JP7004504B2 - Analysis equipment - Google Patents

Description

The present invention relates to a data processing technique, and more particularly to a technique for analyzing data.

With the spread of computers, an increasing number of companies are accumulating sales record information showing the record of selling products to customers in databases and the like. Some companies utilize this sales performance information to create information for supporting the sales activities of users (hereinafter referred to as "sales activity support information") by a computer system. For example, in Patent Document 1, products are classified from various viewpoints, and sales activity support information is created by associating this classification information with sales performance information.

As a method different from Patent Document 1, sales activity support information may be created by associating the attributes of the customer who purchased the product with the sales performance information. In this case, the attribute information of the customer registered at the time of creating the card may be used for the customer who purchased the product using the point card or the credit card.

Japanese Unexamined Patent Publication No. 2001-167203

However, the purchaser who actually purchases the product may be different from the consumer who actually consumes the product. For example, the consumer of a product purchased by a housewife in a household may be the husband of that household. Therefore, the sales activity support information created based on the attributes of the purchaser tends to not reflect the attributes of the consumer, and may not be useful information for the seller of the product.

Based on this background, the present inventor has a group of areas (hereinafter, "area cluster" or "area type") in which the attributes of the inhabitants are similar with respect to the area in which the purchaser resides (hereinafter referred to as "area"). Also called.). That is, sales that are useful to the seller of the product by analyzing the sales performance information from the macro perspective based on the inhabitant attribute of the area cluster in which the purchaser resides, not from the micro perspective based on the attribute of the purchaser of the product. I thought that I could create activity support information.

The present invention has been made in view of the above problems, and one object is to effectively support the sale of goods or services.

In order to solve the above problems, the analyzer of a certain aspect of the present invention is a first acquisition unit that acquires area type data in which a plurality of geographically divided areas are grouped by areas having similar resident attributes. The second acquisition unit that acquires the sales performance data including the data of the area where the customer lives, and the sales performance data of the product or service for each area type. It is provided with an analysis unit that aggregates data and generates analysis information about products or services based on the aggregation result for each area type.

It should be noted that any combination of the above components and the conversion of the expression of the present invention between a method, a program, a recording medium in which the program is stored, and the like are also effective as aspects of the present invention.

According to the present invention, it is possible to effectively support the sale of goods or services.

It is a figure which shows the structure of the sales information analysis system in this embodiment. It is a figure which shows the data structure of the sales promotion performance information. It is a figure which shows the data structure of the product sales information. It is a figure which shows the data structure of the store customer information. It is a figure which shows the data structure of the store sales information. It is a block diagram which shows the functional structure of the sales information analysis apparatus of FIG. It is a figure which shows the data structure of the area cluster information. It is a figure which shows the data structure of the sales promotion effect information. It is a figure which shows the data structure of the product category information. It is a figure which shows the data structure of the product index value information. It is a figure which shows the area sales volume acquired by the area sales volume acquisition part. It is a figure which shows the cluster sales volume calculated by the cluster sales volume calculation part. It is a block diagram which shows the functional structure of the sales promotion effect analysis part. It is a figure which shows the pre-promotion sales volume and the post-promotion sales volume acquired by the sales promotion sales volume acquisition department. It is a figure which shows the pre-promotion sales volume and the post-promotion sales volume aggregated by the sales promotion sales volume totaling unit. It is a figure which shows the sales volume increase rate in each area cluster of the product to be promoted by a sales promotion measure. It is a block diagram which shows the functional structure of the sales volume correction part. It is a figure which shows the data structure of the sales promotion performance information. It is a figure which shows the data structure of the product sales information. It is a figure which shows the pre-promotion sales volume and the post-promotion sales volume acquired by the sales promotion sales volume acquisition department. It is a figure which shows the sales volume after promotion corrected by the sales promotion sales volume correction part. It is a figure which shows the cluster sales volume after correction by a sales volume correction part. It is a block diagram which shows the functional structure of the product share analysis part. It is a figure which shows the category sales volume calculated by the category sales volume calculation part. It is a figure which shows the category sales volume calculated by the category sales volume calculation part. It is a figure which shows the occupancy rate of each product about the product category "coffee". It is a figure which shows the occupancy rate of each product about the product category "coffee". It is a block diagram which shows the functional structure of a purchase factor analysis part. It is a figure which exemplifies the relationship between the inhabitant index value and the cluster sales volume. It is a block diagram which shows the functional structure of a store customer group analysis part. It is a figure which shows the cluster customer ratio of "Kiba store" of FIG. It is a figure which shows the cluster customer ratio of "Kasai store" of FIG. It is a figure which shows the cluster customer ratio of "Funabashi store" of FIG. It is a block diagram which shows the functional structure of the sales condition influence analysis part. It is a figure which shows the sales volume before change and the sales volume after change acquired by the comparative sales volume acquisition part. It is a figure which shows the sales condition sensitivity rate calculated by the sensitivity rate evaluation unit. It is a block diagram which shows the functional structure of the repeat rate analysis part. It is a figure which shows the cluster repeat number calculated by the cluster repeat number calculation part. It is a figure which shows the repeat rate in each area cluster calculated by the repeat rate calculation part. It is a figure which exemplifies the correspondence between the repeat rate and the unit customer sales volume. It is a block diagram which shows the functional structure of the product appeal degree analysis part. Shows the cluster sales volume of the products to be analyzed acquired by the analysis information acquisition unit. It is a figure which shows the cluster index value calculated by the cluster index value calculation part. It is a block diagram which shows the functional structure of the product appeal degree correction part. It is a figure which shows the transition of the cluster sales volume acquired by the sales volume trend calculation part. It is a figure which shows the cluster index value and the rate of change of a cluster sales volume. It is a figure which shows the product index value before correction and the product index value after correction. It is a block diagram which shows the functional structure of a store sales volume prediction part. Shows the estimated sales volume at each store. It is a figure which shows the actual result and the forecast of the sales volume in each store in association with each other. It is a figure which shows the structure of the IDPOS analysis system of 2nd Embodiment. It is a figure which shows the example of IDPOS data. It is a figure which shows the example of the area data. It is a block diagram which shows the functional structure of the analyzer of FIG. 41. It is a figure which shows the example of the area type addition IDPOS data. It is a figure which shows the example of the aggregation result data by the aggregation part. It is a figure which shows the example of the product profile information. It is a figure which shows the example of the product profile information. It is a figure which shows the example of the sales potential information. It is a figure which shows the example of the sales potential information. It is a flowchart which shows the operation of an analyzer.

[First Embodiment]
Before explaining the configuration of the first embodiment in detail, first, an outline will be given.
Now that competition between companies is intensifying, product sellers are required to make appropriate decisions in sales activities. In order to support this decision-making, it is important to utilize sales activity support information. Here, an example of sales activity support information is shown below.

1. 1. Effect information of sales promotion measures. This information shows the effects of sales promotion measures, including advertisements and discounts at stores, and supports the decision-making of sellers regarding the implementation of sales promotion measures.
2. 2. Product share information. Information that indicates the occupancy rate of competing products in the same product category and supports the seller's decision-making regarding product purchases, shelving allocation, etc.
3. 3. Purchasing factor information. Information that shows the influence of customer attributes on product purchases, in other words, the correlation between customer attributes and product sales volume, and supports the seller's decision-making regarding product purchases, shelving allocation, and the like.
4. Store customer base information. This information indicates the structure of the customer base of the store and supports the decision-making of the seller according to the structure of the customer base.

