JP2013182538A - Purchase trend analysis system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a system capable of performing a detailed purchase trend analysis by acquiring data on a purchase trend.SOLUTION: A purchase trend analysis system 1 comprises an alliance terminal 2, an analyzer 3, an analysis terminal 4, and transmission means 5 and 6, and stores purchase data transferred from the alliance terminal. The analyzer 3 performs a purchase trend analysis by designated means, on the basis of analysis means designation data and condition data transferred from the analysis terminal 4, and outputs generated purchase trend analysis data to the analysis terminal 4 in a designated format.

Description

  The present invention relates to a purchase trend analysis system that analyzes customer purchase trends and outputs analysis results.

Conventionally, systems for analyzing customer purchasing behavior and utilizing it for marketing have been developed. For example, a customer merchandise management system (including POS) and a customer merchandise management system (cited reference 1) using a mobile terminal obtains an e-mail address from the customer's terminal and accesses a mobile site containing product information. It is a system that grasps the purchase behavior of customers before purchasing products from this access trend. The product inventory management and sales optimization system, the optimization system, the optimization method, and the optimization program (Cited document 2) are systems that reflect answers to customer questions in product inventory management. .
JP2007-241351 JP2009-205365A JP 2002-123755 A JP2010-252580

  Each of the systems described above is a system for analyzing purchase trends of customers, and a system for managing inventory based on survey results of customer purchase awareness. However, the analysis of customer purchase trends and purchase awareness surveys conducted in this system are performed uniformly, and it is not possible to analyze the attributes and characteristics of each customer in detail and use this in marketing. There wasn't. Further, the purchasing trend and purchasing consciousness in a specific company are presented as an analysis result of the purchasing behavior to the company, and it cannot be said that the system analyzes the purchasing trend across a plurality of stores.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a system capable of obtaining purchase data from a store and performing detailed purchase trend analysis that matches purchase data with customer attributes and the like. Furthermore, it aims at providing the system which unifies and analyzes the purchase behavior of the customer between different industries. In addition, an object of the present invention is to provide an inventory management system that recommends a suitable product using the purchase trend analysis result.

  In order to solve the above problems, a purchasing trend analysis system of the present invention is a system for generating purchasing trend analysis data by analyzing purchasing data of a member, and the system generates an alliance terminal and purchasing trend analysis data. Analysis device and analysis terminal for outputting purchase trend analysis data, transmission means for enabling data transmission between the alliance terminal and the analysis device, and transmission means for enabling data transmission between the analysis device and the analysis terminal The alliance terminal has means for inputting purchase data including a member ID and at least a product code assigned to each product / service and purchase price data, and the member ID and purchase data input by the input means. Means for transmitting to the analyzer, wherein the analyzer is configured to transmit the transmitted member ID and purchase data. Receiving means, means for associating the transmitted member ID with purchase data in a purchase history database, means for determining analysis means based on analysis means designation data transmitted from the analysis terminal, and Based on the condition data transmitted from the analysis terminal, data stored in the purchase history database, and a member master database in which member ID and member attribute data are stored in association with each other, product category code or product code and product / service Analyzing data stored in at least one of a product master database in which name data is associated and stored, and a store master database in which store ID and store data are associated and stored, and generates purchase trend analysis data The analysis terminal includes an analysis means. Means for inputting data and condition data, means for transmitting analysis means designation data and condition data to the analyzer, means for accessing the analyzer and viewing purchase trend analysis data and / or purchasing trends from the analyzer The analysis apparatus has means for downloading and browsing the analysis data, and the analysis device analyzes the data accumulated in the database based on the analysis means designation data and the condition data transmitted from the analysis terminal, and generates the analysis purchase trend Analysis data is output to the analysis terminal.

  In the purchase trend analysis system of the present invention, at least gender data and age data are accumulated in the member master database, and the analysis device is a purchase including a product category code or product code included in the condition data transmitted from the analysis terminal. Extract the member ID that holds the data, count the number of member IDs for each gender data group and age data group based on the member attribute data, gender data group unit and age data group unit for the product category code or product code It has an analysis means which generates purchase trend analysis data which shows the number of member IDs.

  The purchase history database stores purchase date and time data as purchase data, and the analysis device holds purchase data including the product category code or product code included in the condition data transmitted from the analysis terminal within the aggregation period. The number of member IDs to be collected is counted, and the number of member IDs that have the purchase data including the product category code or product code included in the condition data transmitted from the analysis terminal within the counting period is held within the counting period. It has an analysis means for generating purchase trend analysis data indicating a trial rate divided by the number of total member IDs.

  The purchase history database stores purchase date and time data as purchase data, and the analysis device stores purchase data including the product category code or product code included in the condition data transmitted from the analysis terminal within the aggregation period. Count the number of member IDs held, further count the number of member IDs holding the product category code or product data including the product code twice or more, and purchase data including the product category code or product code twice or more Characterized in that it has an analysis means for generating purchase trend analysis data indicating a repeat rate obtained by dividing the number of member IDs holding the product ID by the number of member IDs holding the product category code or the purchase data including the product code. .

  In the purchase history database, purchase unit data indicating the receipt unit at the time of purchase as purchase data is accumulated, and the analysis device has the same product category code or product code included in the condition data transmitted from the analysis terminal. The product category code included in the condition data transmitted from the analysis terminal or the number of conformances for each concurrent purchase product category code or concurrent purchase product code is checked while checking the concurrent purchase product category code or concurrent purchase product code purchased in the unit of concurrent purchase. It has an analysis means for generating purchase trend analysis data indicating a concurrent purchase rate obtained by dividing by the total number of conformity of the concurrent purchase product category code or the concurrent purchase product code purchased in the same unit as the product code. .

  The purchase history database stores the purchase unit data indicating the receipt unit at the time of purchase as purchase data, and the analysis device purchases the product category code or product code included in the condition data transmitted from the analysis terminal. A member who checks the member ID that holds the data and has purchase data that includes the purchased product category code or the purchased product code purchased in the same purchase unit as the product category code or the product code among the matched member IDs A period obtained by extracting an ID and dividing the number of member IDs holding the purchase data including the product category code or the purchase product code by the number of member IDs holding the purchase data including the product category code or the product code Having analysis means for generating purchase trend analysis data indicating the concurrent purchase rate And features.

