Purchase-related behavior analysis apparatus and method

Abstract

To analyze purchase behavior more accurately while considering congestion in a store.SOLUTION: A purchase-related behavior analysis apparatus 1 configured to analyze the reason for giving up purchase in a store includes: a visit state collection unit 13 which collects visit state indicating behavior of a person in the store; a queue condition estimation unit 14 which estimates queue condition of a queue in the store, based on the visit state; a behavior estimation unit 15 which estimates behavior for the purchase of a user for whom the reason for giving up purchase is to be analyzed, based on the visit state; and a purchase give-up reason determination unit 16 which determines, when the behavior indicates giving up purchase, whether the purchase has been given up due to the queue, based on the queue condition and the behavior.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a technique for analyzing user behavior regarding purchases in a store, and in particular to a technique for analyzing the reasons why a user abandons a purchase. [Background technology]

[0002] In restaurants and other stores, customers may abandon their purchases for various reasons. For example, if the store is crowded due to a long line, customers may go to another store. Also, customers may abandon their purchase if the product they planned to purchase is out of stock or the price is higher than expected.

[0003] For this reason, there is a demand for analyzing purchase abandonment by customers who visit a store or immediately before visiting a store, and for example, Patent Document 1 has proposed such a method. Patent Document 1 addresses the issue that "there is no consideration given to the amount of time customers spend wandering around, and it is not possible to obtain analytical information that allows users to accurately grasp the situation of opportunity loss in a store." To address this issue, Patent Document 1 "measures the wandering time of each customer based on video footage taken inside the store, and associates this wandering time with purchase information for each customer," thereby "obtaining analytical information that allows users to accurately grasp the situation of opportunity loss in a store based on video footage taken inside the store." [Prior art documents] [Patent documents]

[0004] [Patent Document 1] International Publication No. 2016 / 166784 Summary of the Invention [Problem to be solved by the invention]

[0005] As mentioned above, reasons for abandoning a purchase include store conditions such as congestion. In this case, customers may move to another store before entering the store. For example, they may wait in line in front of a store and then leave. In this case, customers abandon their purchase because the store is too crowded and they have to wait longer than expected.

[0006] Therefore, the reason for abandoning a purchase cannot be fully analyzed by simply understanding the behavior in the store. Therefore, the present invention aims to realize a more accurate analysis of purchasing behavior that takes into account the congestion state of the store. [Means for solving the problem]

[0007] To solve the above problem, the present invention uses the queue situation to analyze the reason for abandoning a purchase. That is, it is determined whether the reason for abandoning a purchase is a crowded store by using the queue situation and the behavior of users, including their behavior in the queue and their behavior regarding purchases.

[0008] More specifically, the purchase-related behavior analysis device analyzes reasons for abandoning a purchase in a store, and includes a store visit status collection unit that collects store visit status indicating people's behavior in the store, a queue status estimation unit that estimates the queue status at the store based on the store visit status, a behavior estimation unit that estimates the purchase behavior of a user who is the subject of analysis of reasons for abandoning a purchase based on the store visit status, and a purchase abandonment reason determination unit that, if the behavior indicates abandoning a purchase, determines whether the reason for abandoning the purchase is caused by the queue based on the queue status and the behavior.

[0009] It should be noted that the term "purchase" in the present invention includes not only the purchase of goods but also the receipt of services such as eating and drinking. Therefore, the term "store" includes service-providing establishments such as restaurants.

[0010] The present invention also includes a purchase-related behavior analysis method executed by the purchase-related behavior analysis device, a program for causing the purchase-related behavior analysis device to function as a computer, and a storage medium for storing the program. [Effects of the Invention]

[0011] According to the present invention, it is possible to realize a more accurate analysis of purchase-related behavior that takes into account the crowdedness of a store. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a functional block diagram of a purchase-related behavior analysis device 1 according to an embodiment of the present invention. [Figure 2] 1 is a flowchart showing a processing flow according to an embodiment of the present invention. [Figure 3] 1 is a system configuration diagram of a purchase-related behavior analysis system including a purchase-related behavior analysis device 1 according to a first embodiment. [Figure 4] FIG. 1 is a hardware configuration diagram of a purchase-related behavior analysis device 1 according to a first embodiment. [Figure 5] FIG. 10 is a diagram showing building user information 181 used in the first and second embodiments. [Figure 6] FIG. 10 is a diagram showing permanent store information 182 used in the first and second embodiments. [Figure 7] FIG. 10 is a diagram showing temporary store information 183 used in the first and second embodiments. [Figure 8] FIG. 10 is a diagram showing an individual purchase abandonment determination history 184-1 used in the first embodiment. [Figure 9] FIG. 10 is a diagram showing a total purchase abandonment determination history 184-2 used in the first embodiment. [Figure 10] 1 is a flowchart showing a processing flow in the first embodiment. [Figure 11] 10 is a flowchart showing detailed processing of step S16 (purchase abandonment reason determination processing) in the first embodiment. [Figure 12]FIG. 10 is a system configuration diagram of a purchase-related behavior analysis system including a purchase-related behavior analysis device 1 according to a second embodiment. [Figure 13] FIG. 10 is a hardware configuration diagram of a purchase-related behavior analysis device 1 according to a second embodiment. [Figure 14] FIG. 18 is a diagram showing a defector abandonment decision history 184-3 used in the second embodiment. [Figure 15] 10 is a flowchart showing a processing flow in the second embodiment. [Figure 16] 10 is a flowchart showing a processing flow in a third embodiment. [Figure 17] 11 is a flowchart showing a processing flow in a modified example of the third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0013] An embodiment of the present invention will be described below. First, the subject of this embodiment will be described. In this embodiment, the behavior of users at stores within the influence area of ​​a facility will be analyzed. As an example of the influence area of ​​a facility, an area within a predetermined distance from the building, including the building's premises, will be used. Stores will be temporary stores set up within the influence area or permanent stores such as tenants within the building. Here, "within a predetermined distance" can be expressed in the usual units such as kilometers, or in terms of the number of minutes on foot, or in units of addresses on the same or nearby street as the building, such as the number of blocks. In this embodiment, stores include not only the store itself, but also the surrounding area.

[0014] Permanent stores include various retail businesses such as restaurants, grocery stores, apparel stores, and general stores. Temporary stores include food trucks and food stalls, which are examples of restaurants. The equipment for temporary stores can be provided by either the building or the temporary store itself.

[0015] Note that the same industry includes businesses that are in the same industry, such as restaurants, as well as related industries, such as restaurants and food stores, but is not limited to the same industry in the present invention. Furthermore, buildings can be of any type, such as office buildings, residential buildings, or commercial buildings, as long as they are occupied by tenants. Commercial buildings also include station buildings, department stores, and shopping malls. Furthermore, facilities also include amusement parks such as service areas, roadside stations, amusement parks, and zoos. Furthermore, users who are the subject of analysis in this embodiment include those who wish to purchase or are planning to purchase.

[0016] Next, the configuration of this embodiment will be described. Fig. 1 is a functional block diagram of a purchase-related behavior analysis device 1 in this embodiment. In Fig. 1, the purchase-related behavior analysis device 1 has an input unit 11, an output unit 12, a store visit status collection unit 13, a queue status estimation unit 14, a behavior estimation unit 15, a purchase abandonment reason determination unit 16, a policy creation unit 17, and a memory unit 18.

[0017] First, the input unit 11 is connected to a detection device that detects a store visit status indicating the behavior of multiple users visiting a store. Then, the input unit 11 receives the store visit status from the detection device. Here, the detection device may be a camera that captures images of the inside and outside of the store or a user terminal carried by the user. Alternatively, a sensor device such as a TOF (Time of Flight) sensor may be used. It is desirable that the camera be able to capture images of queues at the store. Alternatively, various computers such as smartphones, tablets, and wearable devices such as smartwatches may be used as user terminals. Note that the functions of the detection device may be provided in the purchase-related behavior analysis device 1.

[0018] Furthermore, the output unit 12 outputs the processing results of the purchase-related behavior analysis device 1. For example, the output unit 12 outputs the reasons why users abandon their purchases and the trends in reasons for abandoning purchases at stores. Furthermore, the output unit 12 may output measures for stores that are created based on the reasons for abandoning their purchases. For this reason, the output unit 12 can be realized by a display device or a communication device.

[0019] The store visit status collection unit 13 also collects the store visit status from the detection device via the input unit 11. The store visit status can be image information of people (especially queues made up of users) including users photographed by a camera or location information of user terminals held by the users. The queue status estimation unit 14 estimates the queue status of a queue in a store made up of multiple users based on the collected store visit status. The queue status includes the presence or absence of a queue, and if a queue has been formed, the duration and length (number of people) of the queue, and the congestion status.

[0020] The behavior estimation unit 15 also estimates the behavior of users based on their store visit status and queue status. This behavior includes purchase-related behavior, which indicates the behavior of users in queues and their behavior regarding purchases. Here, the behavior of users in queues includes behavior regarding queues, such as whether the user joined (formed) a queue, whether they approached the end of the queue, or whether they left the queue. The behavior of users regarding purchases includes behavior indicating whether they made a purchase by moving to a cash register or the like. Movement to a cash register or the like includes movement within the store, such as to the ticket vending machine, counter, entrance, or seat.

