JP6458572B2 - Customer behavior management device, customer behavior management method and program - Google Patents

Description

The present invention relates to a customer behavior management device , a customer behavior management method, and a program.

  A cleaning store refers to a store that performs cleaning on clothes such as clothes, bedding such as blankets, hats, curtains, carpets, and the like. When a cleaning request is received from a customer at a cleaning store, the stored item is recorded, a slip is issued, a price is received, the product is sent to a factory for cleaning, and then delivered to the customer.

  This cleaning shop has busy days and busy hours, and there are quiet days and quiet hours. In a laundry, for example, Saturdays and Sundays are busy days, and breaks in the afternoon or after lunch are often busy hours on weekdays. On the other hand, the dry hours of the laundry are often during the daytime on weekdays.

  During busy hours, customer waiting time is lengthened, which can lead to customer churn. During off hours, the human resources and physical resources of the laundry may not be used effectively. Further, the wave of busyness and quietness is not desirable because it wastes the human resources and physical resources of the cleaning shop.

  In addition, for a cleaning company having a plurality of stores, there are busy stores and secluded stores depending on the affiliated stores. Such a difference between busyness and quietness by the store is not desirable because it wastes human resources and physical resources of the cleaning company.

  Conventionally, in order to level such busyness and quietness, it has often been motivated to visit the low season and the quiet shop. For example, the problem of the abstract of Patent Document 1 is described as “Provide an information processing apparatus that appropriately provides users with information on promotions that encourage users to visit restaurants”. “A usage information storage unit 11 that stores usage information that is information related to the use of a restaurant by one or more users, a usage information reception unit 12 that receives usage information and accumulates it in the usage information storage unit 11, Promotion information storage unit 13 for storing promotion information that is information for prompting use of the store, and promotion information for acquiring promotion information corresponding to the usage information stored in the usage information storage unit 11 from the promotion information storage unit 13 An acquisition unit 14; a contact information storage unit 15 that stores one or more contact information that is information indicating a contact corresponding to the user; A transmission unit 16 that transmits the promotion information acquired by the application information acquisition unit 14 to a contact indicated by one or more pieces of contact information stored in the contact information storage unit 15. Yes.

JP 2014-59786 A

  Such motivation is performed by providing points or discounts to customers. In other words, a part of the profit as a cleaning shop is provided to the customer. Therefore, if motivation is provided for all customers, there is a risk of squeezing store profits.

Therefore, an object of the present invention is to provide a customer behavior management apparatus , a customer behavior management method, and a program that motivate only customers who may visit the store.

In order to achieve the above object, the customer behavior management device according to the present invention includes a customer behavior predicting unit that predicts a visit date and time of a customer based on a customer visit history to the store and the customer stop location information, If the store visit date and time predicted by the customer behavior predicting means corresponds to the busy time of the store and the store visit status determining means, and the store visit status determining means determines that the customer visit date and time corresponds to the busy time, An alternative date and time notifying means for notifying the customer of information prompting the customer to visit another date and time , wherein the customer behavior predicting means determines whether or not there is a finished product for the customer when predicting the customer's date and time of visiting the customer. The store visit probability is calculated according to the determined presence / absence of the finished product, and the visit date and time of the customer is also predicted .
Further, the customer behavior management method according to the present invention is a customer behavior management method executed by the customer behavior management device, and the customer visit date and time of the customer based on the customer visit history to the store and the customer stop information. Predicting customer behavior, visiting date and time predicted by the customer behavior predicting step whether or not the store visit status determining step determines whether or not the store busy time, and the store visit status determining step. If it is determined that this is a busy time, an alternative date and time notifying step of notifying the customer of information prompting the customer to visit another date and time, the customer behavior predicting step Determine the presence or absence of finished products related to the customer, calculate the store visit probability according to the determined presence or absence of the finished product, and also predict the customer's visit date and time, And wherein the door.
Further, the program according to the present invention predicts a computer by a customer behavior predicting means for predicting a visit date and time of the customer based on a customer visit history to the store and the customer's visiting location information, and the customer behavior predicting means. If the visit date / time of the customer corresponds to the busy time of the store, the visit status determination means for determining whether the visit date / time corresponds to the busy time of the store, Functioning as an alternative date and time notification means for notifying the customer of the information, and the customer behavior prediction means determines whether or not there is a finished product for the customer in predicting the visit date and time of the customer. The store visit probability is calculated according to the presence or absence, and the visit date and time of the customer is also predicted.

According to the present invention, it is possible to provide a customer behavior management device , a customer behavior management method, and a program that motivate only customers who may visit the store.