  The purchase history database stores purchase unit data indicating the receipt unit at the time of purchase as purchase data, and the analysis device groups the simultaneous purchase rate or period purchase rate by gender and age, and attributes the membership data. It has the analysis means which produces | generates the purchase trend analysis data which consist of separate purchase rate data.

  In the purchase history database, purchase date / time data and purchase price data as purchase data are accumulated, and the analysis device is configured for each product period code or product code included in the condition data transmitted from the analysis terminal for each set period. Purchase count data, sales amount data, and the number of purchase member IDs are collected, and purchase trend analysis data is generated by arranging the product category code or product code in order of purchase count data, sales amount data, and purchase member ID number. It has an analysis means.

  In the purchase history database, purchase date / time data and purchase price data as purchase data are accumulated, and the analysis device is configured for each product period code or product code included in the condition data transmitted from the analysis terminal for each set period. Purchase trend analysis data that counts the number of purchases, sales amount data, and purchase member IDs for each product, and plots the product category code or purchase number data for the product code, sales amount data, and the number of purchase member IDs for each set period It has the analysis means to produce | generate. It is characterized by the above-mentioned.

  In the purchase history database, purchase date / time data and purchase price data as purchase data are accumulated, and the analysis device is configured for each product period code or product code included in the condition data transmitted from the analysis terminal for each set period. Purchase number data and sales amount data for the product category code or the purchase member ID that holds the purchase data containing the product code in units of the number of purchase number data for the product category code or the amount of sales amount data And analyzing means for generating purchase trend analysis data.

  The purchase history database stores the purchase date / time data and purchase price data as purchase data, and the number of purchase date / time data associated with the product category code or product code included in the condition data transmitted from the analysis terminal. From the analysis terminal, the purchase date and time data obtained by aggregating data, the number of times per aggregation period associated with the product category code or product code included in the condition data transmitted from the analysis terminal, and the analysis terminal Generate purchase price data by summing up the purchase price per aggregate period associated with the product category code or product code included in the condition data to be transmitted, and any of purchase date / time data, purchase count data, or purchase price data Generate list data of combinations with or on the horizontal axis or vertical axis It characterized by having a diffraction means.

  In the product master database, a purchase flag is attached to the product code corresponding to the product purchased by the store, and the analyzer has the highest repeat rate among the product codes that do not have the purchase flag of any store. It has an analysis means for generating a stock replacement recommendation data for extracting a predetermined number of codes and displaying a list of products recommended to be put in stock with the extracted product codes or product names associated with the product codes. And

  In the database of the analysis device, a purchase flag is attached to the product code corresponding to the product purchased by the store, and the analysis device is a product having a high purchase rate for the product code with the purchase flag of any store. Extract a predetermined number of product codes that do not have a purchase flag of any store from the code, and display a list of products recommended to be put in stock with the extracted product code or the product name associated with the product code It has the analysis means which produces | generates the stock replacement recommendation data to perform.

  In the database of the analysis device, a purchase flag is attached to the product code corresponding to the product purchased by the store, and the analysis device has the same characteristics as the characteristic data of the higher-order product code in the number of purchases, repeat rate or concurrent purchase rate As a product that recommends extracting a product code that does not have a purchase flag for any store from among product codes that include data, and putting the extracted product code or the product associated with the product code in stock It is characterized in that the recommended stock replacement data to be displayed is generated.

  Alliance terminals and analysis terminals are installed at multiple alliance companies, and purchase history data transmitted from stores of multiple alliance companies is stored in the purchase history database of the analysis equipment. It is characterized by analyzing purchase history data in the alliance and generating purchase trend analysis data.

  According to the purchase trend analysis system of the present invention, an alliance company can select a desired purchase trend analysis according to its business form and acquire a purchase trend analysis result. An analyst can analyze a member's purchasing trend in detail in response to a request from the alliance company, and provide the alliance company with information that visualizes the analysis result. In addition, by constructing the system as a system across a plurality of alliance companies, the purchase trend analysis of one company can be added to the result of the purchase trend analysis of another company. As a result, it is possible to guide more accurate marketing decision-making than to analyze purchasing trends in-house.

The block diagram of the purchase trend analysis system of this invention. The block diagram which shows that the purchasing trend analysis system of this invention is constructed over multiple alliances. The condition input screen of the 1st purchase trend analysis means of a purchase trend analysis system. The histogram showing the purchase trend analysis result of a 1st purchase trend analysis means. The table showing the purchase trend analysis result of the 1st purchase trend analysis means. The table showing the purchase trend analysis result of the 1st purchase trend analysis means. The condition input screen of the 2nd purchase trend analysis means of a purchase trend analysis system. (A) is a scatter diagram showing the purchase trend analysis result of the second purchase trend analysis means of the purchase trend analysis system. (B) is a list showing the purchase trend analysis results of the second purchase trend analysis means. The condition input screen of the 3rd purchase trend analysis means of a purchase trend analysis system. The list showing the purchase trend analysis result of the 3rd purchase trend analysis means. The condition input screen of the 4th purchase trend analysis means of a purchase trend analysis system. (A) is a histogram showing the purchase trend analysis result of the 4th purchase trend analysis means. (B)-(D) are the lists showing the purchase trend analysis result of the 4th purchase trend analysis means. (E) is a list showing the purchase trend analysis results of the fourth purchase trend analysis means. The condition input screen of the 5th purchase trend analysis means of a purchase trend analysis system. (A)-(B) are the lists showing the purchase trend analysis result of the 5th purchase trend analysis means. The condition input screen of the 6th purchase trend analysis means of a purchase trend analysis system. The histogram showing the purchase trend analysis result of the 6th purchase trend analysis means. The table showing the purchase trend analysis result of the 6th purchase trend analysis means. The table showing the purchase trend analysis result of the 6th purchase trend analysis means. The condition input screen of the 7th purchase trend analysis means of a purchase trend analysis system. The table showing the purchase trend analysis result of the 7th purchase trend analysis means. The condition input screen of the 8th purchase trend analysis means of a purchase trend analysis system. The list showing the purchase trend analysis result of the 8th purchase trend analysis means. Graph selection screen. The schematic diagram of the purchase trend analysis system of this invention.