5. Sales condition impact information. Information that shows the influence of the sales conditions of a product on the sales of the product and supports the seller's decision-making regarding the setting of the sales conditions of the product.
6. Repeat rate information. This information indicates the rate at which the same customer repeatedly purchases the same product, and supports the seller's decision-making regarding product purchases, shelving allocation, and the like.
7. Product appeal information. It is information to show how much the product of a certain store is supported by the customer and to support the seller's decision-making regarding product purchasing, shelving allocation, etc.
8. Store sales volume forecast information. It is information to show the prediction of how much the product of a certain store will be purchased by the customer and to support the seller's decision-making regarding product purchase, shelving allocation, etc.

It is desirable that these sales activity support information be created by associating the attributes of the consumer who actually consumes the product with the sales performance of the product. In other words, the purchaser who actually purchased the product may be merely a person who executed the consumer's purchase intention on behalf of the purchaser, and it is useful even if the attributes of the purchaser are associated with the sales performance of the product. It may not be useful sales activity support information.

For example, if the product is a men's hair restorer, the purchaser is often the housewife of the household, while the consumer is often the husband of the household. In this case, even if the sales activity support information is created based on the attributes of the housewife who is the purchaser, the sales activity support information is targeted at the housewife. Therefore, sales activity support information that does not reflect the attributes of consumers may not be useful information for the seller.

There are typically two methods of traditionally acquiring customer attributes. The first method is a method of using the attribute information of the customer registered at the time of card creation for the customer who purchased the product using the point card or the credit card. This first method will be hereinafter referred to as "FSP (Frequent Shoppers Program) method". The second method is a method of using the attribute information of a small number of monitor customers and the attribute information of customers entered in the questionnaire (hereinafter, referred to as "monitor customer method").

However, in the FSP method, the attribute information that can be acquired is the attribute information of the purchaser, and the attribute information of the consumer cannot be acquired. Further, the customer attribute information registered at the time of card creation is only past information, and may be different from the attribute information at the time when the product is purchased. In the case of the monitor customer method, the attributes of consumers can be acquired by selecting monitor customers and respondents to the questionnaire. However, the number of monitor customers and the number of questionnaires are generally small, and when the attributes of customers are determined from the results of a small number of monitor customers and questionnaires, the attributes corresponding to only a small number of people are regarded as the attributes of the entire consumer. In some cases, the accuracy of sales activity support information was low. In addition, the monitor customer method has a problem that it generally takes time because it requires manual labor.

The first embodiment proposes a technique for creating sales activity support information by associating the inhabitant attributes of each area cluster in which the purchaser of the product resides with the sales performance. An area cluster is a grouping of multiple areas with similar inhabitant attributes, typically areas on a town-chome level map, and each area cluster is associated with inhabitant attributes common to these multiple areas. Has been done. Therefore, it is possible to associate the consumer image in the area cluster with the sales performance, and to realize automatic analysis without human intervention. As a result, it is possible to create useful sales activity support information for the seller of the product in a short time.

The above-mentioned sales activity support information is also useful information for product development by product manufacturers and measurement of advertising effectiveness. For example, product share information and purchasing factor information indicating what kind of consumer is purchasing a product allows the manufacturer of the product to match the target customer assumed at the time of product development with the actual consumer of the product. You can check if it is. In addition, based on the effect information of sales promotion measures, product manufacturers can confirm whether product advertisements and corporate advertisements on TV etc. have an impact on consumers who expect a reaction, and the cost of enormous advertising costs can be reduced. It can be used as an index when verifying the effect. In this embodiment, the information is described mainly from the viewpoint of creating support information for the seller of the product, but the information created in the first embodiment appears in a series of supply chains for manufacturing and selling the product. This is useful information for various companies.

Hereinafter, embodiments will be described. In the embodiment, a sales information analysis system for creating the above-mentioned eight sales activity support information will be described.

FIG. 1 shows the configuration of the sales information analysis system according to the present embodiment. In the sales information analysis system 1000, the sales information analysis device 1 is a first user terminal 500a, a second user terminal 500b, and a third user terminal 500b, which are collectively referred to as a user terminal 500, via a LAN (Local Area Network) 400. It is connected to the user terminal 500c. The user terminal 500 is a general web terminal equipped with a web browser. The sales information analyzer 1 has a function as a web server, and transmits a web page, typically an HTML (HyperText Markup Language) file, to the user terminal 500. By displaying this web page on the web browser of the user terminal 500, a user interface for the user to access the sales information analyzer 1 is provided.

The sales information analysis device 1 is also connected to the sales information holding device 300 via the LAN 400. The sales information analysis device 1 analyzes the sales information acquired from the sales information holding device 300 and creates the above-mentioned sales activity support information. The sales information holding device 300 is a device that holds various sales information, and is typically a database server equipped with DBMS (DataBase Management System) software. The sales information holding device 300 includes a sales promotion record information holding unit 302, a product sales information holding unit 304, a store customer information holding unit 306, and a store sales information holding unit 308.

The sales promotion performance information holding unit 302 holds the sales promotion performance information indicating the performance of the sales promotion measures for the product. FIG. 2 shows a data structure of sales promotion performance information. Identification information of the sales promotion measure is recorded in the sales promotion column 310. This sales promotion measure includes various forms of sales promotion measures, from TV commercials and newspaper advertisements to over-the-counter bargain sales at stores. In the target product column 312, products targeted by the sales promotion measures (hereinafter referred to as “sales promotion target products”) are recorded. A plurality of products may be recorded in the sales promotion target product. For example, in the case of a corporate image CM, many products sold by the company are sales promotion target products. The implementation date of the sales promotion measure is recorded in the implementation date column 314. In addition, the sales promotion performance information may record sales promotion measures of various scales from the national scale to the regional scale, and in this case, even if a target area column indicating the area targeted by the sales promotion measures is further provided. good. Return to FIG.

The product sales information holding unit 304 holds product sales information indicating the sales amount per unit period in each area for each product. FIG. 3 shows a data structure of product sales information. In this figure, the sales volume of each product for each area is recorded by the date and time. Product identification information is recorded in the product column 320. In the area column 322, the area where the product was sold is recorded. In the unit period sales amount column 324, the sales amount in a specific unit period such as a daily unit or a weekly unit is recorded. Return to FIG.

The store customer information holding unit 306 holds store customer information indicating how much the customer who purchases the product in each store resides in which area. FIG. 4 shows a data structure of store customer information. Store identification information is recorded in the store column 330. The area in which the customer resides is recorded in the area column 332. In the area customer number column 334, the number of customers residing in each area (hereinafter referred to as "area customer number") is recorded. The number of area customers may be the population of the area itself, but may be the number of customers who actually purchase products at the store among the population of the area. The number of customers in a store may be calculated from the population of the area and the attribute information of the store by using a known analysis method such as an extended Huff model. In the embodiment, the population of the area will be used for the sake of simplicity of explanation. Return to FIG.

The store sales information holding unit 308 holds store sales information indicating the sales volume per unit period for the products sold at each store. FIG. 5 shows a data structure of store sales information. Store identification information is recorded in the store column 340. In the product column 342, the products sold at each store are recorded. In the unit period sales volume column 344, the sales volume in a specific unit period is recorded. Return to FIG.

Note that FIG. 1 shows a sales information analysis system in which each device is connected via a LAN 400, but each device may exist in a physically distant location, and each device may exist in a WAN (Wide Area). It may be connected via Network) or the Internet. Further, the analysis service by the sales information analysis device 1 is not limited to the provision within one company, and the user terminal 500 existing in different companies in the form of ASP (Application Service Provider) or SaaS (Software as a Service). May be provided to.

FIG. 6 is a block diagram showing a functional configuration of the sales information analyzer 1 of FIG. Each block shown in the block diagram of the present specification can be realized by an element such as a computer CPU or a mechanical device in terms of hardware, and can be realized by a computer program or the like in terms of software. It depicts a functional block realized by their cooperation. Therefore, it is understood by those skilled in the art that these functional blocks can be realized in various forms by combining hardware and software, and the present invention is not limited to any of these.