It is a block diagram which shows the outline of the cleaning system containing the customer action management server in 1st Embodiment. It is a figure which shows an example of a customer action history database. It is a figure which shows an example of a customer database. It is a figure which shows an example of a consignment management database. It is a state transition diagram of one individual when finished goods exist in a store. It is a figure which shows a state transition at the time of finishing goods existing in a store, and getting off at Kokubunji Station. It is a figure which shows the state transition in which finishing goods exist in a shop, and there is no position information (home). It is a state transition diagram of one individual when finishing goods are not in a store. It is a figure which shows a state transition without finishing goods in a shop and no position information (home). It is a flowchart of the notification process in 1st Embodiment. It is a block diagram which shows the outline of the cleaning system containing the customer action management server in 2nd Embodiment. It is a block diagram which shows the estimation of the number of customers of a Kokubunji store. It is a block diagram which shows the estimation of the number of visitors of a Hamura store. It is a figure which shows an example of a positional information database. It is a figure which shows an example of a store database. It is a flowchart of the notification process in 2nd Embodiment. It is a flowchart of the store stop prediction process in 2nd Embodiment.

Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the drawings.
<< First Embodiment >>
In the first embodiment, the customer's visit date / time is predicted based on the customer's visit history, and when the predicted visit date / time corresponds to the busy time of the store, an email is sent to the customer urging to visit the store at another date / time. , Level customer visits. Hereinafter, the configuration and operation will be described with reference to FIGS.

FIG. 1 is a configuration diagram showing an outline of a cleaning system 1 including a customer behavior management server 2 in the first embodiment.
The customer behavior management server 2 is a customer behavior management device that manages customer behavior based on a customer behavior history, and includes a ticket gate 4 at each station, a vending machine 5 installed at each location via a network 9, The store terminal 6 is communicably connected. Furthermore, the customer behavior management server 2 can send an email to the smartphone 7 via the network 9. The customer behavior management server 2 includes a customer behavior prediction unit 21, a busy time estimation unit 22, a store visit status determination unit 23, an alternative date and time notification unit 24, a mail transmission unit 28, and a short-range wireless notification unit 29. The customer behavior is managed while referring to the customer behavior history database 31, the customer database 32, and the custody management database 33. In the drawings, the database is simply abbreviated as “DB”.

The ticket gate 4 is an automatic ticket gate device installed at each station, and can communicate with an electronic money card or an electronic money terminal by a short-range wireless communication unit 41 that performs NFC (Near field communication) communication, for example. Here, the short-range wireless communication unit 41 can communicate with the short-range wireless communication unit 71 of the smartphone 7, for example, to settle the railway fare.
The vending machine 5 is installed in each place, and can communicate with an electronic money card or an electronic money terminal by a short-range wireless communication unit 51 that performs NFC communication. Here, the short-range wireless communication unit 51 can communicate with, for example, the short-range wireless communication unit 71 of the smartphone 7 and settle a sale item.

The smartphone 7 is, for example, a mobile phone terminal that includes a touch display and has an electronic money settlement function, and includes a short-range wireless communication unit 71, a wireless communication unit 72, a notification unit 73, and an ID storage unit 74. ing.
The short-range wireless communication unit 71 performs NFC communication with the ticket gate 4 and the vending machine 5 for electronic money settlement, for example. The short-range wireless communication unit 71 transmits identification information stored in an ID storage unit 74 described later to the communication destination.
The wireless communication unit 72 is for accessing the network 9 via a wireless communication path, making a call to another station, or sending / receiving mails to / from other terminals.
The alerting | reporting part 73 is a display, for example, and alert | reports a mail to a customer.
The ID storage unit 74 is a part that stores a unique identification number of the smartphone 7.

This customer behavior management server 2 includes a CPU (Central Processing Unit) (not shown), a ROM (Read Only Memory), a RAM (Random Access Memory), and a storage unit, and the customer behavior management in which the CPU is stored in the storage unit. Each processing unit is realized by executing a program. The customer behavior management server 2 collects the usage history of each customer's smartphone 7 via the short-range wireless communication unit 41 of the ticket gate 4 or the short-range wireless communication unit 51 of the vending machine 5 and stores it in the customer behavior history database 31. To do. As a result, the customer stop history information is recorded in the customer behavior history database 31.
The customer behavior management server 2 further collects store visit history of each customer from the store terminal 6 and stores it in the customer behavior history database 31. As a result, the customer visit history database 31 records the visit history of each customer at each store.

The customer behavior prediction unit 21 predicts the customer visit date and time based on the visit history of the customer to the store and the customer stop information. Each customer's store visit history and stopover information are stored in the customer behavior history database 31.
The busy time estimation unit 22 estimates the busy time of the store.
The store visit status determination unit 23 determines whether the store visit date and time predicted by the customer behavior prediction unit 21 corresponds to the busy time of the store.
The alternative date and time notification unit 24 notifies the customer of information prompting the customer to visit another date and time when the store visit status determination unit 23 determines that the customer's date and time of visiting the store corresponds to a busy time.
The mail transmission unit 28 transmits a useful message to the customer or the smartphone 7 owned by the mail. The short-range wireless notification unit 29 notifies the customer of a useful message from the ticket gate 4 or the vending machine 5 via the short-range wireless communication unit 71 of the smartphone 7 owned by the customer. Useful messages here are, for example, “10 times points when you come to pick up after x” and “10 times points when you pick up on y month and z”. It is a thing to encourage.