(System configuration)
The purchase trend analysis system of the present invention will be described. As shown in FIGS. 1 and 26, the purchasing trend analysis system 1 of the present invention is installed in an alliance terminal (device group) 2 installed in an alliance store of a plurality of companies in the same industry or different industries, and in a general center. It comprises a server analysis device (device group) 3 and an analysis terminal 4 installed in the alliance company. Here, the POS terminal (sales point information management terminal) 23 constituting the alliance terminal 2, the real processing system server 21 (including CPU, RAM, etc.) and / or the batch processing system server 22 (including CPU, RAM, etc.) ) And databases 24 and 25. When constructing a system in a virtual store, the device constituting the alliance terminal 3 can be replaced with a member-owned computer (not shown). The analysis device 3 includes a point system server 31 that accumulates member IDs, product codes, purchase price data, and the like, an analysis system data server 32 (including CPU, RAM, etc.) that performs purchase trend analysis, and an analysis system WEB server 33. (Including CPU, RAM, etc.) and databases 35 and 36. The analysis terminal 4 is an analysis computer 41 including an input device such as a keyboard and an output device such as a monitor and a printer that can access the analysis device 3 to download or view data. The alliance terminal 2 and the analyzer 3 are connected via a dedicated line so that data communication is possible. The analysis terminal 4 and the analysis device 3 are connected by an internet line.

  The purchasing trend analysis system 1 according to the present embodiment can be constructed as a single system. However, as shown in FIG. 1, in order to give exchangeable points to products and services. Existing point systems can also be used. Therefore, in FIG. 1, the server that acquires the member ID and purchase data of the analysis device 3 is the point system server 31, and the database 35 is connected thereto. Here, the point system will be briefly described. The member presents a member ID to be given points at the time of purchase, and the member ID and the product code are read by the POS terminal 23, and the member ID, the product code, and the purchase are purchased. Purchase data including money amount data is transmitted to the point system server 31 in real time or periodically. The point system server 31 receives this and adds and subtracts points according to the purchase amount in a point table (not shown) for the member ID. Further, a recommendation system server 34 may be connected to the point system server 31, and the recommended product name stored in the recommendation database 36 is transmitted to a computer or portable terminal (not shown) held by the member, or the POS terminal 23. It is also possible to construct a recommendation system that prints out a receipt that is output from. By using the point system already constructed in this way, for example, purchasing trends can be analyzed even when purchasing with cash without using a credit card, and more detailed purchasing trends can be analyzed. .

(Alliance terminal)
An alliance terminal 2 is installed in a store of an alliance company that participates in the purchase trend analysis system of the present invention. The POS terminal 23 constituting the alliance terminal 2 has at least input means for reading a member ID and input means for reading a product code assigned to a product / service, and further provides price data and purchase date / time data as purchase data. Can be done. Examples of the input means include a numeric keypad, a magnetic reader, a barcode reader, a two-dimensional barcode reader, and an RFID tag reader.

  In the alliance terminal 2, there are a real-time processing method and a batch processing method as a method for processing data. When using real-time processing for transmitting data to the point system server 31 of the analyzer 3 for each purchase opportunity, the real-time processing system server 21 is connected to the POS terminal 23 of the alliance terminal 2. A database 24 is connected to the real-time processing system server 21 as necessary. The real processing system server 21 and the point system server 31 of the analysis device 3 are connected to the transmission means of both devices through a dedicated line 5 so that data communication is possible, and the member ID and the product code of the product to be purchased at the POS terminal 23. Is read, the purchase data including the member ID and at least the product code is transmitted to the point system server 31 of the analysis device 3.

  When using batch processing for transmitting data to the analysis device 3 for each set data amount or set time, a batch processing system server 22 is constructed in the POS terminal 23 of the alliance terminal 2. The batch processing system server 22 is connected to a database 25. For example, in the product master database, product codes or product category codes assigned to products and services, product names and product categories, and product-specific data, production Product master data associated with product attributes such as company data is accumulated. Moreover, you may build the point grant data regarding the purchase database which accumulate | stores purchase data as needed, and the grant point. Each time the member ID and the product code are read by the POS terminal 23, the member ID, purchase data, and point data are accumulated. When a certain amount of time or a certain amount of data is accumulated in the database, these data are transmitted from the batch processing system server 22 to the point system server 31 of the analyzer 3 via a dedicated line. Note that only one of the real-time processing system server 21 and the batch processing system server 22 may be provided, or they may be used together as shown in FIGS.

(Analysis equipment)
In the purchase trend analysis system 1 of the present embodiment, the analysis device 3 includes a point system unit that receives and stores data, and an analysis system unit that analyzes data. The analysis device 3 receives the data from the alliance terminal 3 and stores the point system server 31, the database 31 connected thereto, the analysis system data server 32 that analyzes the data stored in the database 31, and the purchase trend analysis data It comprises an analysis system WEB server 33 that outputs to the terminal 4. Further, a recommendation system server 34 and a database 36 connected thereto are provided as necessary. In the database 35, a purchase history database is constructed in which the member ID and purchase data transmitted from the alliance terminal 3 are stored at any time. In addition, a member master database that stores member attribute data and a store master database that stores store attribute data are constructed. In the member master database, data relating to member attributes such as the member's gender, age or place of residence, occupation, and preference are stored in association with the member ID. In the store master database, information related to store attributes such as the store location and the sales form is stored in association with the store ID. In addition, when a product master database is not provided on the alliance terminal 2 side, a product master database is also constructed in the database 35. These databases are stored so that the data can be collated with each other based on the condition data.

  The analysis system data server 32 analyzes the data accumulated in the database 35 by the determined purchase trend analysis method based on the program specified by the analysis means designation data and the condition data transmitted from the analysis terminal 4 and purchases the data. Trend analysis data is generated. Further, the purchase trend analysis data calculated by the analysis system data server 32 is processed into output data such as a graph, a scatter diagram, and a list based on the program and transmitted to the analysis system WEB server 33.

(Analysis terminal)
The analysis terminal 4 is connected to the analysis system WEB server 33 via a general-purpose line such as the Internet. The analysis terminal 4 can log in to the analysis system WEB server 33 with an account and a password. The analysis terminal 4 can select and input analysis means designation data for specifying a desired purchase trend analysis method, condition data corresponding thereto, and output format at the first time or each time. Further, output data transmitted from the analysis system WEB server 33 can be downloaded, or the output data of purchase trend analysis data can be browsed by accessing the analysis system WEB server server 33. Further, the output data can be displayed in a desired output format such as a graph, a scatter diagram, and a list by a browser or a spreadsheet program installed in the analysis terminal 4 in advance.