Further, in the sales information analysis device 1, an operation system (hereinafter referred to as "OS") that provides functions and an environment for efficiently using the device and controls the entire device is executed. Has been done. By operating each functional block of the block diagram by such an OS, a plurality of softwares are executed.

The sales information analysis device 1 includes a user interface processing unit 10, a data holding unit 20, and a data processing unit 40. The user interface processing unit 10 is in charge of user interface processing with the user terminal 500. The data holding unit 20 is a storage area for holding various data. The data processing unit 40 executes various data processing based on the data acquired from the user interface processing unit 10, the data holding unit 20, and the sales information holding device 300. Further, the data processing unit 40 also serves as an interface between the user interface processing unit 10 and the data holding unit 20.

User interface processing unit 10:
The user interface processing unit 10 has an analysis request detection unit 12 and an analysis result output unit 14. The analysis request detection unit 12 receives the data transmitted from the user terminal 500 via the LAN 400. For example, it receives analysis request data that specifies a product and requests analysis of the product share for the product category to which the product belongs, or analysis request data that specifies a product and requests analysis of the factors by which the product was purchased. .. The analysis result output unit 14 transmits data to the user terminal 500 via the LAN 400. For example, analysis result data indicating the product share and analysis result data indicating the factor that the product was purchased are transmitted.

Data holding unit 20:
The data holding unit 20 includes an area cluster information holding unit 22, a sales promotion effect information holding unit 24, a product category information holding unit 26, and a product index value information holding unit 28.

The area cluster information holding unit 22 holds the area cluster information indicating the correspondence relationship between the area cluster, each area, and the inhabitant attribute. FIG. 7 shows the data structure of the area cluster information. The area cluster identification information is recorded in the area cluster column 200. A plurality of areas included in the area cluster are recorded in the area column 202. The population of each area is recorded in the population column 204.

In the resident attribute column 206, the resident attribute of the area cluster is recorded. The resident attribute of the area cluster is, in other words, a resident attribute that is commonly applied to a plurality of areas in the area column 202. The inhabitant attributes include demographic attributes that show demographic attribute data for each area cluster, and psychographic attributes that show attribute data related to human psychology such as the values and lifestyles of the residents of each area cluster. included.

Examples of demographic attributes include "nuclear families with relatively small children in their 30s and 40s" and "the income is slightly higher than average and many are college graduates or above". On the other hand, as an example of psychographic attributes, "For women in their twenties, the brand, safety, and economy are very important, but there is almost no environmental orientation." It is environment-oriented. " That is, the area cluster information is information showing the consumer's living environment and way of thinking, in other words, the consumer image of each area cluster.

In the inhabitant index value column 208, the inhabitant index value obtained by converting the inhabitant attribute of the area cluster into an index value is recorded. The inhabitant index value is a plurality of types of conformance values obtained by indexing the degree to which the inhabitant attributes of the area cluster match the plurality of types of evaluation criteria by a predetermined evaluation function. Specific examples of inhabitant index values include average annual income, average number of households, average number of children per household, proportion in their 30s, proportion in their 40s, and the like. In other words, the inhabitant index value is a conversion of the inhabitant attribute into a plurality of numerical values for computer calculation.

An example of how to create area cluster information will be described. To create the area cluster information, attribute information for each area, various published statistical information, original estimation information, and questionnaire result information are used. Statistical information includes age ratio, gender ratio, occupation ratio, educational background ratio, and the like. Estimated information includes average income, average assets, average land prices, etc. The result information of the questionnaire includes the answer results such as lifestyle, values, and consumption taste. Qualitative information is converted into index values by a predetermined evaluation function. The area cluster information creation unit (not shown) divides each area into groups by cluster analysis using various attribute information indexized for each area as a variable. The various index values used in the cluster analysis may be used as the values in the inhabitant index value column 208, and the data to be set in the inhabitant attribute column 206 based on the values in the inhabitant index value column 208 may be set by human judgment. May be good. In addition, the sales performance of various products in each area may be used as a variable in the cluster analysis. Return to FIG.

The sales promotion effect information holding unit 24 holds sales promotion effect information indicating the influence of the sales promotion measure on the sales volume of the sales promotion target product. FIG. 8 shows a data structure of sales promotion effect information. Identification information of the sales promotion measure is recorded in the sales promotion column 210. A product to be promoted is recorded in the target product column 212. The area cluster identification information is recorded in the area cluster column 214. In the sales volume increase rate column 216, the rate at which the sales volume of the product to be promoted is increased by the sales promotion measure (hereinafter referred to as "sales volume increase rate"), in other words, the effect of the sales promotion is recorded.

The product category information holding unit 26 records product category information indicating the correspondence between the product and the product category. FIG. 9 shows a data structure of product category information. The product category is recorded in the product category column 220. In the product column 222, the products associated with each product category are recorded.

The product index value information holding unit 28 is a product in which a plurality of stores are associated with an index value (hereinafter, referred to as “product index value”) indicating how much the product is supported by customers in each store. Holds index value information. FIG. 10 shows a data structure of product index value information. Store identification information is recorded in the store column 230. In the product column 232, the products sold at each store are recorded. In the product index value column 234, the product index value of each product in each store is recorded. Return to FIG.

Data processing unit 40:
The data processing unit 40 includes an area sales volume acquisition unit 42, a cluster sales volume calculation unit 44, a sales promotion effect analysis unit 50, a sales volume correction unit 60, a product share analysis unit 70, and a purchasing factor analysis unit 80. It has a store customer base analysis unit 90, a sales condition influence analysis unit 100, a repeat rate analysis unit 110, a product appeal degree analysis unit 120, a product appeal degree correction unit 130, and a store sales volume prediction unit 140.

The area sales volume acquisition unit 42 accesses the product sales information holding unit 304 to refer to the product sales information, and for each combination of the area and the product, the sales volume of the product in the area for a predetermined period (hereinafter, “area sales volume”). ".) Is acquired. FIG. 11 shows the area sales volume acquired by the area sales volume acquisition unit 42. The area sales volume in the figure is a total of the sales volumes from 10/30 to 11/3 of the product sales information shown in FIG. Return to FIG.

The cluster sales amount calculation unit 44 accesses the area cluster information holding unit 22 and refers to the area cluster information. Then, the area sales volume is totaled for each combination of the area cluster and the product, and the sales volume of the product in the area cluster in a predetermined period (hereinafter, referred to as "cluster sales volume") is calculated. FIG. 12 shows the cluster sales volume calculated by the cluster sales volume calculation unit 44. The cluster sales volume in the figure is a total of the area sales volumes shown in FIG. 11 for each product and each area cluster. Return to FIG.

Sales Promotion Effect Analysis Department 50:
The sales promotion effect analysis unit 50 creates effect information of the sales promotion measure, which is one of the sales activity support information. The effect information of the sales promotion measure in this embodiment is information showing the rate of increase in the sales volume of the product to be promoted by the sales promotion measure for each area cluster. FIG. 13 is a block diagram showing a functional configuration of the sales promotion effect analysis unit 50. The sales promotion effect analysis unit 50 includes a sales promotion sales volume acquisition unit 52, a sales promotion sales volume totaling unit 54, and a sales promotion effect calculation unit 56.

The sales promotion sales volume acquisition unit 52 first accesses the sales promotion performance information holding unit 302, refers to the sales promotion performance information, and identifies the sales promotion target product and the implementation date of the sales promotion measure to be analyzed. For example, when the sales promotion measure “TV advertisement α” in FIG. 2 is the analysis target, the sales promotion target product is specified as “○○ coffee” and the implementation date is specified as “November 1”. The sales promotion sales volume acquisition unit 52 may receive the designated data of the sales promotion measure to be analyzed from the user terminal 500 via the analysis request detection unit 12. Further, regardless of the designation by the user terminal 500, the sales promotion measure recorded in the sales promotion performance information may be automatically acquired as the sales promotion measure to be analyzed.