The customer behavior history database 31 is a database in which a customer, a visit history of the customer to the store, and stopover information are accumulated, and will be described later with reference to FIG.
The customer database 32 is a database in which customer identification information is associated with various information related to the customer, and will be described later with reference to FIG.
The consignment management database 33 is a database for managing consignment items that the cleaning store has entrusted by customers, and will be described later with reference to FIG.

FIG. 2 is a diagram illustrating an example of the customer behavior history database 31.
The customer behavior history database 31 stores a customer behavior history, and includes a smartphone ID column 31a, a finished product column 31b, a date / time column 31c, a location column 31d, and an behavior column 31e. In the customer behavior history database 31, a new record is added by the customer behavior management server 2 every time a customer performs a predetermined behavior. The customer behavior history database 31 includes boarding / alighting as pass information of the ticket gate 4 and purchase as usage information of the vending machine 5.
The smartphone ID column 31a is a column for storing identification information of the customer's smartphone 7 according to the action history.

The finished product column 31b is a column that indicates at which store the finished product for cleaning related to this customer exists at the time of this action history. Customer behavior varies depending on the presence of cleaning finishes. For example, the probability of visiting a store is higher when there is a finished product than when there is no finished product. Therefore, the finished product column 31b is provided to judge customers who have finished products and therefore have a high probability of visiting the store. The finished product column 31b is set by referring to the custody management database 33 described later.
The date / time column 31c is a column for storing the date / time when this action history was recorded.
The place column 31d is a column for storing a place where the action history is recorded.
The behavior column 31e is a column for storing customer behavior related to this behavior history.

FIG. 3 is a diagram illustrating an example of the customer database 32.
The customer database 32 includes a smartphone ID column 32a, a customer ID column 32b, a name column 32c, a home address column 32d, a latitude column 32e, and a longitude column 32f. In the customer database 32, a new record is added by the store terminal 6 every time the store registers a new customer.
The smartphone ID column 32a is a column for storing identification information of the smartphone 7 owned by the customer. The store clerk reads the identification information of the customer's smartphone 7 at the store terminal 6 when registering a new customer.
The customer ID column 32b is a column for storing the identification information of the customer.
The name column 32c is a column for storing the name of the customer.
The home address field 32d is a field for storing the home address of the customer.
The latitude column 32e stores latitude information of the customer's home address, and the longitude column 32f stores longitude information of the customer's home address.

FIG. 4 is a diagram illustrating an example of the custody management database 33.
The custody management database 33 includes a customer ID column 33a, a custody ID column 33b, a product name column 33c, a reception date column 33d, a finishing date column 33e, a delivery date column 33f, and a store column 33g. Composed. In the custody management database 33, a new record is added by the store terminal 6 every time a customer deposits clothes or the like to the cleaning store.
The customer ID column 33a is a column for storing identification information of a customer who owns the custody item.
The custody ID column 33b is a column for storing identification information of the custody, and stores, for example, barcode information of a cleaning tag attached to the custody.
The product name column 33c is a column for storing the product name of this custody product. The store clerk inputs the name of the item to be stored at the store terminal 6.
The reception date column 33d stores the date and time when this custody item was received from the customer.
The finishing date column 33e stores a scheduled date when the consignment is finished.
The delivery date column 33f stores the date and time when the finished consignment item was delivered to the customer.
The store column 33g stores the name of the store that received the checked item.
Note that when the finished product is in the store, the current date and time is after the finishing date and no delivery date is set.

  FIG. 5 is a state transition diagram of one individual when there is a finished product for cleaning. In this state transition diagram, the past 365 days are tabulated, and 104 of them are in a state where finished products for cleaning are in the store. This state transition diagram can be created by the customer behavior management server 2 analyzing the customer behavior history database 31, for example.

First, the state transitions from the state M10 where one day has started to the state M11 without customer position information. The number of transitions is 104 (for 104 days).
There are five transitions from the state M11 where there is no customer location information to the state M19 at the Kokubunji store visit, and 99 transitions to the state M12 at the Kokubunji station.
The transition from state K12 at Kokubunji Station to state M13 at Hamura Station is 98 times, and the transition time is 40 minutes on average. This indicates commuting from the nearest station to the workplace.