[0021] In this way, purchase-related behavior includes the movement of users in the analysis area related to the store. More preferably, the analysis area is an area including a queue. In this way, a queue is formed within the analysis area. For this reason, it can be determined according to location conditions including at least one of the store's location, size, and direction (facing), as represented by the area around the store entrance. Note that the location can be determined by the store address, the area of ​​the facility such as a building, or latitude and longitude information. The size can be determined by the store's area, the number of people it can accommodate, such as the number of seats, etc. The direction (facing) can be determined by the direction in which the store entrance or reception counter faces, or the corridors, roads, elevators, and other facilities to which they face.

[0022] Furthermore, as a purchase-related behavior, a purchase procedure such as the use of electronic money may be used in addition to travel. As a result, whether or not a purchase was made (whether or not the purchase was abandoned) is determined as the user's behavior regarding purchases.

[0023] Furthermore, the purchase abandonment reason determination unit 16 determines the reason for abandoning a purchase based on the queue situation and the behavior, if the behavior indicates abandoning a purchase. More preferably, the purchase abandonment reason determination unit 16 determines whether the queue is the cause. Note that the behavior estimation unit 15 and the purchase abandonment reason determination unit 16 preferably compile the estimation of behavior and the reasons for abandoning a purchase.

[0024] Furthermore, the policy creation unit 17 creates policy information including improvement proposals for purchase abandonment based on the behavior estimation unit 15 and the purchase abandonment reason determination unit 16's behavior estimation and purchase abandonment reason estimation. For example, the policy creation unit 17 creates policy information based on the purchase abandonment determination history 184 described below. Here, this policy information can use recommendation information, which is an improvement proposal for a store, or usage suggestion information, which suggests a store other than the one the user abandoned the purchase at. For this reason, the policy creation unit 17 can be configured with a recommendation information creation unit and a usage suggestion information creation unit.

[0025] The storage unit 18 also stores building user information 181, permanent store information 182, temporary store information 183, purchase abandonment determination history 184, policy information 185, and store visit status information 186. Details of each piece of information will be explained in each embodiment described later, so an overview of each piece of information will be given here.

[0026] First, building user information 181 is information about building users who may become store users. Note that in this embodiment, information about "building" users is handled, but the present invention is not limited to this. For example, it may be information about members of each target store, people who have used the store, or members of a store information service.

[0027] The permanent store information 182 and the temporary store information 183 are stores that are the subject of this embodiment, and are information about the permanent stores in the building described above and the temporary stores that are set up within the affected area. Note that only one of the permanent store information 182 and the temporary store information 183 may be used.

[0028] Furthermore, the purchase abandonment determination history 184 is information indicating the purchase abandonment reason determined by the purchase abandonment reason determination unit 16. Furthermore, the measure information 185 is information indicating the measure created by the measure creation unit 17. Furthermore, the store visit status information 186 is information indicating the store visit status collected by the store visit status collection unit 13. Therefore, the store visit status information 186 can be realized by image information of users (especially, a queue made up of users) photographed by a camera or location information of the user terminals held by the users. This concludes the explanation of the purchase-related behavior analysis device 1 in this embodiment, but the purchase-related behavior analysis device 1 can be realized by various types of computers such as a server or a PC.

[0029] Next, the processing flow of this embodiment will be described. Fig. 2 is a flowchart showing the processing flow of this embodiment. This flowchart is based on the premise that the store visit status collection unit 13 collects store visit status from the detection device via the input unit 11.

[0030] 2, the queue situation estimation unit 14 estimates the queue situation in a store consisting of multiple customers based on the collected store visit status. Also, in step S2, the behavior estimation unit 15 estimates the behavior of users based on the store visit status and the queue situation.

[0031] Then, in step S3, the purchase abandonment reason determination unit 16 determines the reason for abandoning the purchase based on the queue situation and behavior. As a result, if the reason for abandoning the purchase is due to queuing, the process proceeds to step S4. If the reason is other than this (other causes), the process proceeds to step S5. Note that the reason for abandoning the purchase due to queuing includes not only cases where queuing is a direct reason for abandoning the purchase, but also cases where it takes a long time to receive service, such as when queuing is a cause of store congestion or a delay in service provision due to a malfunction of a cash register, which are factors that cause queuing. Note that in step S3, it is desirable that the purchase abandonment reason determination unit 16 determines whether purchase abandonment has occurred based on behavior. However, the determination of purchase abandonment may be performed in another step or by a configuration such as the behavior estimation unit 15.

[0032] As a result, in step S4, the output unit 12 outputs that the reason for abandoning the purchase is due to queues. Furthermore, in step S5, the output unit 12 outputs that the reason for abandoning the purchase is not due to queues. As a result, the system user of the purchase-related behavior analysis device 1 can understand the reason for the user's abandonment of the purchase. Furthermore, in this embodiment, the measure creation unit 17 may create a measure including an improvement plan for the store regarding abandoning the purchase, depending on the determination result of step S3. Furthermore, although the determination result of step S3 is output in the flowchart shown in FIG. 2, the determination result may simply be stored in the storage unit 18 as purchase abandonment determination history 184. This concludes the description of this embodiment, and below, specific examples will be described. [Example]

[0033] In Example 1, analysis subjects are extracted from a plurality of users, purchase-related behavior is determined for each subject, and reasons for abandoning a purchase are determined. Details thereof are explained below. FIG. 3 is a system configuration diagram of a purchase-related behavior analysis system including a purchase-related behavior analysis device 1 in Example 1. In the purchase-related behavior analysis system, the purchase-related behavior analysis device 1, building management systems 2a and 2b, a building management database 3, a temporary store database 4, a permanent store system 5, and a temporary store system 6 are connected to one another via a network 8. A user terminal 7 used by a user is further connected to the network 8.

[0034] First, the purchase-related behavior analysis device 1 is a device that executes the main processing of Example 1. The purchase-related behavior analysis device 1 has an input unit 11 to a storage unit 18, similar to those in Fig. 1. The following description will focus on the differences from Fig. 1.

[0035] The input unit 11 and the output unit 12 are connected to the network 8. The input unit 11 receives customer status from the cameras 23a to 23c and the user terminal 7. The input unit 11 may also acquire purchase information from the permanent store register 52 or the temporary store register 62 as the customer status.

[0036] Furthermore, the output unit 12 outputs the processing results of the purchase-related behavior analysis device 1 to various terminals via the network 8. For this reason, the input unit 11 and the output unit 12 can be realized by a communication device having a communication function via the network 8.

[0037] 1, the store visit status collection unit 13 collects the store visit status via the input unit 11. The store visit status collection unit 13 stores the collected store visit status in the storage unit 18 as store visit status information 186.

[0038] Furthermore, the queue situation estimation unit 14 extracts the visitation status of the target store in the area to be analyzed from the visitation status information 186, and estimates the queue situation in this area to be analyzed.

[0039] The behavior estimation unit 15 also includes a target person extraction unit 151 and a purchase-related behavior determination unit 152. First, the target person extraction unit 151 extracts a target person who is a user whose purchase-related behavior is to be determined from the extracted store visit state. The purchase-related behavior determination unit 152 estimates the purchase-related behavior of the extracted target person based on the extracted store visit state and the estimated queue state. As a result, the behavior estimation unit 15 can estimate the behavior of the target person including the estimated purchase-related behavior. In the first embodiment, the purchase-related behavior is the behavior itself.

[0040] The purchase abandonment reason determination unit 16 also includes an individual reason determination unit 161 and a reason aggregation unit 162. First, the individual reason determination unit 161 determines the subject's reason for abandoning the purchase based on the estimated queue situation and estimated behavior. At this time, the individual reason determination unit 161 determines whether the subject made a purchase or abandoned the purchase using the estimated purchase-related behavior of the subject. Then, the individual reason determination unit 161 stores the determination result in the storage unit 18 as an individual purchase abandonment determination history 184-1.

[0041] Furthermore, the reason tallying unit 162 tally the determination results of the reason for abandoning purchase for multiple subjects. Specifically, the reason tallying unit 162 tally at least the number of subjects whose reason for abandoning purchase was due to queuing. Here, the number of subjects is preferably the total number of people, but if a predetermined condition is met, such as the same subject abandoning a purchase multiple times within a certain period, the number may be counted as one person (one time). More preferably, the reason tallying unit 162 tally subjects whose reason for abandoning purchase was not due to queuing. Furthermore, the reason tallying unit 162 may count subjects who made a purchase. Then, the reason tallying unit 162 stores the tallying result in the storage unit 18 as the tally purchase abandonment determination history 184-2.

[0042] As a result, the purchase abandonment reason determination unit 16 can determine the reason for abandoning a purchase for the target store. In other words, if a predetermined condition is met, such as a large number of subjects who abandoned a purchase due to queuing, it can be determined that the main reason for abandoning a purchase is queuing. Here, "large" means that the number is equal to or exceeds a predetermined threshold, and the ratio of subjects who abandoned a purchase due to reasons other than queuing or who made a purchase is equal to or exceeds a predetermined threshold.

