JP7348646B2 - Solution proposal device, solution proposal method, and solution proposal program - Google Patents

Description

The present invention relates to a solution proposing device, a solution proposing method, and a solution proposing program.

Traditionally, as a method of customer analysis, an analysis method called RFM analysis has been used to rank customers based on three indicators: Recency (date of recent purchase), Frequency (frequency of store visits), and Monetary (volume of purchase amount). .

For example, Patent Document 1 discloses a system that extracts customers who meet conditions from the results of RFM analysis and distributes direct mail to the extracted customers.

JP2003-216867

However, in the system of Patent Document 1, RFM analysis is simply used for direct mail distribution management. Therefore, the original purpose of RFM analysis, which is based on the characteristics of the segment to which the customer belongs, such as newly visiting members, dormant members, etc., and planning and implementing measures to revitalize customer behavior cannot be done as a system. It had not been done.

The purpose of the present invention is to provide a solution proposing device capable of solving the above-mentioned problems and planning and implementing measures to revitalize customer behavior based on customer analysis such as RFM analysis. , a solution proposal method, and a solution proposal program.

In order to solve the above problems, one aspect of the solution proposing device of the present invention is as follows:
an acquisition unit that acquires at least identification information that identifies a user and a plurality of elements representing the user's actions;
a plurality of rows that display one element among the plurality of elements in stages; a plurality of columns that display another element among the plurality of elements in stages and intersect with the plurality of rows; Displaying a matrix comprising cells at intersections of the plurality of rows and the plurality of columns, the cells displaying the number of users belonging to each stage of the one element and each stage of the other element. a display control section;
When the cell is selected, it is possible to set the execution conditions and execution details of the push or pull type measures using communication means such as the Internet for the users whose numbers are displayed in the cell. a setting section,
an execution unit that executes, for the user, the push-type or pull-type measure for which the execution conditions and execution details are set by the setting unit;
Equipped with

According to the above aspect, the matrix displayed by the display control unit includes a plurality of rows in which one of the plurality of elements representing the user's behavior is divided into stages and displayed, and another element among the plurality of elements is displayed. It is displayed in stages, has multiple rows and multiple columns that intersect, and at the intersection of multiple rows and multiple columns, users belonging to each stage of one element and each stage of other elements are displayed. It has a cell that displays the number of . When a cell is selected, it is possible to set the execution conditions and execution details of the push or pull type measures using communication means such as the Internet for the users whose numbers are displayed in the cell. It becomes. Further, the push-type or pull-type measure in which the execution conditions and the execution contents are set is executed for the user by the execution unit. Therefore, by looking at the matrix, it is possible to easily understand the distribution of user behavior, and furthermore, it is possible to propose measures based on user characteristics, implement measures for users, and find appropriate solutions. It becomes possible to make proposals.

Another aspect of the present invention further includes a policy presentation unit that displays an example of the policy when the cell is selected. According to this aspect, the measure presentation unit displays an example of the measure when the cell is selected. Therefore, it is possible to classify user behavior by cell and take appropriate measures for the user.

In another aspect of the present invention, the user's behavior is the user's visit behavior or purchase behavior at a physical store, and the elements include the number of days indicating the most recent visit, the number of visits, the cumulative purchase amount, Alternatively, the acquisition unit includes a value obtained by scoring from these actions, and the acquisition unit represents at least the identification information and the actions of the user from a camera device with facial authentication or a payment system installed in the physical store. Get multiple elements. According to this aspect, the user's purchasing behavior can be analyzed using the matrix based on the number of days indicating the most recent visit, the number of visits, the cumulative purchase amount, or a value obtained by scoring from these behaviors. It is possible to propose and implement measures to improve users' purchasing behavior.

In another aspect of the present invention, the user's behavior is the user's purchasing behavior on a website or access to the website, and the elements include a number of days indicating the most recent visit, a number of visits, The acquisition unit includes at least the identification information and a plurality of pieces of information representing the user's actions, including the cumulative purchase amount or a value scored from these actions. Get the element. According to this aspect, the matrix determines the purchasing behavior of the user or the user's behavior toward the website based on the number of days indicating the most recent visit, the number of visits, the cumulative purchase amount, or the value scored from these behaviors. Access can be analyzed and measures to improve user purchasing behavior or website access can be proposed and implemented.