From the state M13 of getting off at Hamura Station to the state M14 of getting on at Hamura Station, 98 transitions are made, and the transition time is 9 hours on average. This corresponds to working hours at the workplace.
There are 90 transitions from the state M14 at Hamura Station to the state M18 at Kokubunji Station, and the average transition time is 40 minutes. This indicates commuting from the nearest station to the nearest station.

  Further, the state transits 8 times from the state M14 at the Hamura Station to the state M15 using the vending machine 5 in the Hamura Station, and the transition time is 10 minutes on average. The transition from state M15 using the vending machine 5 in the Hamura station premises to state M18 of getting off at Kokubunji station is 8 times, and the transition time is 40 minutes on average. This indicates that the vending machine 5 in the Hamura station was used on the way home.

  There is a transition from state M12 at Kokubunji Station to state M16 at Shinjuku Station, and the transition time is 30 minutes. A transition is made once from the state M16 at the Shinjuku station to the state M17 at the Shinjuku station, and the transition time is 3 hours. A transition is made once from the state M17 at Shinjuku Station to the state M18 at Kokubunji Station, and the transition time is 30 minutes. This shows that I moved from Kokubunji to Shinjuku for private use, spent 3 hours, and returned from Shinjuku to Kokubunji again.

There are 47 transitions from the state M18 at Kokubunji Station to the state M19 at the Kokubunji store, and the transition time is 15 minutes. In addition, the state transits 52 times from the state M18 at the Kokubunji station to the state M20 at the end of the day.
There are 52 transitions from the state M19 at the Kokubunji store to the state M20 at the end of the day. This indicates that the customer went home after receiving the delivery of the cleaning product at the Kokubunji store (cleaning store).

  From this state transition diagram, this customer visits Kokubunji store (cleaning store) when there is a finished product for cleaning, the state M18 when getting off Kokubunji station, and the position information is indefinite since the day starts. Limited to state M11. From this state M18 when the customer gets off Kokubunji Station, the customer visits the Kokubunji store (cleaning shop) in an average of 15 minutes.

FIG. 6 is a graph showing the frequency of state transition when the finished product is in a store and the train gets off at Kokubunji Station. The vertical axis of the graph indicates the number of transitions, the hatched portion indicates the number of visits to the store, the white portion indicates the number of transitions other than visits, and the horizontal axis indicates the time zone. That is, FIG. 6 shows the relationship between the transition from the state M18 of FIG. 5 and the day of the week. The hatched portion of the entire bar graph indicates the store visit frequency.
Here, when we get off Kokubunji Station on Saturday, they are all visiting Kokubunji.
Thus, the store visit probability can be calculated more accurately by multiplying the state transition and the time information (day information).

FIG. 7 is a graph showing the state transition when the finished product is in the store and there is no position information (home). The vertical axis of the graph indicates the number of transitions, the hatched portion indicates the number of visits to the store, the white portion indicates the number of transitions other than visits, and the horizontal axis indicates the time zone. That is, FIG. 7 shows the relationship between the transition from the state M11 of FIG. 5 and the day of the week. The hatched portion of the entire bar graph indicates the store visit frequency.
According to this graph, on Saturdays, Kokubunji stores are visited with high probability from home. Thus, the store visit probability can be calculated more accurately by multiplying the state transition and the time information (day information).

  FIG. 8 is a state transition diagram of one individual when there is no finished product. In this state transition diagram, the past 365 days are totaled, of which 261 days are in a state where there is no finished product for cleaning. This state transition diagram can be created by the customer behavior management server 2 analyzing the customer behavior history database 31, for example.

First, the state transitions from the state M30 in which one day starts to the state M31 in which there is no customer position information. The number of times of transition is 261 times (for 261 days).
There are 42 transitions from the state M31 where there is no customer location information to the state M39 of the Kokubunji store visit, and there are 146 transitions to the state M32 at the Kokubunji station, and the day ends as it is 73 Times.
There are 42 transitions from the state M39 at the Kokubunji store to the state M40 at the end of the day. This indicates that the customer left the cleaning product at home at the Kokubunji store (cleaning store) and then returned home to complete the day.

  There are 134 transitions from the state M32 at Kokubunji Station to the state M33 at Hamura Station, and the average transition time is 40 minutes. This indicates commuting from the nearest station to the workplace.

There are 134 transitions from the state M33 of getting off at Hamura Station to the state M34 of getting on at Hamura Station, and the transition time is an average of 9 hours. This corresponds to working hours at the workplace.
There are 121 transitions from the state M34 at Hamura Station to the state M38 at Kokubunji Station, and the transition time is an average of 40 minutes. This indicates commuting from the nearest station to the nearest station.

  Moreover, it changes 13 times from state M34 of Hamura Station boarding to state M35 using the vending machine 5 in the premises of Hamura Station, and the transition time is an average of 10 minutes. The transition from state M35 using the vending machine 5 in the Hamura station premises to state M38 at Kokubunji station is 13 times, and the transition time is 40 minutes on average. This indicates that the vending machine 5 in the Hamura station was used on the way home.