[0043] The policy creation unit 17 also creates policy information 185 based on an individual purchase abandonment determination history 184-1 and an aggregate purchase abandonment determination history 184-2, which are specific examples of the purchase abandonment determination history 184. The storage unit 18 also stores building user information 181, permanent store information 182, non-permanent store information 183, an individual purchase abandonment determination history 184-1, an aggregate purchase abandonment determination history 184-2, the policy information 185, and store visit status information 186. Here, the individual purchase abandonment determination history 184-1 and the aggregate purchase abandonment determination history 184-2 are specific examples of the purchase abandonment determination history 184. These will be described in detail later.

[0044] Next, the building management systems 2a and 2b are used to manage buildings and are utilized by the owners and managers of the buildings. In the first embodiment, the purchase-related behavior analysis device 1 is realized in a so-called cloud environment used to manage stores in multiple buildings, and therefore multiple building management systems are illustrated. However, this may be a single system, or the functions of the purchase-related behavior analysis device 1 may be provided in the building management system.

[0045] First, the building management system 2a can be realized by a computer system used to manage building A. The building management system 2a is composed of a building management terminal 21a, a building management server 22a, and a camera 23a. The building management terminal 21a notifies the purchase-related behavior analysis device 1 of a request to analyze reasons for abandoning a purchase, and displays the processing results on the purchase-related behavior analysis device 1. The building management system 2a may be used to manage multiple buildings. The camera 23a is an example of a detection device, and captures the analysis target areas of permanent stores and temporary stores in building A. Although FIG. 3 shows one building management terminal 21a and one camera 23a, multiple units may be provided. In particular, a camera 23a may be provided for each permanent store and temporary store.

[0046] Furthermore, building management system 2b is used to manage building B, and is shown as an example configured with building management terminal 21b and camera 23b. Building management terminal 21b is also used to manage the building, and has the same functions as building management terminal 21b. Camera 23b is also an example of a detection device, and takes images of the analysis target area of ​​permanent stores and temporary stores in building B. Note that building management system 2b can also be used to manage multiple buildings.

[0047] The detection device may be provided independently of the building management system 2a or the building management system 2b, as an example shown in Fig. 3 as a camera 23c.

[0048] Here, cameras 23a to 23c will be described. These cameras are installed so as to capture the area of ​​the store to be analyzed. For this reason, cameras 23a to 23c are installed according to the location conditions of the store. In other words, cameras 23a to 23c are installed at positions and orientations according to the location conditions of the store so that the angle of view of each camera covers the area to be analyzed.

[0049] Furthermore, the building management database 3 is a database that stores building A resident user information 31, building A resident permanent store information 32-1, and building B resident permanent store information 32-2. The building A resident user information 31 is the same information as the building user information 181. Therefore, either one of these may be omitted. The building A resident permanent store information 32-1 and building B resident permanent store information 32-2 are each the same information as the permanent store information 182. Note that at least a portion of this information may be stored in the purchase-related behavior analysis device 1 or the building management system 2a or 2b. Conversely, at least a portion of the individual purchase abandonment determination history 184-1 to store visit status information 186 may be stored in the building management database 3.

[0050] The temporary store database 4 is a database that stores temporary store information 41 related to temporary stores. The temporary store information 41 is the same information as the temporary store information 183. Therefore, either one of them may be omitted. The temporary store information 41 may be stored in the purchase-related behavior analysis device 1, the building management systems 2a and 2b, or the building management database 3. Conversely, the temporary store database 4 may store at least a portion of the individual purchase abandonment determination history 184-1 to the store visit status information 186.

[0051] The permanent store system 5 is comprised of a permanent store terminal 51 and a permanent store register 52. The permanent store terminal 51 is a device used by those involved and managers of the permanent store, and can be implemented by a computer. The permanent store terminal 51 accepts and displays policy information and other information related to the permanent store itself. The permanent store register 52 has a payment function for purchases at the permanent store, and can be implemented by a so-called POS register.

[0052] The temporary store system 6 is composed of a temporary store terminal 61 and a temporary store register 62. The temporary store terminal 61 is a device used by those involved and the manager of the temporary store, and can be realized by a computer. The temporary store terminal 61 receives and displays policy information related to the temporary store. The temporary store register 62 has a payment function for purchases at the permanent store, and can be realized by a so-called POS register.

[0053] The permanent store system 5 and the temporary store system 6 and the devices that make them up can be omitted. Furthermore, the permanent store register 52 and the temporary store register 62 may be devices independent of the permanent store system 5 and the temporary store system 6.

[0054] The user terminal 7 is an information terminal held by the user, and can be realized by various computers, such as a smartphone, tablet, or wearable device like a smartwatch. The user terminal 7 has a location detection function using GPS or the like, and notifies the purchase-related behavior analysis device 1 of location information indicating the detected location. The user terminal 7 may also have a payment function realized by a two-dimensional code, electronic money, or the like. This payment function may also be realized by an IC card separate from the user terminal 7.

[0055] The network 8 can be realized by a wide area network such as the Internet, and connects the above-mentioned devices together. The building management terminal 21a, the building management terminal 21b, the permanent store terminal 51, and the temporary store terminal 61 can be realized by a PC, a smartphone, a tablet, or the like. This concludes the explanation of Figure 3. Next, an implementation example of the purchase-related behavior analysis device 1 that performs the main processing of Example 1 will be explained.

[0056] Fig. 4 is a hardware configuration diagram of the purchase-related behavior analysis device 1 in Example 1. In Example 1, the purchase-related behavior analysis device 1 can be realized by a server, which is an example of a computer, in particular, by a cloud. As shown in Fig. 4, the purchase-related behavior analysis device 1 has a processing device 101, a communication device 102, a memory 103, and a secondary storage device 104, which are connected to each other via a communication path.

[0057] First, the processing device 101 can be realized by a processor such as a CPU, and executes calculations in accordance with a purchase-related behavior analysis program 105 stored in a secondary storage device 104, which will be described later. The purchase-related behavior analysis program 105 will be described later. The communication device 102 corresponds to the input unit 11 and output unit 12 in FIG. 3, and is connected to the network 8 to communicate with other devices. Note that, as components corresponding to the input unit 11 and output unit 12, the purchase-related behavior analysis device 1 may be provided with an input device that can be realized by a keyboard or mouse that a system user operates to input data, or a display device that can be realized by a display monitor.

[0058] 3. The memory 103 and the secondary storage device 104 store the purchase-related behavior analysis program 105 and information used for processing by the processing device 101, both of which are stored in the secondary storage device 104. The secondary storage device 104 can be implemented as a so-called storage. The secondary storage device 104 stores the purchase-related behavior analysis program 105, building user information 181, permanent store information 182, temporary store information 183, individual purchase abandonment determination history 184-1, aggregated purchase abandonment determination history 184-2, policy information 185, and store visit status information 186. These will be described later. The secondary storage device 104 may be implemented as various storage media such as an external hard disk drive (HDD), solid state drive (SSD), or memory card, or may be implemented as a device separate from the purchase-related behavior analysis device 1, such as a database system or file server.

[0059] Here, the purchase-related behavior analysis program 105 is composed of a store visit status collection module 106, a queue status estimation module 107, a target person extraction module 108, a purchase-related behavior determination module 109, an individual reason determination module 110, a reason aggregation module 111, and a policy creation module 112. Note that each of these modules may be realized as an individual program or a partial combination.

[0060] The configuration shown in FIG. 3, which performs the same functions as each module, is as follows: Visit status collection module 106: Visit status collection unit 13 Queue situation estimation module 107: Queue situation estimation unit 14 Subject extraction module 108: Subject extraction unit 151 Purchase-related behavior determination module 109: Purchase-related behavior determination unit 152 Individual reason determination module 110: Individual reason determination unit 161 Reason counting module 111: Reason counting unit 162 Policy creation module 112: Policy creation section 17 Therefore, the processing device 101 executes the processes of the store visit status collection unit 13, the queue status estimation unit 14, the target person extraction unit 151, the purchase-related behavior determination unit 152, the individual reason determination unit 161, the reason aggregation unit 162, and the policy creation unit 17 in accordance with the purchase-related behavior analysis program 105. Note that each of these modules may be configured as an individual program. This concludes the description of the configuration of the first embodiment, and next, the information used in the first embodiment will be described.

[0061] Next, the information used in the first embodiment will be described. First, FIG. 5 is a diagram showing building user information 181 used in the first embodiment. The building user information 181 is information about building users, such as employees of buildings related to the permanent store or temporary store that is the subject of analysis. Here, the building users are candidates for users whose reasons for abandoning a purchase are to be determined. Therefore, the building user information 181 has the following items for each building user: ID, working tenant, lunch break time, last restaurant history, and attributes. First, the ID is an item that identifies the building user. The working tenant is an item that indicates the tenant where the building user works. Note that the working tenant may be any tenant, location, etc. related to the building user.

[0062] The lunch break time is an item indicating the lunch break time of the building user in question. In Example 1, the target store is a restaurant, so the lunch break time is used. That is, in Example 1, the target store is a restaurant, so the lunch break time, which is the time when there is a high possibility of the store being used, is used. Therefore, a break time other than the lunch break time may be used, or in the case of a store of another type, a different time may be used depending on the possibility of the store being used.

