation unit 126 calculates action indicators for each user. The action indicator calculation unit 126 may calculate the action indicator (average action indicator) for the entire group by calculating the average value of the action indicators of the user group to be surveyed.

The emotion indicator calculation unit 128 calculates the emotion indicator based on the above-described second type question. Specifically, the emotion indicator calculation unit 128 calculates the emotion indicator for each of the 13 emotion points for each user on an individual basis. The emotion indicator calculation unit 128 calculates the emotion indicators for each user. The emotion indicator calculation unit 128 may calculate the emotion indicators of the entire group for each emotion point by calculating the average value of the emotion indicators of the user group to be surveyed.

The correlation calculation unit 130 calculates the correlation coefficients of the emotion indicator and the action indicator obtained from the plurality of users. Specifically, first, the correlation calculation unit 130 collects, for the “(P1) sympathy”, the emotion indicators (hereinafter, referred to as “emotion indicator (P1)”) of the plurality of users. Next, the correlation calculation unit 130 calculates the correlation coefficient (hereinafter, referred to as “psychology coefficient of correlation”) for a set of the emotion indicators (P1) and a set of the action indicators of a plurality of users as a target based on the formula of a Pearson product-moment correlation coefficient. The correlation calculation unit 130 similarly calculates 13 types of psychology coefficients of correlation for each of the 13 types of emotion indicators.

FIG. 6 is a flowchart showing a calculation process of the psychology coefficient of correlation.

First, the reception unit 134 collects reply data based on the first type question and the second type question from a user who receives a financial service. First, when there is unprocessed reply data (Y in S10), the action indicator calculation unit 126 calculates the action indicator based on the reply to the first type question included in the new reply data (S12). Next, the emotion indicator calculation unit 128 calculates 13 types of emotion indicators based on the replay to the second type questions included in the reply data of the same user (S14). For one user, one type of action indicator and 13 types of emotion indicators are calculated. The same processing is repeated for all reply data.

After calculating the action indicators and the emotion indicators from all the reply data (N in S10), the correlation calculation unit 130 calculates the psychology coefficient of correlation for each emotion indicator (S16). The influence display unit 122 visualizes and displays the influence of the emotion indicator in a predetermined format based on the calculation result (S18). At this time, the improvement notification unit 124 also displays the improvement points. The improvement notification unit 124 notifies the analyst, as an improvement point, the emotion point where the psychology coefficient of correlation is equal to or more than a first threshold, for example, 0.5, and the emotion indicator is equal to or less than a second threshold, for example, 0.

FIG. 7 is a screen diagram of the expectation value screen 140.

The influence display unit 122 displays the expectation value screen 140 according to an instruction from an analyst. A vertical axis indicates the average value of the expectation value (hereinafter, referred to as “average expectation value”). A graph L1 shows a store type bank, and a graph L2 shows an Internet bank. For example, in the user group of the Internet bank, the average expectation value for the “(P3) personality” is “6.44”. On the other hand, in the user group of the store type bank, the average expectation value for the “(P3) personality” is “8.42”. Therefore, it can be seen that the user of the store type bank has a higher expectation for the personality of the servicer than when using the Internet bank.

In addition, in the user group of the Internet bank, the average expectation value for the “(P11) cos” is “8.08”. When using the Internet bank, it can be seen that the user has the expectation for the low cost. On the other hand, in the user group of the store type bank, the average expectation value for the “(P11) cost” is “7.40”. It can be seen that the user of the store type bank has the lower expectation for cost than the Internet bank.

In the user group of the store type bank, the average expectation value for the “(P2) ability” is “7.20”. On the other hand, it can be seen that in the user group of the Internet bank, the average expectation value for the “(P2) ability” is “6.43”, which indicates that expectations are relatively low.

FIG. 8 is a screen diagram of the evaluation value screen 142.

The influence display unit 122 displays the evaluation value screen 142 according to the instruction from the analyst. A vertical axis indicates the average value of the evaluation values (hereinafter, referred to as “average evaluation value”). In the user group of the Internet bank (graph L2), the average value of the evaluation value of the “(P2) ability” is “5.64”, and the average evaluation value of the “(P3) personality” is “5.78”. On the other hand, in the user group of the store type bank (graph L1), the average evaluation value of the “(P2) ability” is “7.00”, and the average evaluation value of the “(P3) personality” is “8.12”.