In order to solve the above problems, one aspect of the solution proposal method of the present invention is as follows:
a step of acquiring, by the acquisition unit, identification information for identifying the user and a plurality of elements representing the user's behavior;
The display control unit displays a plurality of rows in which one element among the plurality of elements is displayed in stages, and another element among the plurality of elements is displayed in stages and intersects with the plurality of rows. a plurality of columns; and a cell at an intersection of the plurality of rows and the plurality of columns, the cell displaying the number of users belonging to each stage of the one element and each stage of the other element. displaying the prepared matrix;
When the cell is selected, the execution conditions and execution details of the push-type or pull-type measure using communication means such as the Internet for the user whose number is displayed in the cell by the setting unit. etc., and the step of setting
The method further includes a step of causing an execution unit to execute the push-type or pull-type measure for the user, in which the execution condition, the execution content, etc. are set.

According to the above aspect, the matrix includes a plurality of rows in which one element out of a plurality of elements representing user behavior is displayed in stages, and another element out of the plurality of elements is displayed in stages. , has multiple rows and multiple columns that intersect, and at the intersection of the multiple rows and multiple columns, a cell that displays the number of users who belong to each stage of one element and each stage of the other element. has. When a cell is selected, it is possible to set the execution conditions and execution details of the push or pull type measures using communication means such as the Internet for the users whose numbers are displayed in the cell. It becomes. Further, the push-type or pull-type measure in which the execution conditions and the execution contents are set is executed for the user by the execution unit. Therefore, by looking at the matrix, you can easily understand the distribution of user behavior, propose measures based on user characteristics, implement measures for users, and propose appropriate solutions. It becomes possible.

In order to solve the above problems, one aspect of the solution proposal program of the present invention is to
A solution proposal program executable by a computer,
The solution proposal program causes the computer to:
obtaining identification information for identifying a user and a plurality of elements representing the user's behavior;
a plurality of rows that display one element among the plurality of elements in stages; a plurality of columns that display another element among the plurality of elements in stages and intersect with the plurality of rows; Displaying a matrix comprising cells at intersections of the plurality of rows and the plurality of columns, the cells displaying the number of users belonging to each stage of the one element and each stage of the other element. step and
When the cell is selected, the execution conditions and execution contents of the push-type or pull-type measure using communication means such as the Internet are set for the user whose number is displayed in the cell. step and
executing the push-type or pull-type measure, in which the execution conditions and the execution contents are set, for the user;

According to the above aspect, the matrix includes a plurality of rows in which one element out of a plurality of elements representing user behavior is displayed in stages, and another element out of the plurality of elements is displayed in stages. , has multiple rows and multiple columns that intersect, and at the intersection of the multiple rows and multiple columns, a cell that displays the number of users who belong to each stage of one element and each stage of the other element. has. When a cell is selected, it is possible to set the execution conditions and execution details of the push or pull type measures using communication means such as the Internet for the users whose numbers are displayed in the cell. It becomes. Further, the push-type or pull-type measure in which the execution conditions and the execution contents are set is executed for the user by the execution unit. Therefore, by looking at the matrix, you can easily understand the distribution of user behavior, propose measures based on user characteristics, implement measures for users, and propose appropriate solutions. It becomes possible.

According to the present invention, it is possible to plan and implement measures to revitalize customer behavior based on customer analysis such as RFM analysis.

1 is a diagram schematically showing a solution proposal system according to an embodiment of the present invention. FIG. 3 is a diagram showing functional blocks of a server in one embodiment. It is a diagram showing functional blocks of a terminal device in one embodiment. It is a diagram showing functional blocks of a solution proposal device in one embodiment. FIG. 3 is a diagram showing functional blocks of a control unit in a solution proposal device in an embodiment. It is a flowchart showing the operation of the solution proposal device in one embodiment. FIG. 3 is a diagram illustrating a user interface of a solution proposal device in one embodiment. FIG. 3 is a diagram illustrating a user interface of a solution proposal device in one embodiment. FIG. 3 is a diagram illustrating a user interface of a solution proposal device in one embodiment. FIG. 3 is a diagram illustrating a user interface of a solution proposal device in one embodiment. FIG. 3 is a diagram illustrating a user interface of a solution proposal device in one embodiment.

<First embodiment>
Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram schematically showing a solution proposal system of this embodiment. FIG. 2 is a diagram showing functional blocks of the server. FIG. 3 is a diagram showing functional blocks of the terminal device. FIG. 4 is a diagram showing functional blocks of the solution proposal device 3.

(Solution proposal system)
As shown in FIG. 1, the solution proposal system 100 of this embodiment includes a terminal device 1, a network 2 such as the Internet, a solution proposal device 3 accessible via the network 2, and a server 4.

The terminal device 1 is, for example, an information processing device such as a personal computer, a tablet, or a smartphone, and is a device that can access a shopping site or the like using a browser via the network 2 .

The network 2 is a communication means that enables data communication between the terminal device 1 and the server 4 and between the server 4 and the solution proposal device 3, and specific examples include the Internet or an intranet. It will be done. In this embodiment, the Internet is used as an example.