[0063] The last restaurant history is an item that indicates the history of the building user's last visit to a store, such as a temporary store, and records the store used and the time. The attributes indicate the attributes of the building user. In the example of Figure 5, address, gender, age, and preferences are used. Here, the attributes, especially preferences, are related to the target restaurant. For this reason, when targeting stores other than restaurants, it is desirable that the attributes, especially preferences, correspond to the type of business of the target store. The preferences indicate the preferences of the building user.

[0064] 6 is a diagram showing permanent store information 182 used in Example 1. The permanent store information 182 is information about permanent stores. Therefore, the permanent store information 182 has the following items: store name, genre, sales, number of customers, number of people that can be accommodated, offered menu 1, offered menu 2, and location conditions.

[0065] First, the store name indicates the name of the permanent store, but any item that can identify the permanent store will suffice. The genre is an item that indicates the genre of the target permanent store. The sales indicates the sales of the relevant permanent store. Note that a profit margin item may be added to the sales.

[0066] The number of customers indicates the number of customers using the corresponding permanent store. For users, a user information item (e.g., membership information) for each user may be provided. The number of customers that can be accommodated indicates the number of customers that can be accommodated at the corresponding permanent store. More preferably, the number of customers that can be accommodated per unit period (e.g., unit time) is used. Here, it is desirable to use data for the same period for sales, number of customers, and number of customers that can be accommodated. Furthermore, representative values ​​such as the average or median for each unit period can be used for sales and number of customers. Furthermore, the number of customers that can be accommodated may be determined based on the number of seats at the corresponding permanent store. In Example 1, the location of the permanent store can be obtained from the permanent store information 32-1 in Building A.

[0067] The offered menu 1 and offered menu 2 are items that indicate menus that are examples of products and services provided at the permanent store. In addition to these, the permanent store information 32-1 in building A may also include a representative menu.

[0068] The location conditions indicate the location conditions of the permanent store, and in the example of Fig. 6, the location, size (area), and direction (entrance) are used. The location conditions can be used to identify the angle of view of the cameras 23a to 23c and to create the policy information 185 that proposes improvements to the store.

[0069] Next, FIG. 7 is a diagram showing temporary store information 183 used in Example 1. The temporary store information 183 is information about temporary stores. Therefore, the temporary store information 183 has the following items: store name, genre, sales, number of customers, menu items 1 and 2, store opening period, and location conditions. First, the store name indicates the name of a temporary store that has already opened or is a potential store opening candidate, but any item that can identify the temporary store will suffice. Furthermore, the items from genre to menu items 2 are the same as those in the permanent store information 182 for the corresponding temporary store.

[0070] The store opening period indicates the period during which the temporary store will be open in Building A. Similarly to the permanent store information 182, the location conditions can be used to identify the angle of view of the cameras 23a to 23c and to create the policy information 185 that proposes improvements to the store.

[0071] Next, Fig. 8 is a diagram showing an individual purchase abandonment determination history 184-1 used in the first embodiment. Fig. 9 is a diagram showing an aggregate purchase abandonment determination history 184-2 used in the first embodiment. These are examples of the purchase abandonment determination history 184 in Fig. 1. First, the individual purchase abandonment determination history 184-1 is information showing a history of the reasons for abandoning a purchase each time a target user abandons a purchase. Therefore, the individual purchase abandonment determination history 184-1 shows each item such as the store name, the target user, the reason for abandoning a purchase, and the date and time.

[0072] The store name is an item indicating the store where the user abandoned the purchase. The target person is an item indicating the user who abandoned the purchase. The reason for abandoning the purchase is an item indicating the reason why the target person abandoned the purchase, and is either a queue or something other than a queue. The date and time is an item indicating the date and time when the purchase was abandoned. The date and time when the purchase was abandoned can be the actual date and time when the purchase-related behavior analysis device 1 determined that the purchase was abandoned.

[0073] Furthermore, the aggregated purchase abandonment determination history 184-2 is information indicating the results of aggregating the individual purchase abandonment determination history 184-1 for each store. For this purpose, the aggregated purchase abandonment determination history 184-2 has the following fields: store name, queue cause (aggregate), and cause. First, the store name indicates the target store. Furthermore, the queue cause (aggregate) indicates the result of aggregating the number of purchase abandonment reasons in the individual purchase abandonment determination history 184-1. Furthermore, the cause indicates the cause of purchase abandonment at the target store (store name) determined according to the queue cause (aggregate). For example, if the queue cause (aggregate) is equal to or greater than a threshold, it is determined to be a queue. Furthermore, the cause may be identified according to the queue cause (aggregate) and the ratio of the number of customers or the number of people that can be accommodated at the relevant store.

[0074] The policy information 185 is created based on at least the aggregated purchase abandonment determination history 184-2, and is information including measures for improving the store, particularly for abandoning purchases. The policy information 185 can be realized as recommendation information for changing the number of stores (number or genre) of stores or usage suggestion information for suggesting stores to users for purchases.

[0075] Furthermore, the store visit status information 186 can use image information of users (especially queues of users) captured by the cameras 23a to 23c and location information from the user terminal 7. It is desirable that these include a time element. As a result, it is possible to grasp the congestion status, including queues, at the store.

[0076] This concludes the explanation of the information used in Example 1, and next, the processing flow of Example 1 will be explained. Fig. 10 is a flowchart showing the processing flow in Example 1. The following explanation will be given along this flowchart, but the processing will mainly use the configuration in Fig. 3. In addition, the following explanation will be given using an example of opening a temporary store in Building A.

[0077] First, in step S11, the store visit status collection unit 13 collects store visit statuses as appropriate via the input unit 11 and stores them in the storage unit 18 as store visit status information 186. Then, in step S12, the queue situation estimation unit 14 extracts the store visit status of the analysis target area to be analyzed from the store visit status information 186. Here, the processing from step S12 onwards may be executed periodically, or may be executed in response to a request from any of the building management terminal 21a, building management terminal 21b, permanent store terminal 51, or temporary store terminal 61. However, more preferably, in addition to the building management terminal 21a related to building A, the permanent store terminal 51 or temporary store terminal 61 that is located in building A is also executed.

[0078] In this case, each of these terminals will notify the purchase-related behavior analysis device 1 of a request in response to an input from the system user. Furthermore, the processes from step S12 onwards may be executed in response to a request from the user terminal 7. In step S11, the store visit status collection unit 13 extracts, as the store visit status, the store visit status for a predetermined period (for example, one hour) or a period designated by the system user.

[0079] In step S13, the queue situation estimation unit 14 estimates the queue situation in the area to be analyzed using the extracted store visit situation.

[0080] Here, the queue status refers to the status of the queue, and can be, for example, whether or not a queue has occurred, the time the queue has been formed, the length of the queue (number of people), or the congestion status. First, the presence or absence of a queue can be determined to have occurred when there are more than a predetermined threshold number of people waiting in line. Furthermore, the congestion status indicates the speed at which people in the queue are moving.

[0081] Furthermore, queue conditions can be estimated by performing image processing on image data or by measuring the location information (within the area to be analyzed) of the user terminal 7. In this way, the store visit status is what is known as measurement data, and queue conditions are an example of people flow data created from this measurement data.

[0082] Furthermore, in steps S14 and S15, the behavior estimation unit 15 estimates the user's behavior based on the extracted store visit state and the created queue situation. To this end, in step S14, the target person extraction unit 151 of the behavior estimation unit 15 extracts a target person from the extracted store visit state. For example, the target person is the user who is detected as the earliest store visit state. In this case, the target person extraction unit 151 extracts, as the target person, users who are waiting in line or who are about to wait in line in the queue indicated by the queue situation estimated by the queue situation estimation unit 14. To this end, the target person extraction unit 151 performs image processing or uses time information included in the location information of the user terminal 7.

[0083] In step S15, the purchase-related behavior determination unit 152 of the behavior estimation unit 15 estimates the purchase-related behavior of the extracted subject based on the extracted store visit state and the estimated queue state.

[0084] Here, purchase-related behavior refers to the behavior of the subject, including behavior in a queue and behavior related to a purchase. Furthermore, behavior in a queue includes behavior related to the queue, such as whether the user joined (or formed) a queue, whether they approached the end of the queue or were absorbed into it, or whether they left the queue.

[0085] Furthermore, the entrance behavior includes behavior indicating whether a purchase was made by moving to a cash register or the like. Movement to a cash register or the like includes movement within the store, such as to the ticket vending machine, counter, window, or seat. Furthermore, purchase-related behavior is determined by performing image processing on image data of the subject's visit to the store or by measuring the location information (within the area to be analyzed) of the user terminal 7. In this way, the behavior estimation unit 15 estimates movement within the area to be analyzed as the subject's behavior.

[0086] It should be noted that AI technology can be used in the image processing in steps S13 to S15.

[0087] Furthermore, if the behavior indicates abandonment of a purchase based on the estimated queue situation and the estimated behavior, the purchase abandonment reason determination unit 16 determines the reason for abandoning the purchase. The determination of the reason for abandoning a purchase in the first embodiment includes determination of the abandonment reason of an individual subject and the abandonment reason of multiple subjects. For this reason, in the first embodiment, the purchase abandonment reason determination unit 16 is provided with an individual reason determination unit 161 and a reason counting unit 162, which respectively execute step S16 and step S17.