  There are 12 transitions from the state M32 at Kokubunji Station to the state M36 at Shinjuku Station, and the transition time is 30 minutes. There are 12 transitions from the state M36 at Shinjuku Station to the state M37 at Shinjuku Station, and the transition time is 3 hours. There are 12 transitions from the state M37 at Shinjuku Station to the state M38 at Kokubunji Station, and the transition time is 30 minutes. This shows that I moved from Kokubunji to Shinjuku for private use, spent 3 hours, and returned from Shinjuku to Kokubunji again.

  There are 146 transitions from the state M38 at Kokubunji Station to the state M40 at the end of the day. In addition, there is no transition from the state M38 at the Kokubunji station to the state M39 at the Kokubunji store.

  From this state transition diagram, it can be seen that the customer visits the Kokubunji store (cleaning store) when there is no finished product for cleaning only in the state M31 where the day starts and the position information is indefinite.

FIG. 9 is a diagram showing a state transition in which there is no finished product in the store and there is no position information (home). The vertical axis of the graph indicates the number of transitions, and the horizontal axis indicates the time zone. That is, FIG. 9 shows the relationship between the transition from the state M31 of FIG. 5 and the day of the week. The hatched portion of the entire bar graph indicates the store visit frequency.
This shows that the customer visits the Kokubunji store only on Saturdays and Sundays. Thus, the store visit probability can be calculated more accurately by multiplying the state transition and the time information (day information).

FIG. 10 is a flowchart of the notification process in the first embodiment.
The customer behavior management server 2 executes this notification processing every time a customer behavior history is received, for example.
First, the busy time estimation unit 22 identifies the busy time of this store (step S10). The busy time is specified based on the custody management database 33, but is not limited thereto, and may be manually input in advance.
If it is determined that busy time exists in this store (step S11 → Yes), the customer behavior prediction unit 21 repeats the processing of steps S12 to S21 for each customer in this store.
First, the customer behavior predicting unit 21 determines whether or not the finished product relating to this customer exists in the store (step S13).

If the finished product is present in the store (step S13 → Yes), the customer behavior predicting unit 21 calculates the state transition in the case of the finished product based on the customer behavior history database 31 (step S14), and this state. Each transition probability in the transition is calculated (step S15), and the process proceeds to step S18.
If there is no finished product in the store (step S13 → No), the customer behavior prediction unit 21 calculates a state transition when there is no finished product based on the customer behavior history database 31 (step S16). Each transition probability is calculated (step S17), and the process proceeds to step S18.

In step S18, the customer behavior prediction unit 21 predicts the store visit probability and store visit time from the current state and state transition of the customer. That is, the customer behavior prediction unit 21 determines whether or not the current state of the customer is a state before visiting the store among the state transitions, and if visiting the store, obtains the average time until the store visit and calculates the predicted store visit time. To do. As described above, the customer behavior prediction unit 21 determines whether or not there is a finished product for the customer, calculates the store visit probability according to the determination result, and predicts the visit date and time of the customer. This is because customer behavior changes depending on the presence or absence of finished products.
If it is determined from this store visit time and the busy time of the store that the customer visits the store during the busy time (step S19 → Yes), the store visit status determining unit 23 separates this customer from the alternative date and time notification unit 24. An e-mail that prompts the customer to visit the store is transmitted (step S20). The alternative date and time notification unit 24 may transmit a message urging the smartphone 7 to visit another date and time when the customer touches the ticket gate 4 or the vending machine 5 with the smartphone 7. The alternative date and time notification unit 24 transmits a message to the smartphone 7 via the network 9 and the ticket gate 4 or the vending machine 5 by the short-range wireless notification unit 29.
In step S21, the customer behavior predicting unit 21 ends the notification process when the process is completed for each customer in the store.

The customer behavior management server 2 of the present embodiment predicts whether it will be received today or when it will come from the ticket gate information and finished product information of the customer at the present time and past history information. When the predicted time corresponds to the busy time of the store, a message prompting the customer to visit the store at another date and time is notified to the customer.
By doing in this way, the customer behavior management server 2 shows the alternative time to a customer who has a high probability of coming to the store during busy hours, and gives the motivation to visit the store at the alternative time. Thereby, the customer visit date and time can be leveled. In addition, since it is not necessary to provide leveling motivation to all customers, the influence on store profits can be suppressed.

<< Second Embodiment >>
In the second embodiment, based on the current location information of the customer, the location information of a plurality of stores, and the location information of the customer's home, a store where the customer may stop is predicted, and the predicted store A message prompting customers to visit the store is notified, and the visitors to each store are leveled. The configuration and operation will be described below with reference to FIGS.