  In the embodiment shown in FIG. 1 and FIG. 26, a single analysis computer 41 serves both as input means for selecting and inputting analysis means designation data, condition data and output format, and means for outputting purchase trend analysis data. However, the functions of the input means and the output means may be assigned to separate analysis computers 41. For example, an analysis computer 41 having input means for selecting and inputting analysis means designation data, condition data, and output format is installed on the general center side, and an analysis computer 41 having output means is installed on the alliance company side. You can also.

(Purchasing data acquisition and storage operations)
A purchase data acquisition operation and a storage operation common to all purchase trend analysis means described below will be described. As a premise of system operation, the member who acquired the member ID provides data such as gender, age, and residence as member master data items, and data such as hobbies and income as optional provision information before using the system. The data relating to the member attributes is stored in the member master database in a state linked to the member ID and member master. Here, the member ID is an arbitrary number of digits (16 digits in this embodiment) assigned to each member, and is recorded or stored in a member card formed by the member or a portable terminal and can be read by the POS terminal. It has become.

  In addition, the product code assigned to various products and services that can be provided at alliance stores is linked to the product master related to data such as product category, product name, and production company as product master data items. Accumulated in the master database. Here, the product code is an arbitrary number of digits (12 digits in this embodiment), and the product is specified by this number. In addition, a specific digit of the product code is a product category code for specifying the product category (major category, medium category, minor category). Further, as necessary, a specific digit number is a production company code indicating a production company, a characteristic code indicating a characteristic of a product, or the like. The specific code is, for example, a specific word included in the product name (for example, “Premium”, “Super”, etc.), product rank (special grade, limited-time product, product being increased in quantity, etc.), or product package color (red , Blue, yellow) is a code indicating the characteristics of the product. Further, the store code assigned to each store of the alliance is stored in the store master database in a state linked to the store master related to the store location, sales form, and other data as the store master data items.

  When selling products / services at an alliance store, the input unit of the POS terminal 23 reads the member ID and also reads the product code of the purchased product / service. The input member ID, purchase data including product code, purchase price data, purchase date / time data, and unit purchase data are both processed by the real processing method or batch processing method and transmitted to the point system server 31. The point system server 31 associates the transmitted member ID with purchase data, and further stores the store code in the purchase history database in the point system server 31. The data is transmitted to the database 35 to the analysis system data server 32 at any time or at regular intervals, or the analysis system data server 32 is going to acquire data.

  On the other hand, the analysis terminal 4 selectively inputs analysis means designation data (histogram analysis, trial & repeat analysis, buy-and-buy matrix analysis, buy-and-buy analysis, ranking analysis, time series analysis, decyl analysis and RFM analysis) for purchase trend analysis. It has become. In addition, condition data corresponding to each analysis means is selected and input. The selected analysis means designation data and condition data are associated with the terminal number of the analysis terminal 4 and an account given for access, and transmitted to the analysis apparatus 3 side via the analysis system WEB server. The transmission of the analysis means designation data and the condition data may be set to be input / transmitted only at the first time and at the time of change, or may be set to be input / transmitted every time the user logs in from the analysis terminal 4. Then, the analysis system data server 32 starts the operation by a predetermined date and time set by the program or by the start input from the analysis terminal 4, and the analysis means of the selected analysis content type starts the operation, and the purchase trend analysis data Is generated and output to the analysis terminal 4.

(Purchase trend analysis operation)
Next, 8 patterns of purchase trend analysis means in the purchase trend analysis system 1 of the present invention will be described. The condition data input here and the generated purchase trend analysis data are merely examples.

(Histogram analysis)
Histogram analysis as the first purchase trend analysis means will be described with reference to FIGS. The histogram analysis is a purchase trend analysis regarding the member attribute for the purchased product category, is processed by the analysis system data server 32, and is output to the analysis terminal 4 via the analysis system WEB server. When analysis means designation data for designating histogram analysis is input from the input unit of the analysis terminal 4 and condition data is input based on the condition input screen (FIG. 3) displayed on the monitor, this data is stored in the analysis system. It is transmitted to the WEB server 33. This condition data input may be performed only for the first time or may be performed each time. In the embodiment shown in FIG. 3, the target member type (all), the target product (product code 12345), the target district (A district), the target day of the week (Monday-Friday), the target time zone (11: 00-15: 00) Display), and the display mode to be output is set.

  When an operation start signal is input by a program incorporated in advance or an input from the analysis terminal 4, the analysis system data server 32 starts analysis based on the set condition data. For example, among the accumulated data in the database 35, the member ID that holds the selected product code (12345) in the purchase data is matched, and the predetermined condition (target district, target day, target) included in the purchase data is included in the purchase data. Filter by time zone. Then, the number of adapted member IDs is determined based on the gender data (for example, male, female, three unknown groups) of the member data linked to the member ID, age data (for example, 1 year old or 10 year old) The purchase trend analysis data including the number of purchase member IDs by sex (gender) group and the number of member IDs by age group with respect to the product code is generated.

  The analysis system data server 32 processes the purchase trend analysis data into two histogram data for each sex by associating the number of purchasing members with the X axis and the age group with the Y axis (FIG. 4A). Further, the analysis system data server 32 transmits the histogram data to the analysis terminal 4 or allows the histogram data to be viewed by logging in from the analysis terminal 4. The analysis terminal 4 displays the histogram data transmitted from the analysis system WEB server 33 in the form of a histogram through a browser or spreadsheet software.

  Further, the analysis system data server 32 calculates the number of purchasing members by sex group and age group for the product category, the ratio data of the number of purchasing members to the total number of purchasing member IDs, and the age data associated with the purchasing member ID. You may make it produce | generate the list data consisting of the average age data (FIG.5 (B)). Moreover, you may make it produce | generate the list data of the number of purchasing member ID according to age (FIG.6 (C)). The analysis terminal 4 displays the list data transmitted from the analysis system WEB server 33 through a browser or spreadsheet software in a matrix format.