The solution proposal device 3 is an information processing device, such as a personal computer or a tablet, which can communicate with the server 4 via the network 2, and the solution proposal program of the present invention is installed in the solution proposal device 3 in advance.

As shown in FIG. 2, the server 4 includes a server communication section 40, a server control section 41, and a server storage section 42. The server communication unit 40 is an interface that performs data communication using the network 2, and is controlled by the server control unit 41. In this embodiment, as an example, the Internet is used as the network 2, and the server communication unit 40 is an interface that communicates data via the Internet.

The server control unit 41 is composed of a CPU, etc., and enables browsing of shopping sites, etc. on the terminal device 1 by executing programs stored in the server storage unit 42, and allows the terminal device 1 to access shopping sites, etc. The analysis data is stored in the server storage unit 32. In this embodiment, access analysis data is acquired using an access analysis application program that is publicly available on the Internet. However, the present invention is not limited to this embodiment. By storing an access analysis application program in the server storage unit 42 and executing this access analysis application program, the access analysis data of the user of the terminal device 1 is sent to the solution proposal device 3 in response to a request from the solution proposal device 3. It may also be provided.

The server storage unit 32 is composed of, for example, an HDD (Hard Disk Drive). The server storage unit 32 stores data and programs for building a shopping site or the like. Further, the server storage unit 32 may store access analysis data by the user of the terminal device 1, a program for acquiring the access analysis data, and the like.

As shown in FIG. 3, the terminal device 1 includes a display section 10, an input section 11, a terminal control section 12, a playback section 13, a terminal communication section 14, and a terminal storage section 15.

The display unit 10 displays a screen of a shopping site or the like. Examples of the display unit 10 include a liquid crystal display, an organic EL display, and the like.

The input unit 11 inputs operation instructions from the user. The input unit 11 includes, for example, a touch panel integrated with the display unit 10, a keyboard, and the like.

The terminal control section 12 is composed of a CPU and the like, and controls the display section 10, the input section 11, the terminal communication section 14, and the terminal storage section 15 by starting an application program.

The terminal communication unit 14 is an interface that performs data communication using the network 2, and is controlled by the terminal control unit 12. In this embodiment, as an example, the Internet is used as the network 2, and the terminal communication unit 14 is an interface that communicates data via the Internet.

The terminal storage unit 15 is composed of an SSD (Solid State Drive) or the like, and an application program for executing the audio data processing of the present invention is installed in advance in the terminal storage unit 15.

As shown in FIG. 4, the solution proposal device 3 includes a display section 30, an input section 31, a control section 32, a communication section 33, and a storage section 34.

The display unit 30 displays a screen for proposing solutions according to the present invention. Examples of the display section 30 include a liquid crystal display, an organic EL display, and the like.

The input unit 31 inputs operation instructions from the user. The input unit 31 includes, for example, a touch panel integrated with the display unit 30, a keyboard, and the like.

The control unit 32 is composed of a CPU and the like, and controls the display unit 30, the input unit 31, the communication unit 33, and the storage unit 34 by starting the solution proposal program of the present invention.

The communication unit 33 is an interface that performs data communication using the network 2, and is controlled by the control unit 32. In this embodiment, as an example, the Internet is used as the network 2, and the communication unit 33 is an interface that communicates data via the Internet.

The storage unit 34 is composed of an SSD (Solid State Drive) or the like, and the solution proposal program of the present invention is installed in the terminal storage unit 15 in advance. Furthermore, the storage unit 34 stores identification information for identifying the user of the terminal device 1, an access destination such as a shopping site by the user, and a plurality of elements representing the user's behavior in association with each other. Details of the identification information, access destination, and multiple elements representing user behavior will be described later.

(Functional block)
FIG. 5 shows functional blocks of the control unit 32 in the solution proposal device 3. The control unit 32 of the solution proposal device 3 functions as an acquisition unit 320, a display control unit 321, a policy presentation unit 322, a setting unit 323, an execution unit 324, and a machine learning unit 325, as shown in FIG.

The acquisition unit 320 acquires at least identification information that identifies the user and a plurality of elements representing the user's behavior by accessing the access analysis program. The access analysis program is a program for acquiring access analysis data by the user of the terminal device 1. The access analysis program may be a program provided by a third party, it may use the access log of the server, or it may create a dedicated program for this system and use it. User behavior is, for example, user purchasing behavior on a website or access to the website, and multiple elements representing user behavior include the number of recent visits, number of visits, and cumulative purchases. Contains monetary amounts or values scored from these actions. Note that by setting the access analysis program, it is possible to link the user ID of the portal site with identification information that identifies the user in the access analysis program. Alternatively, the solution proposal program may import identification information for identifying users in the access analysis program, and link the identification information with the user ID of the portal site.