FIG. 11 is a configuration diagram showing an outline of a cleaning system 1A including a customer behavior management server 2A in the second embodiment. The same elements as those in the first embodiment shown in FIG.
The customer behavior management server 2A of the second embodiment includes a customer behavior prediction unit 21A and a busy day estimation unit 22A that are different from those of the first embodiment, and can further stop by a store prediction unit 25 and a quiet store estimation unit 26. And an alternative store notification unit 27, a location information database 34, and a store database 35.
Unlike the customer behavior prediction unit 21 of the first embodiment, the customer behavior prediction unit 21A predicts the visit date and time of each customer based on the customer visit history of the store.
Unlike the busy time estimation unit 22 of the first embodiment, the busy day estimation unit 22A estimates the busy day of each store.
The possible store stop prediction unit 25 predicts a store where each customer can stop, based on the store database 35, the customer behavior history database 31, and the location information of the customer's home.
The quiet shop estimation unit 26 estimates a quiet shop among the stores that can be visited predicted by the store predictable shop prediction unit 25.
The alternative store notification unit 27 notifies the customer of information that prompts customers to visit a quiet store.
The location information database 34 is a database that gives location information of each ticket gate 4 or vending machine 5, for example, and will be described later with reference to FIG.
The store database 35 is a database that provides location information of each store, and will be described later with reference to FIG.

FIG. 12 is a block diagram showing the estimation of the number of visitors to the Kokubunji store. The vertical axis of the graph indicates the estimated number of visitors, and the horizontal axis indicates the date. That is, FIG. 12 shows the relationship between the number of customers visiting the store and the date.
Here, all the days exceed the busy threshold, indicating that the store is busy.

FIG. 13 is a block diagram showing the estimation of the number of visitors to the Hamura store. Similar to FIG. 12, the vertical axis of the graph indicates the estimated number of customers and the horizontal axis indicates the date.
Here, all the days do not exceed the busy threshold, indicating that it is a quiet shop. That is, by guiding the customers of the Kokubunji store shown in FIG. 12 to the Hamura store, the customers' stores can be leveled as a whole.

FIG. 14 is a diagram illustrating an example of the position information database 34.
The position information database 34 includes a name column 34a, an address column 34b, a latitude column 34c, and a longitude column 34d.
The name column 34a is a column for storing a name for distinguishing the ticket gate 4 or the vending machine 5 used by the customer. The address column 34b is a column for storing addresses of the ticket gate 4 and the vending machine 5 used by the customer.
The latitude column 34c stores the latitude of the ticket gate 4 and the vending machine 5 used by the customer, and the longitude column 34d stores the longitude. It is possible to know the position of the customer from the latitude / longitude information and to easily calculate the distance to the store.

FIG. 15 is a diagram illustrating an example of the store database 35.
The store database 35 includes a location column 35a, an address column 35b, a busy threshold column 35c, a busy day column 35d, a latitude column 35e, and a longitude column 35f.
The location column 35a is a column for storing the location (name) of each store. The address column 35b is a column for storing the address of each store.
The busy threshold value column 35c is a column for storing the busy threshold value of this store. The busy day column 35d is a column for storing busy days of this store.
The latitude column 35e stores the latitude of the store, and the longitude column 35f stores the longitude. Based on the latitude / longitude information and the Hybeni formula, it is possible to know the location of the store and easily calculate the distance between the customer's home and each station on the commuting route.

FIG. 16 is a flowchart of notification processing in the second embodiment.
The customer behavior management server 2A executes this notification processing each time it receives a customer behavior history, for example.
First, the busy day estimation unit 22A identifies the busy day of this store (step S30). The busy day is specified based on the consignment management database 33, but is not limited thereto, and may be manually input in advance.
If it is determined that there is a busy day in this store (step S31 → Yes), the customer behavior predicting unit 21A repeats the processes in steps S32 to S40 for each customer in this store.
First, the customer behavior predicting unit 21A determines whether or not the finished product relating to the customer exists in the store (step S33). If there is no finished product in the store (step S33 → No), the process proceeds to step S40, and the next customer process is repeated. In this way, when there is no finished product in the store, the customer's probability of visiting the store is extremely low, so this customer is excluded from the message transmission target.
If the finished product is present in the store (step S33 → Yes), the customer behavior predicting unit 21A calculates the store visit distribution for the elapsed date after the finish date based on the customer behavior history database 31 (step S34). The store visit probability and the store visit date and time are predicted (step S35).
If the store visit status determination unit 23 determines that the customer will visit the store on a busy day based on the store visit date and time and the busy day of the store (step S36 → Yes), the stopable store predicting unit 25 may make a stopable store prediction process ( 17) is executed (step S37). Further, the non-storey store estimation unit 26 identifies a non-storey store among these stores that can be visited (step S38), and the alternative store notification unit 27 transmits an email urging the customer to visit the non-storey store (step S39). ). When the customer approves the visit to the secluded store, the store transfers the finished product to the secluded store.
The alternative store notification unit 27 may transmit a message urging the smart phone 7 to visit a quiet store when the customer touches the ticket gate 4 or the vending machine 5 with the smart phone 7.
In step S40, the customer behavior predicting unit 21A ends the notification process when the process is completed for each customer in the store.