(Trial / repeat analysis)
The trial / repeat analysis as the second purchase trend analysis means will be described with reference to FIGS. When analysis means designation data for designating trial / repeat analysis is input from the input unit of the analysis terminal 4 and conditions for analysis are input based on a condition input screen (FIG. 7) displayed on the monitor, Is transmitted to the analysis system WEB server 33. This condition input may be performed only for the first time or may be performed each time. In the embodiment shown in FIG. 7, the target member type (all), the target product category (product major category: daily necessities, product middle category: pet goods, product minor category B), target district (Fukuoka Prefecture, Saga Prefecture) , Kumamoto), target days of the week (Saturday, Sunday, public holidays), target time zone (13: 00-18: 00 range), gender group (female), age group (20s, 30s, 40s), counting period (2012) 1st January to 31st January 2012), and the display mode (multiple graphs) to be output is set.

  When an operation start signal for trial behavior analysis is input, the analysis system data server 32 starts analysis. For example, a product code including a product category code (for example, a specific number is assigned to a specific digit) is extracted from the purchase data for the aggregation period in the database 35, and the product ID is the member ID in the purchase data. Are matched, and the member IDs are narrowed down according to predetermined conditions (target district, target day of the week, target time zone) included in the purchase data, and the number of member IDs is tabulated. Further, the total number of member IDs is divided by the total number of member IDs holding purchase data within the total period, and the trial rate for each product code is calculated to generate purchase trend analysis data.

  Further, when an operation start signal for repeat purchase trend analysis is input, the accumulated data in the database 35 is divided by a total period, and a category code (for example, a specific number is assigned to a specific digit) within the total period. Is extracted, the product code matches the member ID in the purchase data, and the number of purchasing member IDs is totaled. Further, a member ID having a product code including the selected category code in the purchase data twice or more is extracted, and the number of the purchasing member ID is totaled. The number of purchase member IDs whose product code is in the purchase history more than once is the number of member IDs whose product code is in the purchase data (that is, the number of purchase member IDs of product codes including the category code B within the counting period). Divide by to calculate the repeat rate and generate purchase trend analysis data.

  The analysis system data server 32 associates the calculated trial rate and repeat rate with the scatter diagram format of the analysis terminal 4, takes the trial rate on the X-axis and the repeat rate on the Y-axis, where trial rate x repeat for each product code Scatter chart data obtained by plotting the rate is generated (FIG. 8A). The analysis terminal 4 displays the scatter diagram data transmitted from the analysis system WEB server 33 through a browser or spreadsheet software. Furthermore, the analysis system data server 32 may generate product code list data for the trial rate and the repeat rate (FIG. 8B). In the embodiment, the analysis in the product category unit has been described. However, the analysis can be performed in the product code unit. By trial analysis, it is possible to know products with a high rate of purchase during the counting period. In addition, it is possible to know a product that has a high probability of being purchased again by a person who has purchased it once through repeat analysis.

(Buying matrix analysis)
The combined purchase matrix analysis as the third purchase trend analysis means will be described with reference to FIGS. The buy-and-buy matrix analysis is an analysis means for calculating a buy-through rate for each product / category and outputting purchase movement analysis data in a matrix format. As a premise, in the purchase history database, purchase unit data indicating receipt units (also referred to as basket units) at the time of purchase as purchase data is accumulated. When the purchase matrix analysis is designated from the input unit of the analysis terminal 4 and the condition to be analyzed is input based on the condition input screen (FIG. 9) displayed on the monitor, this condition is sent to the analysis system WEB server 33. Communicated. This condition input may be performed only for the first time or may be performed each time. In the embodiment shown in FIG. 9, the target member type (all), the target product category (product major classification: daily necessities, product middle classification: pet goods, product minor classification: categories A to E), the target district (Hokkaido, Aomori Prefecture, Akita Prefecture, Miyagi Prefecture, target days of the week (Saturday, Sunday, public holidays), target time zone (13: 00-15: 00 range), gender group (male), age group (30s), counting period (2011) From December 1 to January 31, 2012), the unit of aggregation (simultaneous purchase) is set.

  The first concurrent purchase analysis, which is a simultaneous purchase analysis that analyzes the concurrent purchase rate of a receipt unit (also referred to as a basket unit), is an analysis that outputs the concurrent purchase rate of other products with respect to an arbitrary product in a receipt unit. When the concurrent purchase trend analysis start signal is input, the analysis system data server 32 extracts a product category code or a product code (for example, a specific number is included in a specific digit of the product code) included in the condition data, and the product Matching the product category code or co-purchased product code purchased in the same unit of purchase that indicates that the receipt and basket are the same, for each product category or product, the product category code or co-purchased product Divide the number of matches for each code by the total number of matches of the product category code or the product code purchased in the same unit as the product category code or product code included in the condition data transmitted from the analysis terminal. Generate purchase trend analysis data showing the simultaneous purchase rate And output as list display data (FIG. 10).

  In addition, the period concurrent purchase analysis in which the concurrent purchase analysis for each member is performed as the second concurrent purchase matrix analysis is an analysis for outputting the presence / absence of an arbitrary member's concurrent purchase or the concurrent purchase rate. When the concurrent purchase trend analysis start signal is input, the analysis system data server 32 holds purchase data including a product category code or a product code (for example, a specific number is included in a specific digit of the product code) included in the condition data. The member IDs that have the purchase data including the product category code or the co-purchased product code purchased in the same unit as the product category code or the product code are extracted from the matched member IDs. And a period-by-period purchase rate obtained by dividing the number of member IDs holding the purchase data including the concurrent purchase product category code or the concurrent purchase product code by the number of member IDs holding the purchase data including the product category code or the product code. The purchase trend analysis data shown is generated and output as list display data. It should be noted that matching member attribute conditions and regional conditions may be further narrowed down by members master database items. The analysis terminal 4 displays the list data transmitted from the analysis system WEB server 33 through a browser or spreadsheet software in matrix form.

  Such a concurrent purchase matrix analysis shows a product with a high concurrent purchase rate for an arbitrary product. It is also possible to know the attributes of members who have a high purchase rate.