Next, the sales promotion sales volume acquisition unit 52 accesses the product sales information holding unit 304 to refer to the product sales information, and for the sales promotion target products in each area, the sales volume in a predetermined period before the implementation of the sales promotion measures (hereinafter referred to as “the sales volume”). "Sales volume before sales promotion") is acquired. In addition, the sales volume (hereinafter referred to as "post-sales promotion sales volume") for a predetermined period after the implementation of the sales promotion measures is acquired. Regarding this period, it is sufficient to set a period that is considered appropriate for determining the effect of sales promotion measures based on the experience and experiments of the company. In the above example, the sales volume will be acquired for two days before and after the implementation of the sales promotion measure "TV advertisement α".

FIG. 14A shows the pre-promotion sales volume and the post-promotion sales volume acquired by the sales promotion sales volume acquisition unit 52. In the figure, for "○○ coffee" in Fig. 3, the sales volume from October 30th to October 31st is added up as the pre-promotion sales volume, and the sales volume from November 1st to 11/2 is added up. It shows that it was acquired as a sales volume after promotion. If the sales promotion measures are implemented only in a part of the area as described above, the sales promotion sales volume acquisition unit 52 acquires the information of the area from the sales promotion performance information and obtains the information of the area from the product sales information. You may acquire the pre-promotion sales volume and the post-promotion sales volume only for. Return to FIG.

The sales promotion sales volume totaling unit 54 accesses the area cluster information holding unit 22 to refer to the area cluster information, and totals the pre-promotion sales volume and the post-promotion sales volume for each area cluster. FIG. 14B shows the pre-promotion sales volume and the post-promotion sales volume aggregated by the sales promotion sales volume aggregation unit 54. That is, the pre-promotion sales volume and the post-promotion sales volume of the areas a, b, and c are aggregated, and the pre-promotion sales volume and the post-promotion sales volume of the cluster A are calculated. The same applies to area d, area e, and cluster B. Return to FIG.

The sales promotion effect calculation unit 56 compares the aggregated value of the pre-promotion sales volume and the aggregated value of the post-promotion sales volume for each area cluster, and determines the sales volume increase rate of the sales promotion target product by the sales promotion measures to be analyzed. Calculated for each cluster. Here, the sales promotion effect calculation unit 56 calculates the sales volume increase rate by calculating the ratio of the aggregated values. For example, when the aggregated value of the pre-promotion sales volume and the aggregated value of the post-promotion sales volume are in the state of FIG. 14B, the sales promotion effect calculation unit 56 sets the sales volume increase rate for the cluster A to "1.13". , The sales volume increase rate for cluster B is calculated as "1.64".

The sales promotion effect calculation unit 56 accesses the sales promotion effect information holding unit 24, and records the sales volume increase rate of the sales promotion target product in each area cluster in the sales promotion effect information for the sales promotion measures to be analyzed. Further, when the designation of the sales promotion measure to be analyzed is received from the user terminal 500, the sales volume increase rate in each area cluster calculated for the sales promotion measure is transmitted to the user terminal 500 via the analysis result output unit 14. You may send it.

With the conventional method of analyzing the effects of sales promotion measures, it is clear that the total sales volume of the products to be promoted has increased, but it is not clear what characteristics the sales volume to consumers has increased. Therefore, it was not clear whether the sales promotion measures appealed to the target customer base. According to the sales promotion effect analysis unit 50, the effect of the sales promotion measure can be calculated for each area cluster associated with the customer image from various viewpoints such as demographic attributes and psychographic attributes.

FIG. 14 (c) shows the rate of increase in sales volume in each area cluster of the products targeted for sales promotion by the sales promotion measures. The content of the figure may be displayed on the user terminal 500 as effect information of the sales promotion measure. Product sellers can confirm the effect of sales promotion measures in detail by confirming the rate of increase in sales volume in each area cluster. For example, from the results of FIG. 14 (c), it can be confirmed that the customer of cluster B was strongly appealed, and that different sales promotion measures are required when targeting cluster C.

As a modification, the sales promotion effect analysis unit 50 may include a resident attribute influence analysis unit (not shown). The Resident Attribute Impact Analysis Department conducts regression analysis with the resident attribute of each area cluster as the explanatory variable and the sales volume increase rate in each area cluster as the objective variable due to specific sales promotion measures, and the effect of the resident attribute on the sales volume increase rate. Is calculated as a regression coefficient. By comparing the regression coefficients calculated for each inhabitant attribute, it is possible to identify the inhabitant attribute that has a relatively large effect on the sales volume increase rate. The product seller can implement effective sales promotion measures at a low cost by implementing the above-mentioned specific sales promotion measures only for the area cluster having the inhabitant attribute. Return to FIG.

Sales volume correction unit 60:
The sales volume correction unit 60 removes the increase in sales volume due to the sales promotion measures that are not subject to analysis from the post-promotion sales volume calculated by the sales promotion sales volume totaling unit 54. FIG. 15 is a block diagram showing a functional configuration of the sales volume correction unit 60. The sales volume correction unit 60 includes a sales promotion effect acquisition unit 62, a sales promotion sales volume correction unit 64, a sales promotion sales volume specifying unit 66, and a cluster sales volume correction unit 68.

The sales promotion effect acquisition unit 62 first accesses the sales promotion performance information holding unit 302 to refer to the sales promotion performance information. Then, the sales promotion measures that the sales promotion target products are the same as the sales promotion measures that are the analysis target, and the sales promotion measures that are not the analysis target that are implemented within the predetermined period after the sales promotion measures that are the analysis target are implemented are removed. Specify as a target sales promotion measure. This predetermined period may be a period for acquiring the sales volume after sales promotion for the sales promotion measures to be analyzed, and may be appropriately determined based on the experience and experiment of the company. FIG. 16 shows a data structure of sales promotion performance information. The figure shows that in addition to the sales promotion performance information in Fig. 2, the sales promotion measure "newspaper advertisement θ" that was not analyzed was implemented the day after the sales promotion measure "TV advertisement α" that was analyzed was implemented. There is. In this case, in order to remove the increase in the sales volume of "○○ coffee" due to the "newspaper advertisement θ", the sales promotion effect acquisition unit 62 specifies the "newspaper advertisement θ" as a sales promotion measure to be removed. Return to FIG.

Next, the sales promotion effect acquisition unit 62 accesses the sales promotion effect information holding unit 24, refers to the sales promotion effect information, and acquires the sales volume increase rate of the sales promotion target product by the sales promotion measure to be removed. When the sales promotion effect information is in the state shown in FIG. 8, the sales promotion effect acquisition unit 62 has the sales volume increase rate “1.4” of “○○ coffee” in the cluster A by the “newspaper advertisement θ” and the “1.4” in the cluster B. Acquire the sales volume increase rate of "1.1" for "○○ Coffee". Return to FIG.

The sales promotion sales volume correction unit 64 measures the post-promotion sales volume for each area cluster so that the larger the sales volume increase rate for the sales promotion measures to be removed, the smaller the post-promotion sales volume for the sales promotion measures to be analyzed. Correct to. Specifically, the sales promotion sales volume correction unit 64 overlaps the period for acquiring the post-promotion sales volume for the sales promotion measures to be analyzed and the predetermined period from the implementation date for the sales promotion measures to be removed. To identify. This predetermined period is a period that is strongly influenced by the sales promotion measures to be removed, and may be appropriately determined based on the experience and experiments of the company. The sales promotion sales volume correction unit 64 removes the sales volume increase rate due to the sales promotion measures to be removed from the post-promotion sales volume during the overlapping period.