FIG. 17 is a flowchart of the store-possible store prediction process according to the second embodiment. When processing step S37 of FIG. 16, this stop-at-home store prediction process is called.
The store-predictable store prediction unit 25 predicts the commuting route of the customer based on the customer behavior history database 31 (step S50), and specifies a store within a predetermined distance from each station of the predicted commuting route (step S51). . In addition, in step S52, the store-predictable store prediction unit 25 registers the specified store that can be visited in a column (not shown) of the customer database 32, returns to the notification process of FIG. 16, and executes the process of step S38.

  In the second embodiment, the date of receipt is predicted from the customer visit history, and another store where the customer is likely to stop is predicted from the customer's behavior history, the customer's home location information, and the store location information. . If the predicted date for the customer to pick up is a busy day at this store, the customer is urged to pick up at the alternative store. As a result, customer visits can be leveled as a whole of the affiliated stores.

(Modification)
The present invention is not limited to the above-described embodiment, and can be modified without departing from the spirit of the present invention. For example, there are the following (a) to (f).
(A) When the customer behavior management server guides the customer to another store, the customer behavior management server may notify the customer of GPS (Global Positioning System) position information of the store composed of latitude and longitude. Thereby, the customer can easily go to a new store for the customer by the GPS navigation application of the smartphone 7.
(B) In the first embodiment, the store visit probability may be calculated according to the relationship between the number of days elapsed from the finishing date and the store visit probability.
(C) In the second embodiment, the customer's store visit probability is not limited to the calculation based on the number of days elapsed from the finish date, and may be calculated from the relationship between the day of the week and the store visit probability.
(D) The calculation of the store visit probability is not limited to the above. For example, the relationship between the type of cleaning product and the store visit probability, for example, “ski gloves receive late” and “shirt receives early” is added. May be.
(E) In the second embodiment, the customer's store visit probability may be calculated from the past behavior history of each individual customer. The customer is categorized into each group, and the past behavior history of these groups is calculated. It may be calculated. Here, customer categorization includes, for example, gender, age group, commuting to school, and the like. As a result, the number of samples becomes the number that can be statistically processed, and thus a significant transition probability can be calculated from a short-term action history.
(F) The business condition of the store managed by the customer in the present invention is not limited to the cleaning store, but it is preferable that the store requires two visits of order and receipt, such as a tailor, DPE (Development Printing Enlargement) store, It is not limited to this, and it may be a good business condition at a single visit.

The invention described in the scope of claims attached to the application of this application will be added below. The item numbers of the claims described in the appendix are as set forth in the claims attached to the application of this application.
[Appendix]
<Claim 1>
A customer behavior predicting means for predicting the customer's visit date and time based on the customer's visit history to the store and the customer's visiting location information;
Visit status determination means for determining whether the store visit date and time predicted by the customer behavior prediction means corresponds to the busy time of the store;
If it is determined that the visit date / time of the customer corresponds to a busy time by the store visit status determination means, an alternative date / time notification means for notifying the customer of information prompting a visit at another date / time;
A customer behavior management device comprising:
<Claim 2>
The customer behavior prediction means determines whether or not there is a finished product related to the customer when predicting the visit date and time of the customer, calculates a store visit probability according to the determined presence or absence of the finished product, and also visits the customer to the store Predict the date and time,
The customer behavior management device according to claim 1, wherein:
<Claim 3>
The customer behavior predicting means predicts the visit date and time of the customer from the latest location information of the customer, the stop location information of the customer, and the transition probability to the store based on the presence or absence of the finished product. , Calculate the store visit probability, and also predict the visit date and time of the customer,
The customer behavior management device according to claim 2, wherein:
<Claim 4>
A customer behavior prediction means for predicting the visit date and time of the customer based on the visit history of the customer to the store;
Visit status determination means for determining whether or not the store visit date and time predicted by the customer behavior prediction means corresponds to a busy day of the store;
Based on the location information of a plurality of stores, the customer's stop location information, and the location information of the customer's home, a store-predictable store prediction means for predicting a store where the customer can stop,
Among the stores that can be visited predicted by the store-predictable store predicting means,
If it is determined that the visit date and time of the customer is a busy day by the store visit status determination means, an alternative store notification means for notifying the customer of information that prompts the store to visit the quiet store estimated by the quiet store estimation means;
A customer behavior management device comprising:
<Claim 5>
The customer stop location information includes ticket pass information,
The customer behavior management device according to any one of claims 1 to 4, wherein the customer behavior management device is provided.
<Claim 6>
The customer stop information includes vending machine usage information,
The customer behavior management device according to any one of claims 1 to 4, wherein the customer behavior management device is provided.
<Claim 7>
A customer behavior prediction step for predicting the customer's visit date and time based on the customer's visit history to the store and the customer's stopover information;
The store visit status determining step for determining whether the store visit date and time predicted by the customer behavior predicting step corresponds to the busy time of the store;
If it is determined that the visit date / time of the customer corresponds to a busy time in the store visit status determination step, an alternative date / time notification step for notifying the customer of information prompting the store to visit another date / time;
A customer behavior management method for a customer behavior management device.
<Claim 8>
A customer behavior prediction step for predicting the visit date and time of the customer based on the visit history of the customer to the store;
The store visit status determining step for determining whether the store visit date and time predicted by the customer behavior predicting step corresponds to the busy time of the store;
Based on the location information of a plurality of stores, the customer's stop location information, and the location information of the customer's home, a store predictable store prediction step for predicting a store where the customer can stop,
Of the stores that can be visited predicted by the store-predictable store prediction step,
If it is determined that the visit date and time of the customer corresponds to a busy time in the store visit status determination step, an alternative store notification step for notifying the customer of information that prompts the store to visit the closed store estimated by the closed store estimation step;
A customer behavior management method for a customer behavior management device.