[0088] First, in step S16, the individual reason determination unit 161 determines the reason for abandoning purchase of the target person based on the estimated queue situation and estimated behavior, and creates the individual purchase abandonment determination history 184-1. Here, an example of step S16 (purchase abandonment reason determination process) will be described with reference to FIG.

[0089] 11 is a flowchart showing detailed processing of step S16 (purchase abandonment reason determination processing) in Example 1. First, in step S161, the individual reason determination unit 161 determines whether the extracted subject has reached the cash register of the relevant store. To this end, the individual reason determination unit 161 uses the purchase-related behavior determined by the purchase-related behavior determination unit 152. As a result, if it is determined that the subject has reached the cash register (Y), the process proceeds to step S166. Then, in step S166, the individual reason determination unit 161 determines that the subject has made a purchase.

[0090] Furthermore, if it is determined in step S161 that the person has not reached the cash register (N), the process proceeds to step S162. That is, it is determined that the subject has abandoned the purchase. Furthermore, in step S162, the individual reason determination unit 161 determines whether a queue was formed during the extracted period of visit to the store, based on the presence or absence of a queue in the estimated queue situation. As a result, if a queue was formed (Y), the process proceeds to step S163. Furthermore, if a queue was not formed (N), the process proceeds to step S167. Then, in step S167, the individual reason determination unit 161 determines that the reason for the subject's abandoning the purchase was not due to the queue.

[0091] In step S163, the individual reason determination unit 161 determines whether the extracted subject person lined up in the formed queue. To this end, the individual reason determination unit 161 uses the purchase-related behavior determined by the purchase-related behavior determination unit 152. As a result, if the subject has joined a queue (Y), the process proceeds to step S164. If the subject has not joined a queue (N), the process proceeds to step S167. Then, in step S167, the individual reason determination unit 161 determines that the subject's reason for abandoning the purchase is not due to waiting in line. In step S163, if the purchase-related behavior is near the end of the queue or is absorbed into the queue, it is determined that the subject has joined a queue.

[0092] Furthermore, in step S164, the individual reason determination unit 161 determines whether the queue length in the estimated queue situation is equal to or greater than a predetermined value. The queue length can be determined by the number of users in the queue or the distance from the front to the back. Furthermore, in step S164, it may be determined whether the queue congestion situation in the queue situation is equal to or greater than a predetermined value. In this way, in step S164, the congestion situation due to the queue is determined.

[0093] As a result of the above, if the length of the queue is equal to or greater than a predetermined value (Y), the process proceeds to step S165. On the other hand, if the length of the queue is less than the predetermined value (N), the process proceeds to step S167. Then, in step S165, the individual reason determination unit 161 determines that the reason for the subject's abandonment of purchase is due to the queue. By the above steps S161 to S167, an individual purchase abandonment determination history 184-1 is created. Then, the individual reason determination unit 161 stores the individual purchase abandonment determination history 184-1 in the storage unit 18.

[0094] It is desirable to repeatedly execute steps S14 to S16 for each queue member and user in the area to be analyzed, but they may also be executed for users sampled from the queue members and users in the area to be analyzed.

[0095] 11 is now completed, and we will return to FIG. 10 to explain step S17 and beyond. In step S17, the reason tallying unit 162 tally up the judgment results of the purchase abandonment reasons for multiple subjects. That is, the reason tallying unit 162 tally up the purchase abandonment reasons in the individual purchase abandonment judgment history 184-1 for each store, and create an aggregated purchase abandonment judgment history 184-2.

[0096] In step S18, the policy creation unit 17 uses at least one of the individual purchase abandonment determination history 184-1 and the aggregate purchase abandonment determination history 184-2 to create policy information 185 including a plan to improve congestion represented by a queue. Then, the policy creation unit 17 stores the policy information 185 in the storage unit 18.

[0097] The policy information 185 includes recommendation information for stores and usage suggestion information for users. Here, the recommendation information can be, for example, information indicating a change in the number or type of stores to be opened. Here, if the target store is a temporary store, the policy creation unit 17 creates recommendation information to change at least one of the number and type of temporary stores to be opened or currently opened in the affected area.

[0098] In this case, the policy creation unit 17 may create such recommendation information when at least one of the individual purchase abandonment determination history 184-1 and the aggregate purchase abandonment determination history 184-2 satisfies a predetermined condition. As this predetermined condition, for example, the number of temporary stores due to queues or the ratio of permanent stores due to queues can be used.

[0099] That is, in the first embodiment, it is desirable that the campaign creation unit 17 creates the campaign information 185 when at least one of the individual purchase abandonment determination history 184-1 and the aggregate purchase abandonment determination history 184-2 satisfies a predetermined condition.

[0100] For example, if there is a store in which the reason for abandoning purchase due to queues has occurred a predetermined number of times or more within a certain period in the individual purchase abandonment determination history 184-1 or the aggregate purchase abandonment determination history 184-2, the policy creation unit 17 creates recommended information for the store. Here, the recommended information for the store includes measures to reduce queues, such as changing the location of the store or the position of queues, as well as canceling the store and changing business hours or menus.

[0101] As another example, if there is a target person in the individual purchase abandonment judgment history 184-1 who has abandoned a purchase a predetermined number of times or more within a certain period due to queues, the policy creation unit 17 creates usage suggestion information for the target person. This usage suggestion information includes suggesting stores other than the one where the target person abandoned the purchase.

[0102] Furthermore, for example, information suggesting a store where purchases can be made can be used as the usage suggestion information. Here, if the target store is a temporary store, the policy creation unit 17 creates usage suggestion information that suggests using a permanent store in the relevant building A as the purchase store. In this case, it is desirable for the policy creation unit 17 to identify a permanent store that has no cause or is not a store with a queue in the aggregated purchase abandonment determination history 184-2.

[0103] Furthermore, the policy creation unit 17 transmits the created policy information to related devices via the output unit 12. These related devices include the building management terminal 21a, the building management terminal 21b, the permanent store terminal 51, the temporary store terminal 61, and the user terminal 7. In particular, it is desirable that the recommendation information be transmitted to the building management terminal 21a, the permanent store terminal 51, and the temporary store terminal 61 related to Building A. It is also desirable that the usage suggestion information be transmitted to the user terminal 7.

[0104] This concludes the description of the first embodiment. In the first embodiment, the reason for abandoning purchase is determined for each target user, so that a more precise analysis is possible. [Example]

[0105] In Example 2, people who leave the queue are identified, and these people are treated as subjects to determine the reason for abandoning the purchase. Example 2 will be described below, focusing on the differences from Example 1. First, FIG. 12 is a system configuration diagram of a purchase-related behavior analysis system including a purchase-related behavior analysis device 1 in Example 2. In FIG. 12, devices other than the purchase-related behavior analysis device 1 are indicated by the same characters as in FIG. 3, and therefore descriptions thereof will be omitted.

[0106] 12 also includes an input unit 11, an output unit 12, a store visit status collection unit 13, a queue status estimation unit 14, a behavior estimation unit 15, a purchase abandonment reason determination unit 16, a policy creation unit 17, and a storage unit 18. Here, the input unit 11 to the queue status estimation unit 14 have the same functions as those in the first embodiment, and therefore their description will be omitted.

[0107] The behavior estimation unit 15 also has a queue extraction unit 153, a leaver determination unit 154, and a leaver counting unit 155. First, the queue extraction unit 153 extracts a queue to be analyzed from the store visit state information 186. The leaver determination unit 154 determines leavers from the extracted queue. Then, the leaver counting unit 155 counts the number of determined leavers. As a result, the behavior estimation unit 15 determines leaving a queue as purchase-related behavior and counts the number of leavers as an estimation of behavior.

[0108] Furthermore, the purchase abandonment reason determination unit 16 determines the reason for abandonment based on the number of leavers. At this time, it is desirable that the purchase abandonment reason determination unit 16 also use the queue situation. For example, the purchase abandonment reason determination unit 16 determines the reason for abandonment using, for example, the length of the queue and the ratio of the number of leavers. Then, the behavior estimation unit 15 and the purchase abandonment reason determination unit 16 store the processing results in the storage unit 18 as a leaver purchase abandonment determination history 184-3.

[0109] Furthermore, the policy creation unit 17 creates policy information 185 using the leaver abandonment determination history 184-3 as the reason for abandonment. Furthermore, the storage unit 18 stores building user information 181, permanent store information 182, temporary store information 183, leaver abandonment determination history 184-3, policy information 185, and store visit status information 186. Here, the leaver abandonment determination history 184-3 is a specific example of the purchase abandonment determination history 184 in Example 2. Details of these will be explained later. Furthermore, the building user information 181, permanent store information 182, temporary store information 183, leaver abandonment determination history 184-3, policy information 185, and store visit status information 186 are the same as those in Example 1, so explanations thereof will be omitted.