FIG. 17 shows a data structure of product sales information. Regarding "○○ coffee" in the figure, the unit period sales volume of 11/2 and 11/3 is different from that of Fig. 3, not only the sales promotion measure "TV advertisement α" to be analyzed, but also the sales promotion not to be analyzed. The increase in sales volume due to the measure "newspaper advertisement θ" is added.

FIG. 18A shows the pre-promotion sales volume and the post-promotion sales volume acquired by the sales promotion sales volume acquisition unit 52. The data in this figure is obtained from the product sales information in Fig. 17 for the sales volume of "○○ coffee" from 10/30 to 11/2, two days before and after the implementation of the sales promotion measure "TV advertisement α" to be analyzed. It was done. As described above, since the increase in sales volume due to the sales promotion measure “newspaper advertisement θ” that is not subject to analysis is added, the sales volume after sales promotion is larger than that in FIG. 14 (a). Therefore, in this state, the rate of increase in sales volume due to "TV advertisement α" is calculated to be large.

FIG. 18B shows the post-promotion sales volume corrected by the sales promotion sales volume correction unit 64. The post-promotion sales volume in the figure is the sales volume of XX coffee in each area of 11/2 minus the increase in sales volume due to “newspaper advertisement θ”. Specifically, the sales promotion sales volume correction unit 64 divides the sales volume of 11/2 in the areas a, b, and c belonging to the cluster A by the sales volume increase rate "1.4", and the cluster B The sales volume of 11/2 in the areas d and e belonging to is divided by the sales volume increase rate "1.1".

The sales promotion sales volume totaling unit 54 described above totals the pre-promotion sales volume acquired by the sales promotion sales volume acquisition unit 52 and the post-promotion sales volume corrected by the sales promotion sales volume correction unit 64 for each area cluster. The sales promotion effect calculation unit 56 compares these aggregated values and calculates the sales volume increase rate of the sales promotion target product due to the sales promotion measures to be analyzed.

If the sales promotion measures to be analyzed and the sales promotion measures not to be analyzed are implemented at the same time, the increase in sales volume due to the sales promotion measures not to be analyzed will be added to the post-promotion sales volume of the sales promotion measures to be analyzed. It will be added. As a result, the effect of the sales promotion measures to be analyzed will be calculated to be larger than the actual effect. The sales volume correction unit 60 removes the additional amount due to the sales promotion measures not subject to analysis from the post-sales promotion sales volume of the sales promotion measures to be analyzed. As a result, the sales promotion effect calculation unit 56 can correctly calculate the sales volume increase rate of the sales promotion target product due to the sales promotion measures to be analyzed.

When the area where the sales promotion measures to be removed are implemented is limited, the sales promotion sales volume correction unit 64 limits the area where the sales promotion measures to be removed are implemented, and the sales volume after sales promotion for the sales promotion measures to be analyzed is limited to the area where the sales promotion measures are implemented. May be corrected. For example, if the sales promotion measure to be analyzed is a nationwide TV advertisement and the sales promotion measure to be removed is a regional-limited over-the-counter bargain sale, only the area where the over-the-counter bargain sale was carried out is limited to the area where the over-the-counter bargain sale was carried out. Correct the sales volume. As a result, the rate of increase in the sales volume of the sales promotion target product due to the sales promotion measure to be analyzed can be correctly calculated in consideration of the implementation range of the sales promotion measure.

Further, in the examples shown in FIGS. 18A and 18B, the sales promotion sales volume correction unit 64 corrects the post-promotion sales volume for each area, but the aggregated value of the post-promotion sales volume for each area cluster is calculated. It may be corrected. The sales promotion sales volume correction unit 64 corrects the value of the aggregated value of the post-promotion sales volume during the period strongly affected by the sales promotion measures to be removed according to the sales volume increase rate due to the sales promotion measures to be removed. .. Specifically, the larger the rate of increase in sales volume due to the sales promotion measures targeted for removal, the smaller the value of the aggregated value of post-promotion sales volume for the period strongly affected by the sales promotion measures targeted for removal. , Reduce the aggregated value of post-promotion sales volume. The sales promotion effect calculation unit 56 is a total value of the pre-promotion sales volume aggregated by the sales promotion sales volume totaling unit 54, a total value of the post-promotion sales volume, and a total value corrected by the sales promotion sales volume correction unit 64. To calculate the rate of increase in sales volume of the products to be promoted by the sales promotion measures to be analyzed. Return to FIG.

When the cluster sales volume of the product to be promoted is calculated by the cluster sales volume calculation unit 44, the sales volume correction unit 60 removes the increase in the sales volume due to the sales promotion measure from the cluster sales volume. FIG. 19 shows the cluster sales volume after correction by the sales volume correction unit 60. The cluster sales volume in FIG. 12 is the cluster sales volume in FIG. 12, which is a value including the increase in sales volume due to the sales promotion measure “TV advertisement α”. In the following specific example, it will be described with reference to FIG.

In order to remove the increase in sales volume due to the sales promotion measure, the sales promotion sales volume acquisition unit 52 first acquires the pre-promotion purchase volume of the product to be promoted. Here, since the product to be promoted is "○○ coffee" in FIG. 3 and the implementation date of the sales promotion measure is 11/1, the sales promotion sales volume acquisition unit 52 is "○○ coffee" in each area of 10/31. Will be acquired. Next, the sales promotion sales volume totaling unit 54 totals the pre-promotion purchase volume of the sales promotion target product for each area cluster. In the case of "○○ coffee" in FIG. 3, the aggregated value of 10/31 for cluster A including areas a, b, and c is 265, and the aggregated value of 10/31 for cluster B including areas d and e is 150. Become. This is shown in the pre-promotion sales volume in FIG.

The sales promotion sales volume specifying unit 66 determines whether the aggregation period of the cluster sales volume calculated by the cluster sales volume calculation unit 44 includes the implementation date and time of the sales promotion measures for the products corresponding to the cluster sales volume. judge. When it is determined that it is included, the cluster sales volume for a predetermined period in which the influence of the sales promotion measure is expected to be high is specified from the implementation date of the sales promotion measure. For this period, an appropriate period may be set based on the experience and experiment of the company.

For example, the sales promotion measure "TV advertisement α" implemented on November 1st for "○○ coffee" will greatly affect the sales volume of "○○ coffee" on November 1st and 11/2. At this time, the sales promotion sales volume specifying unit 66 specifies the cluster sales volume of 11/1 and 11/2 from the cluster sales volume as the cluster sales volume greatly affected by the “television advertisement α”. This is shown in the sales promotion impact of FIG. The sales promotion impact in FIG. 3 is the total of the sales volume of “○○ coffee” in 11/1 and 11/2 in FIG. 3 for each area cluster. Return to FIG.

The cluster sales volume correction unit 68 corrects the cluster sales volume based on the pre-promotion sales volume that is not affected by the sales promotion measures. One method is to replace the cluster sales volume affected by the sales promotion measures with the pre-promotion sales volume. For example, the value of the sales promotion influence in FIG. 19, that is, the aggregated value of the sales volume of 11/1 and 11/2 is the value in the pre-promotion sales volume column, here, the pre-promotion sales volume is set to 2 for 2 days. Replace with the doubled value. Then, the cluster sales volume that is not affected by the sales promotion measures, that is, the sales volume of 10/30, 10/31, and 11/3 is totaled. The cluster sales volume corrected in this way is shown in the corrected cluster sales volume in FIG.

In another method of correcting the cluster sales volume, the sales promotion effect analysis unit 50 described above calculates the sales volume increase rate of the sales promotion target product due to the sales promotion measures to be removed. The cluster sales volume correction unit 68 corrects the cluster sales volume affected by the sales promotion measures according to the value of the sales volume increase rate. For example, the cluster sales volume affected by the sales promotion measures may be divided by the sales volume increase rate and aggregated with the cluster sales volume not affected by the sales promotion measures.