FIG. 9 is a screen diagram of an emotion indicator screen 144.

The influence display unit 122 displays the emotion indicator screen 144 according to the instruction from the analyst. The vertical axis indicates the average value of the emotion indicators (evaluation value-expectation value) of the plurality of users (hereinafter, referred to as “average emotion indicator”). As described above, a high emotion indicator indicates “satisfaction higher than expected”, and a low emotion indicator indicates “satisfaction lower than expected”.

In the user group of the store type bank (graph L1), the average emotion indicator (P4: effectiveness of risk management) is “−1.30”. This means that the user is strongly dissatisfied with “(P4) effectiveness of risk management” in the store type bank. On the other hand, in the user group of the store type bank, the average emotion indicator (P3: personality) is “−0.30”. This means that the “(P3) personality” in the store type bank generally meets the user's expectation.

In the user group of the Internet bank (graph L2), the average emotion indicator (P8: speedy) is “0.04”. This means that the Internet bank almost meets the user's expectation for the “(P8) Speedy”.

FIG. 10 is a screen diagram of a psychology correlation screen 146.

The influence display unit 122 displays the psychology correlation screen 146 according to the instruction from the analyst. The vertical axis indicates the psychology coefficient of correlation of the emotion indicator and the action indicator (CX). In the case of the store type bank (graph L1), the psychology coefficient of correlation of the “(P3) personality”, which is one of the emotion points, is “0.95”. Therefore, giving a good impression to the customer about the personality of the servicer is likely to lead to the improvement in the action indicator. According to the emotion indicator screen 144 of FIG. 9, since the average emotion indicator (P3: personality) is “−0.30”, which is never low, but it can be seen that there is room for further improvement for the “(P3) personality”.

In the store type bank, the psychology coefficient of correlation of the “(P4) effectiveness of risk management” is “−0.45”. Therefore, it is considered that the improvement in risk management is unlikely to lead to improvement in the action indicator. According to the emotion indicator screen 144 of FIG. 9, the average emotion indicator (P4: effectiveness of risk management) is very low as “−1.30”. The “(P4) effectiveness of risk management” has a lot of room for improvement, which is not unlikely to lead to the improvement in the customer experience.

The improvement notification unit 124 refers to the emotion indicator and the psychology coefficient of correlation to notify an analyst of an improvement point. As described above, in the present embodiment, the improvement notification unit 124 notifies, as an improvement point, an emotion point where the psychology coefficient of correlation is 0.5 or more and the emotion indicator is 0 or less. Therefore, according to the emotion indicator screen 144 of FIG. 9 and the psychology correlation screen 146 of FIG. 10, the improvement notification unit 124 proposes the “(P3) personality”, the “(P5) transparency”, the “(P6) always”, and the “(P8) speedy” as the improvement point for the store type bank (graph L1). Further, the improvement notification unit 124 proposes the “(P3) personality”, the “(P4) effectiveness of risk management”, and the “(P7) simple” as the improvement points for the Internet bank (graph L2).

FIG. 11 is a screen diagram of a customer psychology analysis screen 148 related to the Internet bank.

The influence display unit 122 displays the customer psychology analysis screen 148 as an analysis result for the user group of the Internet bank according to the instruction from the analyst. The horizontal axis indicates the average emotion indicator, and the vertical axis indicates the psychology coefficient of correlation. The right side indicates that the (average) emotion indicator is higher, that is, the degree of satisfaction with respect to the expectation is higher. The upper side indicates that the psychology coefficient of correlation is higher, that is, the emotion indicator and the action indicator are strongly positively correlated.

According to the customer psychology analysis screen 148, the emotion indicator for the convenience is generally large, and the psychology coefficient of correlation is also large. For this reason, it is considered that in order to further improve the convenience of the Internet bank, the profitability can be realized by concentrating the management resources.

FIG. 12 is a screen diagram of a customer psychology analysis screen 150 related to the store type bank.