[0110] 13 is a hardware configuration diagram of the purchase-related behavior analysis device 1 in Example 2. In Example 2 as well, the purchase-related behavior analysis device 1 can be realized by a server, which is an example of a computer, in particular, by a cloud. Therefore, in Example 2 as well, similar to Example 1, the purchase-related behavior analysis device 1 includes a processing device 101, a communication device 102, a memory 103, and a secondary storage device 104, which are connected to each other via a communication path.

[0111] However, it differs from the first embodiment in that the configuration of the purchase-related behavior analysis program 105 and the defector abandonment determination history 184-3 are stored in the secondary storage device 104. Here, the difference in the configuration of the purchase-related behavior analysis program 105 will be described.

[0112] In comparison with the first embodiment, the second embodiment uses a matrix extraction module 113, a leaver determination module 114, and a leaver counting module 115. First, the matrix extraction module 113 realizes the function of the matrix extraction unit 153. Furthermore, the leaver determination module 114 realizes the function of the leaver determination unit 154. Furthermore, the leaver counting module 115 realizes the function of the leaver counting unit 155. These may also be realized by individual programs, as in the first embodiment.

[0113] Next, the leaver abandonment judgment history 184-3, which is information specific to the second embodiment, will be described. FIG. 14 is a diagram showing the leaver abandonment judgment history 184-3 used in the second embodiment. This is an example of the purchase abandonment judgment history 184 in FIG. 1. The leaver abandonment judgment history 184-3 is information indicating the situation of leaving a queue and the cause of the corresponding leaving based on this. Here, leaving a queue can be inferred as abandoning a purchase due to the queue. Therefore, in the second embodiment, the situation of leaving, particularly the number of leavers, is used to determine whether purchase abandonment due to the queue has occurred at the corresponding store.

[0114] Therefore, the leaver abandonment judgment history 184-3 has the following items: store name, queue, time period of occurrence, leavers, number of leavers (percentage), and cause.

[0115] First, the store name indicates the name of the store where the dropout occurred, but any item that can identify a permanent store will suffice. The queue is an item that identifies the detected queue. The occurrence time period is an item that indicates the time period during which the queue occurred. The dropouts are an item that includes users who have left the queue, as well as users who have changed direction while heading to the end of the queue and have given up on their purchase.

[0116] The number of leavers (ratio) is an item that indicates the number of relevant leavers and their ratio to the number of purchasers. Furthermore, as for the cause, if the number of leavers (ratio) is above a certain level, queuing is considered to be the cause. Furthermore, if a queue exceeding a threshold occurs within a predetermined period (one day, one week), the policy information 185 may be created.

[0117] The item for leavers can be omitted. In particular, when image processing is used, it is desirable to omit it, since it is not possible to identify individual users. On the other hand, when the user terminal 7 is used, it is possible to identify users, so it is possible to provide an item for leavers.

[0118] This concludes the description of the information used in the second embodiment, and next, a description will be given of the processing flow of the second embodiment. Fig. 15 is a flowchart showing the processing flow in the second embodiment. The following description will be given along this flowchart, but the processing will mainly use the configuration in Fig. 12.

[0119] First, steps S11 to S13 are almost the same as those in the first embodiment, and therefore the description thereof will be omitted. However, in step S13, the queue situation estimation unit 14 stores the estimated results in the queue and occurrence time period of the leaver purchase abandonment determination history 184-3.

[0120] In steps S21 to S23, the behavior estimation unit 15 estimates the status of leaving the queue as the behavior of the customer based on the extracted store visit status and the created queue status. This behavior estimation process is repeatedly executed for each queue. In addition, it is desirable to periodically execute the behavior estimation process, such as when the queue clears or once a day.

[0121] First, in step S21, the queue extraction unit 153 extracts a target queue from among the queues indicated by the estimated queue situation. This extraction may be performed in order of occurrence or in a predetermined store order.

[0122] Furthermore, in step S22, the leaver determination unit 154 determines a leaver from the extracted matrix. That is, the occurrence of a leaver is detected. To this end, the leaver determination unit 154 performs image processing or uses time information included in the location information of the user terminal 7. Note that AI technology can be utilized in this image processing. Then, the leaver determination unit 154 stores information on the determined leaver in the leaver abandonment determination history 184-3. Specifically, the leaver determination unit 154 stores the ID of the user determined as a leaver in the leaver. Note that, as described above, if the leaver item is omitted, it is sufficient to determine in step S22 that a leaver has been present. Alternatively, the leaver determination unit 154 may assign a virtual ID to the detected user and store this.

[0123] In step S23, the leaver counting unit 155 counts the number of leavers determined in step S22. Then, the leaver counting unit 155 stores this as the number of leavers. In addition, the leaver counting unit 155 may calculate a ratio from the counted number of leavers and the number of people in line in the estimated queue situation, and store this in the leaver abandonment determination history 184-3.

[0124] In step S24, the purchase abandonment reason determination unit 16 determines the number of leavers for each matrix in the leaver purchase abandonment determination history 184-3. That is, it is determined whether there are zero leavers, many leavers, or few leavers. As a result, it is determined whether there are many purchase abandonments due to queues in the corresponding matrix. Here, a large number of leavers means that the number or rate of leavers is greater than or equal to a predetermined threshold.

[0125] Furthermore, if the number or ratio of leavers in each matrix of the leaver abandonment judgment history 184-3 is greater than or equal to a predetermined number, it may be judged to be high. Furthermore, a number of leavers of 0 means that there are no leavers, and may also be 0 if the number is less than a predetermined number. Furthermore, a low number of leavers means that the number of leavers is between 0 and a high number. As a result, if there are 0 leavers, the process transitions to step S26. Furthermore, if there are many leavers, the process transitions to step S26. Furthermore, if there are few leavers, the process transitions to step S27.

[0126] Furthermore, in step S25, the purchase abandonment reason determination unit 16 determines that there was no purchase abandonment for the corresponding queue, and stores "none" as the cause in the leaver abandonment determination history 184-3. Furthermore, in step S26, the purchase abandonment reason determination unit 16 determines that the reason for purchase abandonment for the corresponding queue is a cause other than a queue, and stores "other than a queue" as the cause in the leaver abandonment determination history 184-3. Furthermore, in step S27, the purchase abandonment reason determination unit 16 determines that the reason for purchase abandonment for the corresponding queue is a queue, and stores "queue" as the cause in the leaver abandonment determination history 184-3.

[0127] In step S28, the policy creation unit 17 uses the leaver purchase abandonment determination history 184-3 to create policy information 185 including a plan to improve congestion represented by a queue. Then, the policy creation unit 17 stores the policy information 185 in the storage unit 18.

[0128] Here, in the second embodiment, similarly to the first embodiment, the policy creation unit 17 creates recommendation information about stores where queues are the cause, and creates usage suggestion information for users stored as dropouts. In this case, the policy creation unit 17 may create recommendation information about stores where the number of queues stored as the cause is a predetermined number or more. In addition, the policy creation unit 17 may create usage suggestion information for users stored as dropouts a predetermined number or more.

[0129] This concludes the explanation of Figure 15, but step S24 can also be implemented in the following modified example. The purchase abandonment reason determination unit 16 classifies each number of leavers in the leaver purchase abandonment determination history 184-3 into multiple ranks. For example, if there are 0 leavers, it is marked as X, if there are 1 to 10 leavers, it is marked as △, and if there are 11 or more leavers, it is marked as O. Points are also awarded for each rank. For example, X is given as 0 points, △ is given as 0.5 points, and O is given as 1 point. The purchase abandonment reason determination unit 16 then tallies the number of points for each store and regards this as the cause of the store in question. In other words, if the number is equal to or greater than a threshold, it is determined that the reason for abandoning purchase at the store in question is due to queues.

[0130] This concludes the explanation of Example 2. In Example 2, dropouts are extracted from the queue and the reason for abandoning the purchase is determined using this, which enables more efficient analysis. Also, in the above example, analysis was performed for each store, but permanent stores and temporary stores may be grouped separately and processing may be performed for each group. In other words, the processing of each flowchart may be executed. [Example]

[0131] Next, in Example 3, the behavior of users in a specified area when they abandon a purchase is tracked to determine the reason for abandoning a purchase at a store, particularly a temporary store. In Example 3, policy information 185 is created based on the reason for abandoning a purchase at this store. For example, in Example 3, reasons for abandoning a purchase at multiple stores are tallied, and policy information 185 is created based on this. In Example 3, it is also determined whether a user has abandoned a purchase based on the tracked behavior of the user, and the results are tallied to determine the reason for abandoning a purchase. Details of this are explained below.

[0132] The configuration and information of Example 3 are the same as those of Example 1, and therefore a description thereof will be omitted. Next, Fig. 16 is a flowchart showing the processing flow in Example 3. The following description will be given along this flowchart, but the processing will be carried out using the configuration of Fig. 3. The following description will be given using an example of opening a temporary store in Building A.

[0133] In the processing flow of the third embodiment, first, steps S11 to S15 in Fig. 10 and step S161 in Fig. 11 are executed in the same manner. Then, if it is determined in step S161 that the transaction has not reached the cash register (N), the process proceeds to step S301.