In yet another method of correcting the cluster sales volume, the cluster sales volume correction unit 68 accesses the sales promotion effect information holding unit 24 to acquire the sales volume increase rate of the sales promotion target product by the sales promotion measure to be removed. The cluster sales volume correction unit 68 corrects the cluster sales volume affected by the sales promotion measures according to the value of the sales volume increase rate. For example, the cluster sales volume affected by the sales promotion measures may be divided by the sales volume increase rate and aggregated with the cluster sales volume not affected by the sales promotion measures.

According to the sales volume correction unit 60, it is possible to remove the increase in sales volume due to the sales promotion measure from the cluster sales volume of the products affected by the sales promotion measure. The increase in sales volume due to sales promotion measures is often in a temporary state, and the sales volume including the increase often does not reflect the ability of the product itself. By removing the influence of the sales promotion measures by the sales volume correction unit 60, it is possible to calculate the cluster sales volume that reflects the ability of the product itself.

By appropriately setting the period for which the effects of sales promotion measures should be eliminated, the rapid and temporary increase in sales volume due to sales promotion measures can be eliminated, while the increase in sales volume due to the improvement of the product itself. Remains in the cluster sales volume. In other words, the increase in sales volume due to the increase in name recognition of the product due to sales promotion measures should not be excluded because it is considered that the sales volume will increase due to the improvement of the ability of the product itself. Since such effects will continue even after the period when the influence of sales promotion measures should be eliminated, the increase in sales volume due to the improvement of the product itself can be reflected in the cluster sales volume.

The cluster sales volume corrected by the cluster sales volume correction unit 68 is used in the product share analysis unit 70, the purchasing factor analysis unit 80, the sales condition impact analysis unit 100, the repeat rate analysis unit 110, and the product appeal analysis unit 120, which will be described later. can. As a result, the influence of the increase in sales volume due to the sales promotion measure can be removed from the sales activity support information created in each functional block, and the sales activity support information can be created based on the ability of the product itself. A specific example thereof will be described in detail in the product share analysis unit 70, which will be described later. Return to FIG.

Product Share Analysis Department 70:
The product share analysis unit 70 creates product share information, which is one of the sales activity support information. The product share information in the present embodiment is information indicating the ratio of the sales volume of a certain product in the product category for each area cluster. FIG. 20 is a block diagram showing a functional configuration of the product share analysis unit 70. The product share analysis unit 70 includes a category sales volume calculation unit 72 and an occupancy rate calculation unit 74.

The category sales volume calculation unit 72 accesses the product category information holding unit 26 and refers to the product category information. Then, the cluster sales volume calculated by the cluster sales volume calculation unit 44 is aggregated for each combination of the area cluster and the product category, and the sales volume for a specific product category in each area cluster in a predetermined period (hereinafter, "category sales"). It is called "quantity".) Is calculated. FIG. 21A shows the category sales volume calculated by the category sales volume calculation unit 72. The category sales volume in the figure is the aggregate of the cluster sales volume in Fig. 12 for each area cluster for "○○ coffee", "△△ coffee", and "□□ blend" that belong to the product category "coffee".

Further, the category sales volume calculation unit 72 may calculate the category sales volume by aggregating the cluster sales volume corrected by the cluster sales volume correction unit 68 described above. FIG. 21B shows the category sales volume calculated by the category sales volume calculation unit 72. The category sales volume in the figure is the cluster sales volume shown in Fig. 12 for "△△ coffee" and "□□ blend", and the corrected cluster sales volume in Fig. 19 for "○○ coffee" for each area cluster. It is a total. Return to FIG.

The occupancy rate calculation unit 74 calculates the ratio of the category sales amount for each area cluster to the cluster sales amount of each product in each area cluster as the occupancy rate. For example, since the category sales volume of the cluster A in FIG. 21 (a) is “2795” and the cluster sales volume of “○○ coffee” in FIG. 12 is “1475”, the occupancy rate is calculated as 53%. FIG. 22A shows the occupancy rate of each product in the product category “coffee”. The content of the figure may be displayed on the user terminal 500.

Further, the occupancy rate calculation unit 74 may calculate the ratio of the cluster sales amount corrected by the cluster sales amount correction unit 68 and the category sales amount obtained by totaling the corrected cluster sales amount as the occupancy rate. FIG. 22B shows the occupancy rate of each product in the product category “coffee”. The figure shows the ratio between the cluster sales volume of "○○ Coffee", which is obtained by removing the increase in sales volume due to the sales promotion measure "TV Advertising α", and the category sales volume, which is the total of these. In this figure, the increase in sales volume of “○○ coffee” due to sales promotion measures has been removed, so the occupancy rate of “△△ coffee” and “□□ blend” is larger than in Fig. 22 (a). It has become.

The category sales volume calculation unit 72 may receive the designated data of the product to be analyzed from the user terminal 500 via the analysis request detection unit 12. At this time, the occupancy rate calculation unit 74 calculates the occupancy rate of the analysis target product in the product category to which the analysis target product belongs, and transmits it to the user terminal 500 as product share information via the analysis result output unit 14. Similarly, the category sales amount calculation unit 72 may receive the designated data of the product category to be analyzed from the user terminal 500. At this time, the occupancy rate calculation unit 74 calculates the occupancy rate of each product in the product category to be analyzed and transmits it to the user terminal 500.

According to the product share analysis unit 70, the product seller can confirm the share of each product in the product category for each area cluster. That is, detailed competitive product analysis is realized. As mentioned above, since the area cluster is associated with the consumer image, the product seller can compare and examine which product is supported by which consumer in the product category. This makes it possible to make appropriate decisions regarding the purchase of products, shelving allocation, and the like. In addition, by calculating the occupancy rate based on the cluster sales volume excluding the increase in sales volume due to sales promotion measures, the product seller can also confirm the share of each product by the ability of the product itself. Return to FIG.

Purchasing Factor Analysis Department 80:
The purchasing factor analysis unit 80 creates purchasing factor information, which is one of the sales activity support information. The purchasing factor information in the present embodiment is information indicating the resident attribute that influences the purchase of the product by the customer. FIG. 23 is a block diagram showing a functional configuration of the purchasing factor analysis unit 80. The purchasing factor analysis unit 80 includes a resident attribute acquisition unit 82 and a resident attribute evaluation unit 84.

The resident attribute acquisition unit 82 receives the designated data of the product to be analyzed from the user terminal 500 via the analysis request detection unit 12. The inhabitant attribute acquisition unit 82 first acquires the cluster sales amount calculated by the cluster sales amount calculation unit 44, which is the sales amount of the product to be analyzed in each area cluster. Next, the inhabitant attribute acquisition unit 82 identifies an area cluster associated with each cluster sales volume. Finally, the inhabitant attribute acquisition unit 82 accesses the area cluster information holding unit 22 to refer to the area cluster information, acquires the inhabitant attribute of the specified area cluster, and associates it with each cluster sales volume.

For example, it is assumed that the cluster sales volume of the analysis target product “○○ coffee” is in the state shown in FIG. At this time, the resident attribute acquisition unit 82 acquires the cluster sales amount “1475” in the cluster A and the cluster sales amount “960” in the cluster B for the analysis target product “○○ coffee”. The inhabitant attribute acquisition unit 82 refers to the area cluster information in FIG. 7, and sets the inhabitant attribute of cluster A "high ratio in 30s" and the like to the cluster sales volume "1475" and the inhabitant attribute of cluster B "50s". "High ratio" is associated with the cluster sales volume "960". Return to FIG.