The display control unit 321 causes the display unit 30 to display a solution proposal matrix. Details of the solution proposal matrix will be described later.

When a cell in the solution proposal matrix is selected, the policy presentation unit 322 displays an example of a policy for the user associated with the cell. The number of users with a predetermined attribute is displayed in the cell, and the users with the attribute are associated with each other. Details of cells, predetermined attributes, the number of users, and policies for users will be described later.

The setting unit 323 sets execution conditions, execution details, etc. of a push-type or pull-type measure using a communication means such as the Internet for the user whose number is displayed in the cell. Details of the push-type or pull-type measures, as well as the conditions for implementing the measures and the details of the measures, will be described later.

The execution unit 324 executes the measures set by the setting unit 323 as appropriate.

The machine learning unit 325 uses the identification information for identifying the user acquired by the acquisition unit 320, multiple elements representing the user's behavior, and the measures set by the setting unit 323 as training data, and uses deep learning or the like to Evaluate the optimal measures for each segment to which users belong. "Multiple elements representing user behavior" include reactions to sent emails, reactions such as clicks in the case of A/B tests, time until opening in the case of push notifications, and visits to physical stores. Frequency, access frequency of web pages, etc. Such data can be used as teacher data.

(Operation example)
Next, an example of the operation of this embodiment will be described with reference to FIGS. 6 to 11. FIG. 6 is a flowchart showing the operation of the solution proposal device 3 in this embodiment. 7 to 11 are diagrams showing user interfaces of the solution proposal device 3 in this embodiment.

In this embodiment, as an example, the user of the solution proposal device 3 operates a membership-based portal site that posts news, etc., and in order to increase access to this portal site by the users of the terminal device 1, the user We will explain how to implement these measures.

In order for the user of the terminal device 1 to become a member of the portal site, it is necessary to input items such as name, address, and e-mail address on the membership registration page. The user ID of the member may be arbitrarily determined by the user of the terminal device 1, or may be automatically assigned by the solution proposal program. Membership information such as the user ID and email address of the user of the terminal device 1 who has completed membership registration can be communicated with the storage unit 34 of the solution proposal device 3, the server storage unit 42 of the server 4, or the solution proposal device 3. It is stored in the storage section of another server connected to the network 2. In this embodiment, as an example, it is assumed that the member information is stored in the storage unit 34 of the solution proposal device 3.

The user of the solution proposal device 3 starts a solution proposal program in the solution proposal device 3 when taking measures for users to increase access to the portal site. The user of the solution proposal device 3 selects execution of creation of a matrix for solution proposal on the user interface of the solution proposal program (not shown).

When execution of creation of a matrix for solution proposal is selected, the acquisition unit 320 acquires access analysis data for the portal site during a predetermined period using an access analysis program that is publicly available on the Internet. (Figure 6: S1). The control unit 32 causes the storage unit 34 to store the acquired access analysis data.

The access analysis data includes the number of sessions (visits) of each user of the terminal device 1 to the portal site, the last access date, the number of accesses to articles by category, and the like. Such access analysis data is stored in the storage unit 34 in association with a user ID as identification information for identifying a user. Therefore, the storage unit 34 stores the user ID as identification information for identifying the user, the user's access destination (for example, the URL of the portal site), and a plurality of elements representing the user's behavior (for example, the session for the portal site). (number of visits), last access date, number of accesses to articles by category, etc.) are stored in association with each other.

Next, the display control unit 321 creates and displays a matrix for solution proposal ((FIG. 6: S2). First, the display control unit 321 classifies the access analysis data. The display control unit 321 The display control unit 321 classifies the number of sessions (visits) and the last access date to the portal site as multiple elements representing user behavior.The display control unit 321 sets the number of sessions (visits) to 90 or more and 30 or more, for example. , 10 or more sessions, 2 or more sessions, and 1 session, and classifies the user IDs to belong to these sessions.Next, the display control unit 321 classifies the last access date from 0 to 1 day ago, from 2 to 10 days ago. The user IDs are divided into the previous day, 10 to 30 days ago, and 30 to 90 days ago, and the classified user IDs are further classified as belonging to these last access dates.

Next, the display control unit 321 displays on the display unit 30 a solution proposal matrix in which the graded last access date is expressed in multiple rows and the graded number of sessions (visits) is expressed in multiple columns. let FIG. 7 shows an example of a matrix 50 for solution proposal. As shown in FIG. 7, the target period of the acquired access analysis data is displayed in the upper row 51 of the matrix 50. Further, in the plurality of rows 52 of the matrix 50, the last access date is displayed in stages of 0 to 1 day ago, 2 to 10 days ago, 10 to 30 days ago, and 30 to 90 days ago. Furthermore, the number of sessions (visits) classified into 90 or more times, 30 or more times, 10 or more times, 2 or more times, and 1 time is displayed in a plurality of columns 53 of the matrix 50.