(Buy and buy analysis)
The concurrent purchase analysis as the fourth purchase trend analysis means will be described with reference to FIGS. The concurrent purchase analysis is performed using the data generated by the concurrent purchase matrix analysis. The concurrent purchase analysis is a means for analyzing a specified product category or product and another specified product category or product, or a combined purchase rate of all product categories or products. Similar to the concurrent purchase matrix analysis, this analysis can be performed by selecting the simultaneous concurrent purchase analysis and the period concurrent purchase analysis. When a condition to be analyzed is input from the input unit of the analysis terminal 4 based on a condition input screen (FIG. 11) displayed on the monitor, this condition is transmitted to the analysis system WEB server 33. This condition input may be performed only for the first time or may be performed each time. In the embodiment shown in FIG. 11, the target member type (all), the target product category (beer), the target product category (paper diapers), the target area (Tokyo, Kanagawa, Saitama, Chiba), the target day of the week (Saturday, Sunday, public holidays), target time zone (17: 00-22: 00), gender group (all), age group (all), aggregation period (no designation), aggregation unit (simultaneous purchase) is set .

  When the simultaneous purchase trend analysis start signal is input, the analysis system data server 32 extracts the product code including the product category code A (beer) from the accumulated data in the database 35 for the accumulated data in the aggregation period, The same purchase unit product category code B with the same purchase unit data such as a receipt or a basket is matched. For the matched product category, the number of customers (the number of receipts) of the corresponding product (the customer) and the source product is divided by the number of customers (the number of receipts) of the source product, and the concurrent purchase rate is calculated for each receipt. Generate purchase trend analysis data consisting of

  Explaining the period-by-period purchase analysis related to aggregation for each member. When a concurrent purchase trend analysis start signal is input, the purchase unit is the same as an arbitrary product code in the purchase history of a member ID in the accumulated data within the aggregation period. The product code is matched, and the concurrent purchase rate is calculated by dividing the number of concurrent purchase members of the corresponding product (buy-to party) and purchase source product by the number of purchase members of the purchase source product in the aggregation period, and from the concurrent purchase rate for each member ID Generate purchase trend analysis data.

  As an example of output data, a graph of the concurrent purchase rate for any product by sex (FIG. 12 (A)), a list of presence / absence of concurrent purchase by member attribute (FIG. 13 (B)), for any product There is a list of the number of concurrent purchases, the number of concurrent purchase member IDs, and the concurrent purchase rate (FIG. 13D). In the embodiment, the analysis of the product category unit has been described, but the analysis can also be performed in the product code unit. Then, the list data transmitted from the analysis system WEB server 33 through a browser or spreadsheet software is displayed in a matrix.

(Ranking analysis)
Ranking analysis as the fifth purchasing trend analysis means will be described with reference to FIGS. Ranking analysis is a function that displays a list of product codes, the number of sales by product category code, the sales amount, and the number of purchasing members arranged in descending order. When the time series analysis start signal is input, the analysis system data server 32 accesses the database and corresponds to the number of member IDs including the selected product category code or product code in the purchase data and the product category code or product code. The purchase trend analysis data is generated by collecting the number of sales to be made and the sales amount of the product category code or the product code. Then, ranking list data is generated in descending order of the number of member IDs, the number of sales, and the number of sales amounts for the product category code or the product code. The analysis terminal 4 displays the list data transmitted from the analysis system WEB server 33 through a browser or spreadsheet software in matrix form.

(Time series analysis)
A time series analysis as a sixth purchasing behavior analysis means will be described with reference to FIGS. The time series analysis is a function of analyzing the daily or weekly changes in the number of sales by product code, product category code, sales amount, and number of customers. When the time series analysis start signal is input, the analysis system data server 32 extracts a product code including the selected product category code, and the product category code or the number of member IDs in which the product code is in purchase data and the product category code. Or a line graph that plots the number of sales and the sales amount for a product code and plots them in units of set periods (eg, hourly, daily, monthly, weekly, hourly) to show the transition of these three elements. Data (FIG. 18) is generated. Further, ranking list data is generated in descending order of the number of member IDs, the number of sales, and the amount of sales for a certain period for the product code (FIGS. 19 and 20). The analysis terminal 4 displays the line graph data transmitted from the analysis system WEB server 33 through a browser or spreadsheet software, and displays the list data in a matrix.

(Decyl analysis)
A decyl analysis as a seventh purchase trend analysis means will be described with reference to FIGS. The decyl analysis is a decyl analysis function that ranks member IDs according to superiority or inferiority of purchase behavior. When condition data such as the target member, counting period, region, and time zone are input and transmitted (FIG. 21) and a decyl analysis start signal is input, the analysis system data server 32 accesses the database and purchases a member ID. Decyl analysis list data is generated by ranking the member IDs by element, and by the number of member IDs, grouped by rank according to a predetermined set rank for each of the three elements of the amount, the number of visits, and the unit price of customers (see FIG. 22). In addition, grouped member IDs and product codes associated in the purchase history database may be ranked for each repeat rate and number of purchases. The analysis terminal 4 displays the list data transmitted from the analysis system WEB server 33 through a browser or spreadsheet software in matrix form.

(RFM analysis)
The RFM analysis as the eighth purchase behavior analysis means will be described with reference to FIGS. The RFM analysis is an RFM analysis function that combines the recent purchase (Recency), how often (Frequency), and how much money (Monetary). The Recency analysis function is a function for extracting the latest purchase date in a predetermined period for each member ID. The frequency analysis function is a function for extracting the number of purchases within a predetermined period for each member ID. The Monetary analysis function is a function for extracting the purchase amount in a predetermined period for each member ID. When the RFM analysis start signal is input, R program F analysis, F test M analysis, M test M analysis, R test M analysis, and F test F analysis are performed by the program. As an example of F-F analysis, a designated period of F1 (for example, the month before last) and a designated period of F2 (for example, the previous month) are selected, and the number of member IDs for each number of purchases during this period is tabulated. List data representing the number of member IDs is generated for each purchase time group of two purchasers, four purchasers, six purchasers, eight purchasers, ten purchasers and more. Further, the member ID may be associated with the purchase time group, and the data may be further aggregated for each member data associated with the member ID.

  Next, format selection from the analysis terminal will be described based on FIG. In order to make the purchase trend analysis result data shown in the first to eighth graphs into a comparative graph and a plurality of graphs, the analysis terminal can select the output format of the purchase trend analysis result data. According to the example of the present invention, the purchase trend analysis result data can be selected from single graph display and multiple graph display for each product, area and day of the week / time period. The selected format is stored in the analysis computer, and when accessed from the analysis computer, a graph or a list of the selected format is displayed on the analysis computer.