The inhabitant attribute evaluation unit 84 compares the cluster sales volume of each area cluster and identifies an area cluster having a relatively large cluster sales volume. Then, the inhabitant attribute of the area cluster is determined as the inhabitant attribute that has a high correlation with the sales volume of the product to be analyzed. For example, the cluster sales volume of "○○ coffee" is larger in cluster A than in cluster B. Therefore, the inhabitant attribute evaluation unit 84 determines the inhabitant attribute of cluster A, such as "the ratio of people in their thirties is high", as the inhabitant attribute that has a high correlation with the sales volume of "○○ coffee". The resident attribute evaluation unit 84 transmits to the user terminal 500 the purchase factor information indicating the resident attribute determined to have a high correlation with the sales amount of the product to be analyzed, via the analysis result output unit 14.

According to the purchasing factor analysis unit 80, it is possible to identify the inhabitant attribute that influences the purchase of the product by the customer. That is, the product seller confirms the purchase factor information created by the purchase factor analysis unit 80, and makes a decision on the product assortment and shelving allocation according to the attributes of the customers of each store. The sales volume can be increased.

Next, a modified example of determining the inhabitant attribute that has a high correlation with the product to be analyzed will be described. The inhabitant attribute acquisition unit 82 refers to the area cluster information and acquires the inhabitant index value together with the inhabitant attribute. The inhabitant attribute evaluation unit 84 uses the cluster sales volume in each area cluster as the objective variable, and the regression analysis using the inhabitant index value in each area cluster as the explanatory variable, and the influence of each inhabitant index value on the cluster sales volume as the regression coefficient. calculate. The inhabitant attribute evaluation unit 84 may select an explanatory variable by using a known test method such as a t-test or an F-test and a known selection method such as a variable decrease method or a variable increase method. The inhabitant attribute evaluation unit 84 determines the inhabitant attribute associated with the inhabitant index value having a relatively large absolute value of the regression coefficient as the inhabitant attribute having a high correlation with the sales volume of the product to be analyzed, and determines the inhabitant attribute in the user terminal 500. Send.

FIG. 24 illustrates the relationship between the inhabitant index value and the cluster sales volume. The content of the figure may be displayed on the user terminal 500 as purchase factor information. The numerical values in the figure show the regression coefficient calculated by the above-mentioned regression analysis, and it is shown that the inhabitant index value with a blank numerical value does not have a significant effect on the cluster sales volume. For example, product sellers say that "△△ coffee" is purchased by low-income customers, "○○ coffee" is not often purchased by households with many children, while "△△ coffee" is the number of children. From this result, it can be confirmed that the coffee is often purchased by households with a large number of coffee.

According to this modification, the correlation between the sales volume of each product and the inhabitant attribute can be specified in detail. That is, the product seller can understand the reason why each product is purchased by the customer. For example, it is possible to grasp the influence of the average income of residents on the sales volume of each product. As a result, product sellers increase the sales volume of products by purchasing appropriate products and allocating shelves according to the attributes of customers at each store, such as purchasing many products that match the average income of major customers. Can be made to. Return to FIG.

Store Customer Base Analysis Department 90:
The store customer base analysis unit 90 creates store customer base information, which is one of the sales activity support information. The store customer segment information in the present embodiment is information indicating the ratio of the number of customers residing in each area cluster to the total number of customers in the store for each combination of the store and the area cluster. FIG. 25 is a block diagram showing a functional configuration of the store customer demographic analysis unit 90. The store customer demographic analysis unit 90 includes a cluster customer number calculation unit 92 and a cluster customer ratio calculation unit 94.

The cluster customer number calculation unit 92 receives the designated data of the store to be analyzed from the user terminal 500 via the analysis request detection unit 12. The cluster customer number calculation unit 92 accesses the store customer information holding unit 306, refers to the store customer information, aggregates the number of area customers for each combination of the store and the area cluster, and counts the number of customers residing in each area cluster. (Hereinafter referred to as "the number of cluster customers") is calculated. For example, in the case of "Kiba store" in FIG. 4, the number of cluster customers in cluster A is "600", which is the number of customers in area a, and the number of cluster customers in cluster C is the aggregated value of the number of customers in area f and area i. It becomes 1150 ". In reality, the larger the store, the larger the area where the customers of the store live, so the number of area customers in many areas is aggregated for each area cluster, and the number of cluster customers is calculated for each area. To.

The cluster customer ratio calculation unit 94 calculates the ratio of the number of cluster customers to the total number of customers in the store (hereinafter, referred to as “cluster customer ratio”) for each combination of the store and the area cluster. For example, in the case of the Kiba store in FIG. 4, the total number of customers is "2550", which is the sum of the values in the area customer number column 334. The ratio of the total number of customers to the number of cluster customers "600" in cluster A is "23%", and the ratio of the total number of customers to the number of cluster customers "1150" in cluster C is "45%".

FIG. 26A shows the cluster customer ratio of the “Kiba store” in FIG. FIG. 26B shows the cluster customer ratio of the “Kasai store” in FIG. FIG. 26 (c) shows the cluster customer ratio of the “Funabashi store” in FIG. The cluster customer ratio calculation unit 94 transmits store customer demographic information indicating the cluster customer ratio of each area cluster to the user terminal 500 for the analysis target store via the analysis result output unit 14.

According to the store customer demographic analysis unit 90, the ratio of customers residing in each area cluster of the store to be analyzed can be calculated. For example, from FIG. 26 (a), the product seller can confirm that the customers of the “Kiba store” have many residents of cluster C, and the store strategy that emphasizes the inhabitant attributes of cluster C is implemented at the “Kiba store”. Can be done. Return to FIG.

Sales Condition Impact Analysis Department 100:
The sales condition impact analysis unit 100 creates sales condition impact information, which is one of the sales activity support information. The sales condition influence information in the present embodiment is information indicating the sales volume increase rate for each area cluster before and after the setting of the sales condition of a certain product. FIG. 27 is a block diagram showing a functional configuration of the sales condition influence analysis unit 100. The sales condition influence analysis unit 100 includes a comparative sales volume acquisition unit 102 and a sensitivity rate evaluation unit 104.

The comparative sales volume acquisition unit 102 receives the designated data of the product to be analyzed from the user terminal 500 via the analysis request detection unit 12. For the product to be analyzed, the specified data of the period before the change of the sales condition and the period after the change of the sales condition may be received from the user terminal 500, and the date and time of the change of the sales condition of each product is held in a storage unit (not shown). You may. The comparative sales volume acquisition unit 102 refers to the cluster sales volume of the analysis target product calculated by the cluster sales volume calculation unit 44, and refers to the cluster sales volume for a predetermined period before the change of sales conditions (hereinafter, “sales volume before change””. ) And the cluster sales volume for a predetermined period after the change of sales conditions (hereinafter referred to as "changed sales volume") are acquired.

FIG. 28A shows the sales volume before the change and the sales volume after the change acquired by the comparative sales volume acquisition unit 102. The sales volume before and after the change in the figure are the cluster sales volume and sales conditions before the change of the sales conditions of the analysis target product "□□ Blend" among the cluster sales volumes calculated by the cluster sales volume calculation unit 44. Shows the cluster sales volume after the change. In the following example, it will be described based on the figure.

The sensitivity rate evaluation unit 104 calculates the ratio of the sales volume after the change to the sales volume before the change (hereinafter, referred to as “sales condition sensitivity rate”) for each area cluster. For example, in the case of FIG. 28A, the sales condition sensitivity rates from cluster A to cluster D are "1.41", "1.13", "1.4", and "1.25", respectively. .. FIG. 28B shows the sales condition sensitivity rate calculated by the sensitivity rate evaluation unit 104. The sensitivity rate evaluation unit 104 transmits the sales condition influence information indicating the sales condition sensitivity rate of the analysis target product in each area cluster to the user terminal 500 via the analysis result output unit 14.