In the matrix 50, a cell 54 at the intersection of a plurality of rows 52 and a plurality of columns 53 displays the graded last access date and the number of users belonging to the graded number of sessions (visits). Ru. The number of users displayed in cell 54 is determined by counting the number of user IDs belonging to each last access date and number of sessions (visits).

By displaying such a solution proposal matrix 50 on the display unit 30, it is possible to understand at a glance what class the user belongs to. For example, in the matrix 50 shown in FIG. 7, the number of users whose last access date was 0 to 1 day ago and who had one session (visit) is displayed in the cell 54 as "5000 users." Therefore, it can be understood at a glance that there are 5000 new users. Furthermore, in the matrix 50 shown in FIG. 7, the number of users whose last access date was 30 to 90 days ago and whose number of sessions (visits) was 90 or more is displayed as "2000 users" in the cell 54. Therefore, it can be understood at a glance that there are 2000 dormant or former active users.

FIG. 8 shows a display example when any cell 54 of the matrix 50 is clicked. In the example shown in FIG. 8, when the cell 54 in which the last access date was 2 to 10 days ago, the number of sessions (visits) was 10 or more, and the number of users was "1000 users" was clicked in the matrix 50. It shows. As shown in FIG. 8, when a certain cell 54 is clicked, the display control unit 321 displays the cell 54 so that the background color of the cell 54 is different from that of other cells 54.

Furthermore, when a certain cell 54 is clicked (FIG. 6: S3), the policy presentation unit 322 displays an example of the policy to the user associated with the cell 54 (FIG. 6: S4).

First, when a certain cell 54 is clicked (FIG. 6: S3), the measure presentation unit 322 displays a measure corresponding to the cell 54 in the measure display area 55 at the bottom of the matrix 50, as shown in FIG. An example of measures for the user is displayed (FIG. 6: S4). An example of the measures may be set in advance for each cell 54.

Furthermore, the machine learning unit 325 may automatically suggest an example of an optimal facility according to the user's conditions, the results of the measures, etc., using machine learning, deep learning, or the like. In this case, the machine learning unit 325 uses the identification information for identifying the user acquired by the acquisition unit 320, multiple elements representing the user's behavior, and the measures set by the setting unit 323 as training data for deep learning. etc., to evaluate the optimal measures for each segment to which the user belongs. "Multiple elements representing user behavior" include reactions to sent emails, reactions such as clicks in the case of A/B tests, time until opening in the case of push notifications, and visits to physical stores. Frequency, access frequency of web pages, etc.
One example is the analysis of email content and reactions to the email. For example, if the content of the email uses the keywords ``summer sale'' or ``〇〇 campaign,'' it will analyze how people reacted to those keywords. Based on this analysis, the content of the next email to be sent can be suggested.
Furthermore, when the content of the email includes product recommendations, the recommended products may be sent out depending on what kind of products the user has reacted to.
The above data can be used as teacher data. The policy presentation unit 322 may display the highly evaluated policy obtained by the machine learning unit 325 as the optimal policy.

In the example shown in FIG. 8, the proposed measure is set for the cell 54 whose last access date was 2 to 10 days ago, the number of sessions (visits) was 10 or more, and the number of users was displayed as "1000 users." Is displayed. In the policy proposal display area 55, the attributes of the user associated with this cell 54 indicate that the user has visited the site ``2 to 10 days before the last access date and 10 or more times but less than 30 times.'' Under this attribute, the characteristic of the user associated with this cell 54 is ``more accesses to ``sports'' articles than the tendency of total PV. ” is shown. The characteristics of which articles are accessed the most can be extracted from access analysis data.

Under this feature, two facility proposals are shown. As the first facility plan, a facility plan that says "Send an access e-mail to users who have not accessed the site for 10 days or more" is displayed, and on the right side of this first facility plan, there are a setting button 56a and a delete button 56b. is displayed. Also, under the first facility proposal, the second facility proposal is ``Publish a ``Sports'' article once every three days. ” is displayed. A setting button 57a and a delete button 57b are also displayed on the right side of this second facility proposal.

When the user of the solution proposal device 3 clicks the setting button 56a (FIG. 6: S5; YES), the setting unit 323 performs the policy setting process (FIG. 6: S6) with “last access date 2 to 10 days ago and 10 times”. A measure is set to send an access invitation email to users who have visited the site less than 30 times, that is, to users who have not accessed the site for 10 days or more.

In addition, the execution unit 324 sends an access invitation email to users who have made “visits 2 to 10 days before the last access date and at least 10 times but less than 30 times,” that is, users who have not accessed the site for more than 10 days. Implement measures as appropriate. The content of the access invitation email can be registered when setting the policy, as will be described later.