(Inventory management system)
An inventory management system that can be optionally added to the purchase trend analysis system of this embodiment will be described. As a premise of the inventory management system, each store selects a product purchased in the store for the product master, and sets a purchase flag for each store in the product code.

  A first form of the inventory management system is an inventory management system using repeat rate data of trial / repeat analysis. It is recommended that the analysis system data server 32 extract a predetermined number of product codes having a higher repeat rate from the product codes for which the store's purchase flag is not set, and put the product names associated with the product codes in stock. The product name to be displayed is displayed in the list, and the inventory replacement recommendation list data is generated. Further, as a function to be added, a predetermined number of product codes lower in the repeat rate are extracted from the product codes for which any store has a purchase flag, and the repeat rate is selected from product codes for which the store has no purchase flag. Equivalent number of high-order product codes are extracted, and the product name linked to the product code with the lowest repeat rate is displayed in the list as the recommended product name to be removed from stock, and the purchase flag is not set Of the product codes, the product name linked to the product code with the highest repeat rate is displayed in the list as the product name recommended to be put in stock, and the inventory replacement recommendation list data is generated. This stock replacement recommendation list data can be viewed through a browser from an analysis computer installed in an alliance company through an analysis system WEB server. As a result, the alliance company can purchase products with a higher repeat rate in its own company and other alliance companies for store inventory management.

  Further, the second form is an inventory management system using concurrent purchase rate data of concurrent purchase analysis. This system extracts the product code for which any store has a purchase flag, and sets the purchase rate for the product code for which the purchase flag is set among the product codes for which the purchase flag is not set to the product code with the higher purchase rate. The associated product name is displayed in the list as a product name that is recommended to be put in stock, and inventory replacement recommendation list data is generated. This stock replacement recommendation list data can be viewed through a browser from an analysis computer installed in an alliance company through an analysis system WEB server. As a result, alliance companies purchase products with a high buy-in ratio for arbitrary products at their own companies and other alliance companies, or display products with a high simultaneous purchase rate next to any products for store inventory management. be able to.

  Furthermore, the third form is a product linked to a product code having the same characteristic code as that of the product code higher in the repeat rate of the trial / repeat analysis or the product code higher in the combined purchase rate. The product name linked to the code is displayed in the list as the product name recommended to be put in stock, and the inventory replacement recommended list data is generated. If a specific example is shown, about the goods which contain the word "premium" in a goods name, the goods code containing a specific number will be given to the specific digits of the goods category code linked | related to this, and it will accumulate | store in a goods master database. If the sales of “Premium Beer” are good, there is an example in which “Premium Chocolate” having a characteristic code including the same word “Premium” is displayed as a recommended product name. This stock replacement recommendation list data can be viewed through a browser from an analysis computer installed in an alliance company through an analysis system WEB server. As a result, the alliance company can purchase merchandise having merchandise characteristics favored by members, such as high-end products, limited-time products, and increased products, for store inventory management.

  These inventory management systems performed by the purchasing trend analysis system server are recommended by decyl analysis, time series analysis, and purchasing trend analysis for each area, recommended product names by member rank, and recommendation by time series. It is also possible to generate data for product names and recommended product names by area.

  According to the purchase trend analysis system 1, an alliance company can select a desired analysis type, analyze a member's purchase trend in detail, and provide the alliance company as information visualizing the analysis result. . In addition, by constructing the system as a system across a plurality of alliance companies, the purchase trend analysis of one company can be added to the result of the purchase trend analysis of another company. As a result, more accurate marketing decision making can be derived than purchasing trend analysis using data only for the company. The purchase trend analysis system 1 can be constructed by using an existing point grant system. Further, it is possible to construct an inventory management system as described above, and it is also possible to propose inventory management in accordance with customer purchase behavior based on purchase trend analysis data.

  In the present embodiment, the purchase trend analysis of one alliance company has been described as an example. However, as shown in FIG. 2, the purchase trend analysis system 1 can be constructed across a plurality of alliance companies.

DESCRIPTION OF SYMBOLS 1 Purchase trend analysis system 2 Alliance terminal 21 Real processing system server 22 Batch processing system server 23 POS terminal 24,25 Database 3 Analyzer 31 Point system server 32 Analysis system data server 33 Analysis system WEB server 34 Recommendation system server 35,36 database 4 Analysis terminal 41 Computer for analysis

Claims (15)