[0134] Furthermore, in step S301, the individual reason determination unit 161 determines whether the destination of the user who abandoned the purchase is within a predetermined area or outside the predetermined area. Here, the predetermined area refers to the above-mentioned affected area, such as the premises of building A where a temporary store is located, or the temporary store area within that area where the temporary store is located. Furthermore, the individual reason determination unit 161 determines whether the destination is within a predetermined area or outside the predetermined area, using the purchase-related behavior determined by the purchase-related behavior determination unit 152. In other words, the individual reason determination unit 161 determines the destination based on the behavior of the user who abandoned the purchase estimated by the behavior estimation unit 15.

[0135] If the result of step S301 is that the destination is within the predetermined area, the process proceeds to step S302. If the destination is outside the predetermined area, the process proceeds to step S307. Here, the destination within the area can be assumed to mean that the customer will purchase the item from another temporary store within the predetermined area instead of the temporary store where they gave up on purchasing the item.

[0136] Then, in step S302, the individual reason determination unit 161 determines that the reason for the user's abandonment of the purchase is a mismatch in preferences. In other words, it is determined that the user's preferences do not match the temporary store for which the purchase was abandoned. Also, in step S303, the reason tallying unit 162 tally the number of times that a mismatch in preferences has been determined at a predetermined interval. In response to this, in step S304, the reason tallying unit 162 determines whether the number of times tallied in step S303 is equal to or greater than a predetermined number. As a result, if it is equal to or greater than the predetermined number (Y), the process proceeds to step S305. If it is less than the predetermined number (N), the process proceeds to step S321.

[0137] Here, the number of times in steps S303 and S304 may be the total number of times a purchase was abandoned or the ratio of the total number of times to all users including those who made a purchase. Furthermore, the predetermined number of times or the predetermined ratio used in step S304 may be set according to the number of customers of the temporary store, the number of customer movements, or the genres offered. Then, in step S305, the reason tallying unit 162 determines that the reason for abandoning a purchase at the temporary store, i.e., the store, is a preference mismatch.

[0138] In step S306, the individual reason determination unit 161 determines the queue situation estimated by the queue situation estimation unit 14. That is, the individual reason determination unit 161 classifies the situation into one of (1) to (3). (1) The length of the queue at each store (the relevant temporary store and other temporary stores) in the specified area is equal to or longer than a specified length. (2) The length of the queue at the relevant temporary store and some other temporary stores is longer than the specified length. (3) The queue length at each store (the relevant temporary store and other temporary stores) in the specified area is less than a specified length. In other words, there are no stores where the queue length is longer than a specified length.

[0139] Here, the individual reason determination unit 161 determines the queue length in the same manner as in step S164. If none of (1) to (3) applies, the process proceeds to step S321, and a message indicating that there is no measure information to be proposed is output. Furthermore, in (1), the queue length at all stores does not have to be greater than a predetermined value. For example, the queue length at stores that are equal to or greater than a first threshold, such as 90% or more of all stores, may be greater than a predetermined value. In addition, in (3), the queue length at all stores does not have to be less than a predetermined value. For example, the queue length at stores that are less than a second threshold, such as 10% or less of all stores, may be greater than a predetermined value. And in (2), the queue length at stores that are less than the first threshold but greater than a second threshold may be greater than a predetermined value.

[0140] In the case of (1), in step S307, the individual reason determination unit 161 determines that the reason for abandoning the purchase of the user in question is queueing. In step S308, the reason counting unit 162 counts the number of times that it was determined to be queueing in step S307 at a predetermined cycle. In response to this, in step S308, the reason counting unit 162 determines whether the number of times counted in step S308 is equal to or greater than a predetermined number of times. As a result, if it is equal to or greater than the predetermined number of times (Y), the process proceeds to step S310. If it is less than the predetermined number of times (N), the process proceeds to step S321. The determination in step S309 is performed in the same manner as in step S305.

[0141] Then, in step S310, the reason tallying unit 162 determines that the reason for abandoning the purchase at the relevant temporary store, i.e., the store, is excessive demand. In other words, it is determined that there are too many customers in the entire predetermined area (each store), and therefore supply is not keeping up with demand.

[0142] In the case of (2), in step S311, the individual reason determination unit 161 determines that the reason for abandoning the purchase of the user in question is queueing. In step S312, the reason tallying unit 162 tally the number of times that it was determined to be queueing in step S311 at a predetermined cycle. In response to this, in step S313, the reason tallying unit 162 determines whether the number of times tallied in step S312 is equal to or greater than a predetermined number of times. As a result, if it is equal to or greater than the predetermined number of times (Y), the process proceeds to step S314. If it is less than the predetermined number of times (N), the process proceeds to step S321. The determination in step S313 is also performed in the same manner as in steps S305 and S309.

[0143] Then, in step S314, the reason tallying unit 162 determines that the reason for abandoning the purchase at the relevant temporary store, i.e., the store, is biased. In other words, it is determined that popularity is biased (for example, concentrated) toward some stores in the entire predetermined area (each store) where queues of a predetermined length or more are formed.

[0144] In the case of (3), in step S315, the individual reason determination unit 161 determines that the reason for abandoning the purchase of the user in question is other than queueing. In step S316, the reason tallying unit 162 tally the number of times that it was determined in step S315 that the reason was other than queueing, at predetermined intervals. In response to this, in step S317, the reason tallying unit 162 determines whether the number of times tallying in step S316 is equal to or greater than a predetermined number of times. As a result, if it is equal to or greater than the predetermined number of times (Y), the process proceeds to step S318. If it is less than the predetermined number of times (N), the process proceeds to step S321. The determination in step S318 is also performed in the same manner as in steps S305, S309, and S313.

[0145] Then, in step S318, the reason tallying unit 162 determines that the reason for abandoning the purchase at the relevant temporary store, i.e., the store, is insufficient demand (excess supply). In other words, it is determined that there is more supply than demand because there are fewer customers in the entire predetermined area (each store).

[0146] As described above, the reasons for abandoning purchases at each store are determined in steps S305, S310, S314, and S318. In response to this, in step S319, the reason tallying unit 162 tallyes the reasons for abandoning purchases at each store determined in steps S305, S310, S314, and S318. The reason tallying unit 162 tallyes the number of each of the reasons for abandoning purchases, including preference mismatch, excess demand, bias, and insufficient demand.

[0147] Then, in step S320, the policy creation unit 17 creates policy information 185 in accordance with the number of reasons for abandoning purchase that have been tallied for each store. At this time, the policy creation unit 17 stores the policy information 185 in the storage unit 18.

[0148] Here, a specific example of creating the policy information 185 will be described. The policy creating unit 17 extracts reasons for abandoning purchases in a predetermined area according to the numbers calculated from preference mismatch, excess demand, bias, and insufficient demand. This extraction includes extracting the most common reason or reasons that are above a predetermined threshold.

[0149] Then, the policy creation unit 17 creates policy information 185 including a proposal according to the extracted reason. In the case of a preference mismatch, the policy creation unit 17 creates policy information 185 proposing to replace the temporary store that is the relevant store. In this case, the policy creation unit 17 identifies a store that matches the preference of the relevant user in step S302 as a store to replace the temporary store. For this purpose, the policy creation unit 17 searches the temporary store information 183 for a temporary store that has a menu offering that matches the attributes of the building user information 181.

[0150] Furthermore, if the reason for abandoning a purchase in a predetermined area is excessive demand, the policy creation unit 17 creates policy information 185 that proposes adding temporary stores to the predetermined area and increasing the number of stores. The policy creation unit 17 determines the number of temporary stores to add based on the queue situation determined in step S306 and the number of times tallied in step S308. In other words, the number of stores is determined based on at least one of the queue length, the number of queues, and the number of times. The longer or larger these are, the more stores are added. Note that the temporary stores to be added may be similar stores to temporary stores already open in the predetermined area, similar to the bias described below.

[0151] Furthermore, if the reasons for abandoning purchases in a predetermined area are biased, the policy creation unit 17 creates policy information 185 that proposes increasing the number of similar stores. To this end, the policy creation unit 17 identifies the menu items offered at the store determined in step S306 to have a queue length equal to or longer than a predetermined length from the temporary store information 183. Then, the policy creation unit 17 identifies other temporary stores that have the identified menu items offered from the temporary store information 183, and sets these as similar stores. Note that similar stores only need to have at least a portion of the menu items that match those of the store with a queue length equal to or longer than a predetermined length.

[0152] Furthermore, the number of stores to be increased can be determined according to at least one of the length of queues, the number of queues, and the frequency of queues, as in the case of excessive demand. In Example 3, similar stores were identified using menus, but other characteristics such as genre may also be used. Furthermore, since the increase in similar stores includes the replacement of some of the existing temporary stores, the total number of stores does not necessarily have to increase, and the number of stores may be maintained or decreased.

[0153] Furthermore, if the reason for abandoning purchases in the predetermined area is insufficient demand, the policy creation unit 17 creates policy information 185 that proposes repurposing the predetermined area. In other words, the temporary stores are withdrawn from the predetermined area and converted to another use such as a bicycle parking lot. Note that the temporary stores to be withdrawn are not limited to all the temporary stores, but may be only some of the temporary stores. Furthermore, withdrawal also includes not opening a new store after the opening period of a currently opened temporary store has expired (including cancellation of a planned store opening) and withdrawal during the opening period.