Further, when the user of the solution proposal device 3 clicks the setting button 57a (FIG. 6: S5; YES), the setting unit 323 performs a policy setting process (FIG. 6: S6) in which "compared to the trend of total PV" A measure is set in which a "sports" article is posted once every three days for users who "access articles a lot", that is, users associated with the selected cell 54.

In addition, the execution unit 324 sends out "sports" articles once every three days to users who access "sports" articles more often than the trend of total PV, that is, users associated with the selected cell 54. Implement measures to post.

An add new measure button 58 is displayed at the bottom of the example of the above measures, and the user of the solution proposal device 3 can add a new measure by clicking the add new measure button 58. can. In the flowchart of FIG. 6, when the new policy addition button 58 is clicked, the setting unit 323 determines that setting of a new policy has been selected (FIG. 6: S7; YES), and as a new policy setting process, the setting unit 323 displays 30 displays a new measure addition screen (FIG. 6: S8). Further, the setting unit 323 performs appropriate processing according to the input during the new policy setting processing (FIG. 6: S8).

FIG. 9 shows an example of the new policy addition screen 60. As shown in FIG. 9, a condition target 61 is displayed in the upper part of the new policy addition screen 60. The condition target 61 displays what kind of user is the target user who will implement the new measure. In the example shown in FIG. 9, it is shown that the target users are those who meet the conditions of "last accessed 2 to 10 days ago and visited 10 or more times."

A policy name 62 is displayed below the condition target 61. In the example shown in FIG. 9, "New Policy" is displayed as the policy name 62. The policy name may be input arbitrarily by the user, or an appropriate name may be automatically input according to the action selected by the user.

Policy settings 63 are displayed below the policy name 62. The policy setting 63 is divided into a plurality of items, and for example, a condition target 63a, a condition 63b, an action target 63c, an action 63d, a processing date 63f, a processing time 63g, and a stop date 63h can be set. The condition target 63a determines whether the policy to be set targets all users who fall under the condition target 61, or whether it targets a specific user among the users who fall under the condition target 61. can be selected.

The condition 63b can be set to cover the entire range of the conditions displayed in the condition target 61 or a part of the range. For the action target 63c, the target for performing the action as a measure can be set in detail based on the relationship between the condition target 61 and the condition 63b. As the action 63d, an action to be performed as a measure can be selected. For the processing date 63f, it is possible to select whether the operation is to be performed every day or on a specific day. For the processing date 63g, the time to perform the operation can be selected. As the stop date 63h, the day on which the operation is to be stopped can be selected.

A policy setting completion button 63i is displayed below the stop date 63h, and when the user clicks the policy setting completion button 63i, the setting unit 322 sets the policy based on the conditions and operations set in the policy setting 63. is set for the corresponding cell 54.

FIG. 10 shows an example of settings on the new policy addition screen 60. In the example shown in FIG. 10, "all users" is selected as the condition target 63a, and that the difference between today's date and the last access date is less than 9 days is selected as the condition 63b. Further, as the operation target 63c, a user who corresponds to the condition target 61 and who also corresponds to the condition 63b is selected.

Sending mail is selected as the action 63d. When e-mail sending is selected as the action 63d, an e-mail input field 63e is displayed as shown in FIG. The mail input field 63e displays a mail title input field 63e1, a mail text input field 63e2, and a mail female view button 63e3. The user can optionally input the title of the email in the email title input field 63e1. The user can optionally input the text of the email in the email text input field 63e2. In this way, when sending an email is selected as a measure, the content of the email to be sent to users who meet the conditions can be set in advance.

After inputting the title of the email in the email title 63e1 and the text of the email in the email text input field 63e2, the user clicks the email preview button 63e3 to display a preview of the email (not shown). As a result, the user can confirm in advance the content, format, etc. of the email to be sent.

In the example shown in FIG. 10, "every day" is selected as the processing date 63f, and "9:00 AM" is selected as the processing time 63g. Furthermore, "undetermined" is selected as the stop date 63h.

After setting the policy as described above, when the user clicks the policy setting completion button 63i, the policy is set for the corresponding cell 54.

Even when a new policy is set as described above, the execution unit 324 appropriately executes the new policy.

The explanation returns to FIG. 8. When the above measures are set, the set measures are executed by the execution unit 324. As a result, as shown by the arrows on the cells 54 with different background colors in FIG. It is expected that the number of sessions will increase to more active users with 30 or more sessions. The effectiveness of the measures can be confirmed by acquiring similar access analysis data after a certain period of time and performing the solution proposal process as described above. As a result, the so-called PDCA cycle can be implemented. Furthermore, the machine learning unit 325 may evaluate the effectiveness of such measures.