A system for analyzing purchase data of members to generate purchase trend analysis data, the system comprising an alliance terminal, an analysis device for generating purchase trend analysis data, an analysis terminal for outputting purchase trend analysis data, and the alliance A communication means that enables data transmission between a terminal and the analysis device, and a transmission means that enables data transmission between the analysis device and the analysis terminal;
The alliance terminal has means for inputting purchase data including a member ID and at least a product code assigned to each product / service and purchase price data, and the analysis device receives the member ID and purchase data input by the input means. Having means to communicate to
The analysis device includes means for receiving the transmitted member ID and purchase data, means for storing the transmitted member ID and purchase data in association with each other in a purchase history database, and analysis means transmitted from the analysis terminal. A means for determining an analysis means based on designated data, and a data stored in the purchase history database, and a member ID and member attribute data are stored in association with each other based on the condition data transmitted from the analysis terminal. Stored in at least one of a member master database, a product master database in which product category codes or product codes are stored in association with product / service name data, and a store master database in which store IDs are stored in association with store data. Analyzing generated data and generating purchase trend analysis data And it has a stage,
The analysis terminal includes means for inputting analysis means designation data and condition data, means for transmitting the analysis means designation data and condition data to the analysis device, and means for accessing the analysis device and browsing purchase trend analysis data And / or a means for downloading and browsing purchase trend analysis data from the analysis device,
The analysis device analyzes data accumulated in a database based on analysis means designation data and condition data transmitted from the analysis terminal, and outputs the generated purchase trend analysis data to the analysis terminal. Purchasing trend analysis system.   At least gender data and age data are accumulated in the member master database, and the analysis device extracts the member ID holding the purchase data including the product category code or the product code included in the condition data transmitted from the analysis terminal. , Purchasing trend indicating the number of member IDs for each sex data group and age data group for each product category code or product code by counting the number of member IDs for each sex data group and age data group based on member attribute data The purchase trend analysis system according to claim 1, further comprising analysis means for generating analysis data.   The purchase history database stores purchase date and time data as purchase data, and the analysis device holds purchase data including the product category code or product code included in the condition data transmitted from the analysis terminal within the aggregation period. The number of member IDs to be collected is counted, and the number of member IDs that have the purchase data including the product category code or product code included in the condition data transmitted from the analysis terminal within the counting period is held within the counting period. The purchase trend analysis system according to claim 1 or 2, further comprising analysis means for generating purchase trend analysis data indicating a trial rate divided by the number of total member IDs.   The purchase history database stores purchase date and time data as purchase data, and the analysis device stores purchase data including the product category code or product code included in the condition data transmitted from the analysis terminal within the aggregation period. Count the number of member IDs held, further count the number of member IDs holding the product category code or product data including the product code twice or more, and purchase data including the product category code or product code twice or more Characterized in that it has an analysis means for generating purchase trend analysis data indicating a repeat rate obtained by dividing the number of member IDs holding the product ID by the number of member IDs holding the product category code or the purchase data including the product code. The purchase trend analysis system according to any one of claims 1 to 3.   In the purchase history database, purchase unit data indicating the receipt unit at the time of purchase as purchase data is accumulated, and the analysis device has the same product category code or product code included in the condition data transmitted from the analysis terminal. The product category code included in the condition data transmitted from the analysis terminal or the number of conformances for each concurrent purchase product category code or concurrent purchase product code is checked while checking the concurrent purchase product category code or concurrent purchase product code purchased in the unit of concurrent purchase. It has an analysis means for generating purchase trend analysis data indicating a concurrent purchase rate obtained by dividing by the total number of conformity of the concurrent purchase product category code or the concurrent purchase product code purchased in the same unit as the product code. The purchase trend analysis system according to any one of claims 1 to 4.   The purchase history database stores the purchase unit data indicating the receipt unit at the time of purchase as purchase data, and the analysis device purchases the product category code or product code included in the condition data transmitted from the analysis terminal. A member who checks the member ID that holds the data and has purchase data that includes the purchased product category code or the purchased product code purchased in the same purchase unit as the product category code or the product code among the matched member IDs A period obtained by extracting an ID and dividing the number of member IDs holding the purchase data including the product category code or the purchase product code by the number of member IDs holding the purchase data including the product category code or the product code Having analysis means for generating purchase trend analysis data indicating the concurrent purchase rate Purchasing trends analysis system according to any one of claims 1 to 5, characterized.   The purchase history database stores purchase unit data indicating the receipt unit at the time of purchase as purchase data, and the analysis device groups the simultaneous purchase rate or period purchase rate by gender and age, and attributes the membership data. 7. The purchase trend analysis system according to claim 5, further comprising an analysis unit that generates purchase trend analysis data including separate purchase rate data.   In the purchase history database, purchase date / time data and purchase price data as purchase data are accumulated, and the analysis device is configured for each product period code or product code included in the condition data transmitted from the analysis terminal for each set period. Purchase count data, sales amount data, and the number of purchase member IDs are collected, and purchase trend analysis data is generated by arranging the product category code or product code in order of purchase count data, sales amount data, and purchase member ID number. The purchase trend analysis system according to any one of claims 1 to 7, further comprising an analysis unit.   In the purchase history database, purchase date / time data and purchase price data as purchase data are accumulated, and the analysis device is configured for each product period code or product code included in the condition data transmitted from the analysis terminal for each set period. Purchase trend analysis data that counts the number of purchases, sales amount data, and purchase member IDs for each product, and plots the product category code or purchase number data for the product code, sales amount data, and the number of purchase member IDs for each set period 9. The purchasing trend analysis system according to claim 1, further comprising an analyzing means for generating.   In the purchase history database, purchase date / time data and purchase price data as purchase data are accumulated, and the analysis device is configured for each product period code or product code included in the condition data transmitted from the analysis terminal for each set period. Purchase number data and sales amount data for the product category code or the purchase member ID that holds the purchase data containing the product code in units of the number of purchase number data for the product category code or the amount of sales amount data The purchase trend analysis system according to any one of claims 1 to 9, further comprising analysis means for generating purchase trend analysis data.   The purchase history database stores the purchase date / time data and purchase price data as purchase data, and the number of purchase date / time data associated with the product category code or product code included in the condition data transmitted from the analysis terminal. From the analysis terminal, the purchase date and time data obtained by aggregating data, the number of times per aggregation period associated with the product category code or product code included in the condition data transmitted from the analysis terminal, and the analysis terminal Generate purchase price data by summing up the purchase price per aggregate period associated with the product category code or product code included in the condition data to be transmitted, and any of purchase date / time data, purchase count data, or purchase price data Generate list data of combinations with or on the horizontal axis or vertical axis Purchasing trends analysis system according to any one of claims 1 to 10 characterized by having a diffraction means.   In the product master database, a purchase flag is attached to the product code corresponding to the product purchased by the store, and the analyzer has the highest repeat rate among the product codes that do not have the purchase flag of any store. It has an analysis means for generating a stock replacement recommendation data for extracting a predetermined number of codes and displaying a list of products recommended to be put in stock with the extracted product codes or product names associated with the product codes. The purchase trend analysis system according to claim 4.   In the database of the analysis device, a purchase flag is attached to the product code corresponding to the product purchased by the store, and the analysis device is a product having a high purchase rate for the product code with the purchase flag of any store. Extract a predetermined number of product codes that do not have a purchase flag of any store from the code, and display a list of products recommended to be put in stock with the extracted product code or the product name associated with the product code The purchase trend analysis system according to any one of claims 5 to 7, further comprising analysis means for generating recommended stock replacement data.   In the database of the analysis device, a purchase flag is attached to the product code corresponding to the product purchased by the store, and the analysis device has the same characteristics as the characteristic data of the higher-order product code in the number of purchases, repeat rate or concurrent purchase rate As a product that recommends extracting a product code that does not have a purchase flag for any store from among product codes that include data, and putting the extracted product code or the product associated with the product code in stock 8. The purchase trend analysis system according to claim 4, wherein the inventory replacement recommendation data to be displayed in a list is generated.   Alliance terminals and analysis terminals are installed at multiple alliance companies, and purchase history data transmitted from stores of multiple alliance companies is stored in the purchase history database of the analysis equipment. 15. The purchase trend analysis system according to claim 1, wherein purchase trend analysis data is generated by analyzing purchase history data in the alliance.