The influence display unit 122 displays the customer psychology analysis screen 150 as the analysis result for the user group of the store type bank according to the instruction from the analyst. The horizontal axis indicates the average emotion indicator, and the vertical axis indicates the psychology coefficient of correlation.

According to the customer psychology analysis screen 150, some of the emotion indicators related to the convenience are considered to greatly contribute to the improvement in the action indicator and furthermore, to the profitability, but the effect on Internet banking cannot be expected. On the other hand, for the “(P3) personality”, it is already possible to provide a high level of satisfaction, but there is room for further improvement. In addition, it is understood that the emotion indicator of the “(P01) sympathy” is high, but it is difficult to improve the action indicator.

As above, the customer analysis device 100 has been described based on the embodiment.

According to the present embodiment, by defining the action indicator by a plurality of elements that cancel each other's noise thought, it is possible to enhance the positive correlation between the profitability and the action indicator. In particular, it is understood that for “service that is hard to recommend to a human” such as a financial service, the CX, which is a new action indicator, has a higher correlation than the NPS.

In the present embodiment, the plurality of emotion indicators are set as emotional factors affecting the action indicator. By calculating the psychology coefficient of correlation as the magnitude of the influence of each emotion indicator on the action indicator, it is possible to search for effective emotion points to improve the action indicator. According to such a control method, it is easy to efficiently and rationally reform organization business by intensively inputting limited management resources to improve appropriate emotion points.

Note that the present invention is not limited to the above embodiments and modifications, and can be embodied by modifying the components without departing from the gist. Various inventions may be formed by appropriately combining a plurality of components disclosed in the above embodiments and modifications. In addition, some components may be deleted from all the components shown in the above-described embodiments and the modifications.

[Modification]

In the present embodiment, the description has been given in which the average value of the recommendation value, the continuation value, and the purchase value is calculated as the action indicator. As the modification, the action indicator may be calculated by performing the weighted average on these three values. For example, a ternary weighting factor may be adjusted so that a positive correlation coefficient with a business indicator indicating profitability is increased.

Claims

1. A customer analysis device, comprising:

an action indicator calculation unit that acquires replies from a user, who receives a predetermined service, to one or more defined first type questions and calculates an action indicator based on one or more replies in order to survey an action desire for the service;

an emotion indicator calculation unit that acquires replies from the user, who receives the service, to a plurality of defined second type questions and calculates a plurality types of emotion indicators based on the plurality of replies in order to survey a customer emotion for the service;

a correlation calculation unit that calculates correlation coefficients for the action indicator and each of the plurality types of emotion indicators; and

an influence display unit that graphically displays an influence of each of the plurality of emotion indicators on the action indicator based on the calculated correlation coefficient.

2. The customer analysis device according to claim 1, wherein the action indicator calculation unit calculates the action indicator based on a reply from a user to each of a first type question for surveying a purchase desire of the service, a first type question for surveying a continuation desire of the service, and a first type question for surveying a recommendation desire of the service.

3. The customer analysis device according to claim 1, wherein the emotion indicator calculation unit acquires two types of replies of a degree of expectation before receiving the service and a degree of satisfaction after receiving the service, and acquires a difference value between the degree of satisfaction and the degree of expectation as a reply to the second type question.

4. The customer analysis device according to claim 1, further comprising:

an improvement notification unit that notifies a user of a second type question in which the correlation coefficient with the action indicator is equal to or greater than a first threshold and the difference value as the reply is equal to or less than a second threshold.

5. A customer analysis program causing a computer to execute:

a function of acquiring replies from a user, who receives a predetermined service, to one or more defined first type questions and calculating an action indicator based on one or more replies in order to survey an action desire for the service;

a function of acquiring replies from the user, who receives the service, to a plurality of defined second type questions and calculating a plurality types of emotion indicators based on the plurality of replies in order to survey a customer emotion for the service;

a function of calculating correlation coefficients for the action indicator and each of the plurality types of emotion indicators; and

a function of graphically displaying an influence of each of the plurality of emotion indicators on the action indicator based on the calculated correlation coefficient.