As described above, according to the present embodiment, segments divided by two or more axes such as last access date and number of sessions are expressed in the solution proposal matrix, and the characteristics of the segments are posted. Based on this, we will propose an example of measures based on the characteristics of the segment, such as email distribution, push notifications, and site display settings, so that you can easily understand user behavior on services such as providing portal sites. In addition, measures for users can be handled all-in-one using only the solution proposal device 3.

In the example described above, the access status of the membership portal site can be grasped, and based on this access status, measures can be taken for users to increase access.

Examples of measures include not only the above-mentioned email distribution, but also so-called push-type measures using communication means such as the Internet, such as SMS notifications and push notifications to applications. In addition to push-type measures, it is also possible to carry out so-called pull-type measures using communication means such as the Internet, such as so-called A/B tests, such as posting more "sports" articles. good.

Further, by applying filtering such as male/female to the entire population of the matrix, it is possible to target more segments.

(Modified example)
The above embodiments are illustrative, and various modifications can be made without departing from the scope of the invention.

In the embodiment described above, the manner in which each cell 54 in the matrix 50 is selected has been described, but the present invention is not limited to such an aspect. For example, as shown in FIG. 11, by selecting any column representing the number of sessions, all cells 54 corresponding to that column may be selected. In the example shown in FIG. 11, by selecting the column 53a where the number of sessions is "1 time", the number of sessions is "1 time" and the last access date is "0 to 1 day ago" or "2 to 10 days ago". ”, “10 to 30 days ago”, and “30 to 90 days ago” are selected. Similarly, by selecting any row representing the last access date, all cells 54 corresponding to that row may be selected.

The matrix 50 does not need to be limited to a matrix with two axes of Recency (how many days of recent visits) and Frequency (number of visits); it is sufficient that it can perform segmentation based on two variables and implementation of measures for the segments.

For example, measures can be taken using a segment called the Score/Monetary matrix, which is based on the value scored based on behavior to measure the visitor's purchase intention, and the Monetary, which is the total purchase amount including the visitor's past. . Examples of behavioral scoring include:

Click the link in the email +3
Fill out the form +5
Visit to advantageous product information page +3

In the example described above, the provision of a portal site that posts news articles and the like was cited as an example of the service provided by the user of the solution proposal device 3, but the present invention is not limited to such an aspect. For example, it is applicable to all kinds of websites such as shopping sites and curation sites. In the case of a shopping site, the matrix 50 as described above should be created by not only digitizing the number of recent visits or the number of visits, but also digitizing and classifying the cumulative purchase amount. Bye.

Further, the present invention is not limited to websites, but can also be applied to user visit behavior or purchase behavior at physical stores. In this case, the user's actions can be converted into data by, for example, reading a membership card or the like. When promoting users' purchasing behavior at a physical store, it is necessary to convert the number of recent visits, number of visits, cumulative purchase amount, or values scored from these behaviors into data and classify them. , the solution proposal process may be performed in the same manner as described above.

In addition, identification information for identifying a user may be obtained not only by the user's input of membership information or the automatic provision of a user ID, but also by, for example, in the case of a physical store, the identification information may be obtained based on an image taken by a camera device equipped with facial recognition. You can also do this.

In the above-described embodiment, a mode has been described in which a policy is set for the cell 54 in the matrix 50 and the policy, such as email transmission, is executed. However, the present invention is not limited to this embodiment, and the display control section 321 may simply display the matrix 50 on the display section 30. By displaying the matrix 50 on the display unit 30, it is possible to understand at a glance what class the user belongs to, and it becomes possible to take appropriate measures.

Alternatively, the setting section 322 may enable the setting of a measure in a cell 54 in the matrix 50, and by selecting the cell 54, the set measure may be displayed on the display section 30. In this case, it becomes possible to individually implement measures according to the displayed measures.

In the above-described embodiment, a mode in which the matrix 50 is displayed in two dimensions has been described, but the present invention is not limited to such a mode, and the matrix 50 may be displayed in three dimensions.

In the above-described embodiment, an example of a measure is described in which an e-mail is sent to a specific user, but the present invention is not limited to such an example. For example, for new users, measures include changing the screen to be displayed for each user so that a screen for new users is displayed.

In the above-described embodiment, a mode has been described in which user identification information and user behavior are analyzed using an access analysis program, but the present invention is not limited to such a mode. For example, when a user makes a payment with electronic money using a card or smartphone at a physical store, the user's identification information and information about the user's behavior obtained by the payment system are obtained, The user's behavior may also be analyzed. In addition, the identification information that identifies the user is not only based on the image of the camera device with face authentication, but also the identification information that identifies the user based on, for example, the footsteps of the user's walking recorded with a voice recorder. It's okay.