1,1A Cleaning system 2,2A Customer behavior management server (Customer behavior management device)
21, 21A Customer behavior prediction unit 22 Busy time estimation unit 23 Visit status determination unit 24 Alternative date and time notification unit 25 Stoppable store prediction unit 26 Leisure store estimation unit 27 Alternative store notification unit 28 Email transmission unit 29 Short-range wireless notification unit 31 Customer Action history database 32 Customer database 33 Custody management database 34 Location information database 35 Store database 4 Ticket gate 41 Near field communication unit 5 Vending machine 51 Near field communication unit 6 Store terminal 7 Smartphone 71 Near field communication unit 72 Wireless communication Unit 73 Notification unit 74 ID storage unit 9 Network

Claims (6)

A customer behavior predicting means for predicting the customer's visit date and time based on the customer's visit history to the store and the customer's visiting location information;
Visit status determination means for determining whether the store visit date and time predicted by the customer behavior prediction means corresponds to the busy time of the store;
If it is determined that the visit date / time of the customer corresponds to a busy time by the store visit status determination means, an alternative date / time notification means for notifying the customer of information prompting a visit at another date / time;
With
The customer behavior prediction means determines whether or not there is a finished product related to the customer when predicting the visit date and time of the customer, calculates a store visit probability according to the determined presence or absence of the finished product, and also visits the customer to the store Predict the date and time,
A customer behavior management device characterized by that. The customer behavior predicting means predicts the visit date and time of the customer from the latest location information of the customer, the stop location information of the customer, and the transition probability to the store based on the presence or absence of the finished product. , Calculate the store visit probability, and also predict the visit date and time of the customer,
The customer behavior management device according to claim 1, wherein: The customer stop location information includes ticket pass information,
The customer behavior management device according to claim 1 or 2, wherein The customer stop information includes vending machine usage information,
The customer behavior management device according to any one of claims 1 to 3, wherein the customer behavior management device is provided. A customer behavior management method executed by a customer behavior management device,
A customer behavior prediction step for predicting the customer's visit date and time based on the customer's visit history to the store and the customer's stopover information;
The store visit status determining step for determining whether the store visit date and time predicted by the customer behavior predicting step corresponds to the busy time of the store;
If it is determined that the visit date / time of the customer corresponds to a busy time in the store visit status determination step, an alternative date / time notification step for notifying the customer of information prompting the store to visit another date / time;
Including
The customer behavior prediction step determines the presence / absence of a finished product related to the customer in predicting the visit date and time of the customer, calculates a store visit probability according to the determined presence / absence of the finished product, and also visits the customer to the store Predict the date and time,
A customer behavior management method characterized by the above. Computer
A customer behavior prediction means for predicting the customer's visit date and time based on the customer's visit history to the store and the customer's stopover information;
Visit status determination means for determining whether or not the store visit date and time predicted by the customer behavior prediction means corresponds to the busy time of the store,
An alternative date and time notifying unit for notifying the customer of information prompting the customer to visit another date and time when the customer visit date and time is determined to be a busy time by the store visit status determining unit;
Function as
The customer behavior prediction means determines whether or not there is a finished product related to the customer when predicting the visit date and time of the customer, calculates a store visit probability according to the determined presence or absence of the finished product, and also visits the customer to the store Predict the date and time,
A program characterized by that.

Images