[0154] Note that, similarly to the first and second embodiments, the policy creation unit 17 may create the policy information 185 when the reason for abandoning purchase satisfies a predetermined condition. Furthermore, if there are reasons for abandoning purchase (such as inconsistencies in preferences) for two or more temporary stores (outlets) as a result of step S3319, the policy creation unit 17 may create the policy information 185 based on these. Furthermore, the predetermined condition for the reason for abandoning purchase may be set based on the reason for abandoning purchase.

[0155] Then, in step S321, the measure creation unit 17 outputs the result to the related device via the output unit 12. That is, if the reason for abandoning the purchase at the store has been identified (steps S305, S310, S314, and S318), the measure creation unit 17 transmits the created measure information 185 to the related device via the output unit 12. On the other hand, if the reason for abandoning the purchase has not been identified, the measure creation unit 17 transmits to the related device via the output unit 12 a message indicating that no measure is necessary.

[0156] This concludes the explanation of FIG. 16. Steps S310, S314, and S318 theoretically do not coexist because the queue conditions are different. However, it is possible that they may coexist depending on the timing at which the queue conditions are estimated. For this reason, the reasons for abandoning a purchase at a store may all coexist: preference mismatch, excess demand, imbalance, and insufficient demand. As a result, the policy information 185 may include all of these corresponding suggestions.

[0157] Next, a modified example of the third embodiment will be described. In the process of Fig. 16, each reason for abandoning a purchase is tallied, and action information 185 is created based on the reason for abandoning a purchase extracted from the tallied result. In contrast, in this modified example, action information 185 is created based on each reason for abandoning a purchase. This will be described in detail below. Fig. 17 is a flowchart showing the processing flow in the modified example of the third embodiment.

[0158] In Fig. 17, steps S301 to S318 are the same processes as those in Fig. 16, and therefore the description thereof will be omitted. Then, in steps S320-1 to S320-4, the measure creation unit 17 creates measure information 185 according to the reason for abandoning purchase at the identified store.

[0159] In step S320-1, the policy creation unit 17 creates policy information 185 proposing replacing the temporary store in question. Also, in step S320-2, the policy creation unit 17 creates policy information 185 proposing adding a temporary store to a predetermined area and increasing the number of stores. Also, in step S320-3, the policy creation unit 17 creates policy information 185 proposing increasing the number of similar stores. Furthermore, the policy creation unit 17 creates policy information 185 proposing changing the use of the predetermined area. These are executed in the same manner as in FIG. 16.

[0160] As a result, based on the reasons for abandoning purchase identified as reasons for abandoning purchase at the store, respective policy information 185 is created. Then, similarly to FIG. 16, the results are output in step S321.

[0161] This concludes the explanation of the modified example. In the third embodiment, it is possible to propose appropriate measures according to the store visit situation, such as when there is excess demand, when customers are concentrated in one area, or when the number of customers varies due to characteristics such as menus or genres. [Explanation of symbols]

[0162] 1...purchase-related behavior analysis device, 11...input unit, 12...output unit, 13...store visit status collection unit, 14...queue status estimation unit, 15...behavior estimation unit, 151...target extraction unit, 152...purchase-related behavior determination unit, 153...queue extraction unit, 154...dropout determination unit, 155...dropout counting unit, 16...purchase abandonment reason determination unit, 161...individual reason determination unit, 162...reason counting unit, 17...measure creation unit, 18...memory unit, 181...building user information, 182...permanent store information, 183...temporary store information, 184...purchase abandonment determination history, 18 4-1...Individual purchase abandonment judgment history, 184-2...Aggregate purchase abandonment judgment history, 184-3...Leaver purchase abandonment judgment history, 185...Measure information, 186...Store visit status information, 2a...Building management system, 21a...Building management terminal, 22a...Building management server, 23a...Camera, 2b...Building management system, 21b...Building management terminal, 23b...Camera, 23c...Camera, 3...Building management database, 4...Temporary store database, 5...Permanent store system, 6...Temporary store system, 7...User terminal, 8...Network

Claims

1. A purchase-related behavior analysis device that analyzes reasons for abandoning purchases in stores, a store visit status collection unit that collects store visit status indicating the behavior of people in the store; a queue status estimation unit that estimates a queue status at the store based on the store visit status; a behavior estimation unit that estimates the purchase behavior of a user who is a target of analysis of reasons for abandoning a purchase based on the store visit status; The purchase-related behavior analysis device has a purchase abandonment reason determination unit that, when the behavior indicates purchase abandonment, determines whether the reason for purchase abandonment is due to the queue, based on the queue situation and the behavior.

2. 2. The purchase-related behavior analysis device according to claim 1, the store visit status collection unit collects store visit statuses of a plurality of users at the store, The purchase abandonment reason determination unit an individual reason determination unit that determines the reason for abandoning purchase for each of the plurality of users; A purchase-related behavior analysis device having a reason counting unit that counts the reasons for abandoning the purchase for each of the plurality of users.

3. 3. The purchase-related behavior analysis device according to claim 2, Further, a policy creation unit is provided for creating recommendation information using the collected reasons for abandoning purchase, The store is a temporary store, The purchase-related behavior analysis device, wherein the recommendation information is recommendation information for changing at least one of the number and genre of the temporary stores.

4. 4. The purchase-related behavior analysis device according to claim 3, The purchase-related behavior analysis device, wherein the policy creation unit creates the recommendation information when the reason for abandoning a purchase due to the queue satisfies a predetermined condition.

5. 5. The purchase-related behavior analysis device according to claim 4, The predetermined condition is set based on the number of times that the reason for abandoning a purchase is due to queuing, or the ratio of the reasons for abandoning a purchase due to queuing to reasons other than queuing.

6. 3. The purchase-related behavior analysis device according to claim 2, The store is a temporary store, A purchase-related behavior analysis device having a user suggestion information creation unit that creates usage suggestion information suggesting to the user that they use a permanent store located within a specified range of the temporary store if the number of times that the reason for abandoning a purchase is due to the queue, or the ratio of the reasons for abandoning a purchase due to the queue to the ratio of reasons for abandoning a purchase due to factors other than the queue, is above a specified threshold.

7. 2. The purchase-related behavior analysis device according to claim 1, The behavior estimation unit is a purchase-related behavior analysis device that estimates, as the behavior, the movement of the user in an analysis target area for the store.

8. 8. The purchase-related behavior analysis device according to claim 7, The behavior estimation unit sets the analysis area for each store, and the analysis area is set according to location conditions of the store.

9. 2. The purchase-related behavior analysis device according to claim 1, the store visit status collection unit collects store visit statuses of a plurality of users at the store, The store visit status is image information of the plurality of users or location information of the users.

10. 2. The purchase-related behavior analysis device according to claim 1, The behavior estimation unit a leaver determination unit that determines a person leaving the queue based on the store visit status and the queue status; a leaver counting unit that counts the number of leavers, The purchase-related behavior analysis device, wherein the purchase abandonment reason determination unit determines whether the reason for abandoning a purchase is due to the queue, based on the counted number of abandoners.

11. 4. The purchase-related behavior analysis device according to claim 3, the behavior estimation unit estimates the behavior of the user in a predetermined area having a plurality of stores; the individual reason determination unit determines the reason for the user abandoning the purchase based on the behavior of the user; the reason tallying unit tallying the reasons for abandoning a purchase of a plurality of the users and determining the reason for abandoning a purchase at the store; The policy creation unit creates the recommendation information based on the determined reasons for abandoning purchases at the store.

12. The purchase-related behavior analysis device according to claim 11, The purchase-related behavior analysis device wherein the policy creation unit compiles reasons for abandoning a purchase at a plurality of the stores, and creates the recommendation information based on the compiled reasons for abandoning a purchase at the plurality of stores.

13. The purchase-related behavior analysis device according to claim 12, the policy creation unit creates the recommendation information when the reason for abandoning purchase due to the queue satisfies a predetermined condition; The recommendation information is created based on the aggregated reasons for abandoning purchases at at least two of the plurality of stores.

14. The purchase-related behavior analysis device according to claim 12, the policy creation unit creates the recommendation information when the reason for abandoning purchase due to the queue satisfies a predetermined condition; The predetermined condition is set based on the reasons for abandoning purchases at at least two of the plurality of stores that have been collected.

15. A purchase-related behavior analysis method executed by a purchase-related behavior analysis device that analyzes reasons for abandoning a purchase in a store, the purchase-related behavior analysis method comprising: a store visit status collection unit; a queue status estimation unit; a behavior estimation unit; and a purchase abandonment reason determination unit, the store visit status collection unit collects store visit status indicating the behavior of people in the store; the queue status estimation unit estimates a queue status at the store based on the store visit status; the behavior estimation unit estimates the purchase behavior of the user who is the subject of analysis of reasons for abandoning purchase based on the store visit state; The purchase-related behavior analysis method, wherein the purchase abandonment reason determination unit determines, if the behavior indicates purchase abandonment, whether the reason for purchase abandonment is caused by the queue, based on the queue situation and the behavior.

Images