<Industrial applicability>
INDUSTRIAL APPLICATION This invention can be utilized in the user technical field which analyzes the user's purchasing behavior or access on a website, or the user's purchasing behavior at a physical store, and proposes a solution to improve the user's behavior.

1 Terminal device 2 Network 3 Solution proposal device 4 Server 10 Display section 11 Input section 12 Terminal control section 14 Terminal communication section 15 Terminal storage section 30 Display section 31 Input section 32 Control section 33 Communication section 34 Storage section 40 Server communication section 41 Server Control unit 42 Server storage unit 30 Server communication unit 31 Server control unit 32 Server storage unit 100 Solution proposal system 320 Acquisition unit 321 Display control unit 322 Policy presentation unit 323 Setting unit 324 Execution unit

Claims (7)

an acquisition unit that acquires at least identification information that identifies a user and a plurality of elements representing the user's actions;
a plurality of rows that display one element among the plurality of elements in stages; a plurality of columns that display another element among the plurality of elements in stages and intersect with the plurality of rows; Displaying a matrix comprising cells at intersections of the plurality of rows and the plurality of columns, the cells displaying the number of users belonging to each stage of the one element and each stage of the other element. a display control section;
When the cell is selected, at least the identification information and a plurality of elements representing the user's behavior are used as training data to select each of the rows and the columns by machine learning or deep learning. a machine learning unit that proposes an example of a push-type or pull-type optimal measure using the Internet, which is an optimal measure for each segment to which the user belongs, expressed as
a policy presentation unit that displays an example of the proposed policy;
a setting section that can set execution conditions and execution details of the measure;
A solution proposing device comprising: an execution unit that executes, for the user, the measure for which the execution condition and the execution content are set by the setting unit. further comprising a policy presentation unit that displays an example of the policy when the cell is selected;
The solution proposing device according to claim 1, characterized in that: The user's behavior is the user's store visit behavior or purchase behavior at a physical store, and the elements include the number of recent visits, the number of visits, the cumulative purchase amount, or scoring based on these behaviors. Contains the value for which
The acquisition unit acquires at least the identification information and a plurality of elements representing the user's behavior from a facial authentication camera device or a payment system installed in the physical store.
The solution proposing device according to claim 1 or 2, characterized in that: The user's behavior is the user's purchasing behavior on a website or access to the website, and the elements include the number of recent visits, the number of visits, the cumulative purchase amount, or Contains values scored from behavior,
The acquisition unit acquires at least the identification information and a plurality of elements representing the user's behavior based on access analysis software of the website.
The solution proposing device according to claim 1 or 2, characterized in that: The element includes the user's reaction to the implemented measure;
The solution proposing device according to any one of claims 1 to 4, characterized in that: a step of acquiring, by the acquisition unit, identification information for identifying the user and a plurality of elements representing the user's behavior;
The display control unit displays a plurality of rows in which one element among the plurality of elements is displayed in stages, and another element among the plurality of elements is displayed in stages and intersects with the plurality of rows. a plurality of columns; and a cell at an intersection of the plurality of rows and the plurality of columns, the cell displaying the number of users belonging to each stage of the one element and each stage of the other element. displaying the prepared matrix;
When the cell is selected, at least the identification information and a plurality of elements representing the user's behavior are used as training data to select each of the rows and the columns by machine learning or deep learning. a step of proposing an example of a push-type or pull-type optimal measure using the Internet, which is an optimal measure for each segment to which the user belongs, expressed as
Displaying examples of the proposed measures;
a step of setting execution conditions and execution details of the measure;
a step of causing an execution unit to execute the push-type or pull-type measure for the user, in which the execution condition and the execution content are set;
A solution proposal method characterized by: A solution proposal program executable by a computer,
The solution proposal program causes the computer to:
obtaining identification information for identifying a user and a plurality of elements representing the user's behavior;
a plurality of rows that display one element among the plurality of elements in stages; a plurality of columns that display another element among the plurality of elements in stages and intersect with the plurality of rows; Displaying a matrix comprising cells at intersections of the plurality of rows and the plurality of columns, the cells displaying the number of users belonging to each stage of the one element and each stage of the other element. step and
When the cell is selected, at least the identification information and a plurality of elements representing the user's behavior are used as training data to select each of the rows and the columns by machine learning or deep learning. a step of proposing an example of a push-type or pull-type optimal measure using the Internet, which is an optimal measure for each segment to which the user belongs, expressed as
Displaying examples of the proposed measures;
a step of setting execution conditions and execution details of the measure;
executing the push-type or pull-type measure for which the execution condition and the execution content have been set for the user;
A solution proposal program characterized by:

Images