The sensitivity rate evaluation unit 104 also refers to the cluster customer ratio calculated by the cluster customer ratio calculation unit 94, and the cluster customer ratio for the area cluster whose sales condition sensitivity rate is equal to or higher than a predetermined threshold value is equal to or higher than a predetermined threshold value. Identify the store that is. Then, the area cluster, the sales condition sensitivity rate, and the store are associated with each other and transmitted to the user terminal 500.

For example, assume that the threshold value of the sales condition sensitivity rate is "1.4" and the threshold value of the cluster customer ratio is "40%". At this time, the sensitivity evaluation unit 104 identifies the clusters A and C from FIG. 28 (b). Then, referring to the cluster customer ratios shown in FIGS. 26 (a) to 26 (c), the “Kasai store” is specified as a store having a high cluster customer ratio in cluster A, and as a store having a high cluster customer ratio in cluster C. Identify the "Kiba store". From this result, the product seller can confirm the stores where the change of the sales conditions is particularly effective.

According to the sales condition influence analysis unit 100, it is possible to calculate the sensitivity rate of the product to the sales conditions for the customers residing in each area cluster. By confirming this sensitivity rate, the product seller can easily set the sales conditions of the product appropriately. A typical example of setting sales conditions is product pricing, which makes it easier for product sellers to make decisions about product pricing strategies. In addition, by associating an area cluster with a high sensitivity rate with a store whose main customer is a resident of the area cluster, the product seller can easily formulate a strategy for each store regarding the setting of sales conditions of the product. For example, it is possible to make a decision to reduce the price of a certain product only for stores whose main customers are residents of an area cluster that is highly sensitive to the price of a certain product.

Repeat rate analysis unit 110:
The repeat rate analysis unit 110 creates repeat rate information, which is one of the sales activity support information. The repeat rate information of the present embodiment is information indicating the rate at which the same product is repeatedly sold to the same customer in each area cluster. FIG. 29 is a block diagram showing a functional configuration of the repeat rate analysis unit 110. The repeat rate analysis unit 110 includes an area repeat number acquisition unit 112, a cluster repeat number calculation unit 114, a repeat rate calculation unit 116, and a repeat rate evaluation unit 118.

In the product sales information shown in FIG. 3, the number of area repeats (not shown) is further recorded for each product and area, and the number of times the product is repeatedly sold to the same customer in a predetermined period (hereinafter, "number of area repeats"). ".) Is recorded. For example, for "△△ coffee" in the period from 10/30 to 11/3 in FIG. 3, the number of area repeats in the area a, the number of area repeats in the area b, and so on are recorded. The number of area repeats may be counted as 2 times, 3 times, ... If the same customer purchases at different timings, and if the same customer purchases repeatedly, how many times. However, it may be counted only once.

The area repeat number acquisition unit 112 accesses the product sales information holding unit 304 to refer to the product sales information, and acquires the area repeat number in a predetermined period for each combination of the area and the product.

The cluster repeat number calculation unit 114 accesses the area cluster information holding unit 22 to refer to the area cluster information, and totals the number of area repeats for each combination of the area cluster and the product. Hereinafter, this aggregated value will be referred to as "the number of cluster repeats". For example, the number of area repeats for areas a, b, and c is aggregated to calculate the number of cluster repeats for cluster A. FIG. 30A shows the number of cluster repeats calculated by the cluster repeat number calculation unit 114. The figure shows the number of cluster repeats for “△△ coffee”, and the cluster sales volume in the figure shows the cluster sales volume calculated by the cluster sales volume calculation unit 44.

The repeat rate calculation unit 116 receives the designated data of the product to be analyzed from the user terminal 500 via the analysis request detection unit 12. The repeat rate calculation unit 116 calculates the ratio of the cluster sales amount and the number of cluster repeats for the product to be analyzed (hereinafter, referred to as “repeat rate”). For example, when the number of cluster repeats of the analysis target product “△△ coffee” is in the state of FIG. 30 (a), the repeat rate of “△△ coffee” in cluster A is calculated to be “8%”. FIG. 30B shows the repeat rate in each area cluster calculated by the repeat rate calculation unit 116.

The repeat rate evaluation unit 118 identifies an area cluster having a relatively large repeat rate for the repeat rate in each area cluster calculated by the repeat rate calculation unit 116. The repeat rate evaluation unit 118 may specify, for example, an area cluster having the highest repeat rate, or may specify an area cluster in which a repeat rate equal to or higher than a predetermined threshold is calculated. When the latter threshold value is "20%", the repeat rate evaluation unit 118 identifies the area clusters B and C from the repeat rate shown in FIG. 30 (b). The repeat rate evaluation unit 118 transmits the specified area cluster to the user terminal 500 via the analysis result output unit 14.

According to the repeat rate analysis unit 110, the repeat rate can be calculated for each area cluster. The product seller can grasp what kind of attribute the consumer has repeatedly purchased the product. In addition, it is possible to make decisions on product sales strategies according to the attributes of major customers. For example, by sending direct mail or the like to residents living in an area cluster with a high repeat rate, it is possible to effectively increase the sales volume of products.

In the first modification, the repeat rate evaluation unit 118 indicates the sales volume per person (hereinafter referred to as “unit population sales volume”) indicating the ratio of the cluster sales volume to the population of the area cluster for each of the plurality of area clusters. Is calculated, and the repeat rate is associated with the unit population sales volume.

Specifically, the repeat rate evaluation unit 118 accesses the area cluster information holding unit 22 to refer to the area cluster information, and acquires the population of the area cluster for which the repeat rate has been calculated. For example, when the area cluster information is in the state shown in FIG. 7, the populations of areas a, b, and c are aggregated and "2150" is acquired for the population of cluster A. Similarly, "1150", "2000", and "900" are acquired for cluster B, cluster C, and cluster D.

The repeat rate evaluation unit 118 calculates the unit population sales volume for each area cluster. For example, for cluster A, since the population is "2150" and the cluster sales volume is "505", the unit population sales volume "0.23" is calculated. Similarly, for cluster B, cluster C, and cluster D, "0.27", "0.09", and "0.09" are calculated. The repeat rate evaluation unit 118 associates the repeat rate with the unit customer sales volume and transmits the data to the user terminal 500 via the analysis result output unit 14.

FIG. 31 illustrates the correspondence between the repeat rate and the unit customer sales volume. In the figure, regarding the customer's purchasing tendency for "△△ coffee", the customer of cluster A has a large unit customer sales volume but a small repeat rate, and the customer of cluster C has a small unit customer sales volume but a large repeat rate. Is shown.

According to the first modification, the product seller can make a detailed sales strategy in terms of sales volume and repeat rate. For example, from the results shown in Fig. 31, for stores whose main customers are cluster A customers, a large amount of "△△ coffee" should be purchased at the beginning of the sale, and the number of purchases should be significantly reduced over time. Can make a decision. For stores whose main customers are Cluster C customers, the purchase amount of "△△ coffee" itself may be small, but since there are many repeat customers, "△△ coffee" will be continuously purchased so that it will not be out of stock. You can decide that you should.

In the second modification, the repeat rate evaluation unit 118 refers to the cluster customer ratio calculated by the cluster customer ratio calculation unit 94, and determines the cluster customer ratio for the area cluster whose repeat rate is equal to or higher than a predetermined threshold. Identify stores that are above the threshold of. Then, the area cluster whose repeat rate is equal to or higher than a predetermined threshold value and the store whose main customer is the resident of the area cluster are associated with each other and transmitted to the user terminal 500.

According to the second modification, the product seller can decide the sales strategy for each store according to the repeat rate of a certain product. For example, for a store whose main customer is a resident of an area cluster with a high repeat rate of a certain product, if the product is not currently sold, it will be sold immediately, and then the product will be continued so that it will not be out of stock. You can make decisions such as checking. Return to FIG.

Product appeal analysis department 120:
The product appeal degree analysis unit 120 creates product appeal degree information, which is one of the sales activity information. The product appeal degree information in the present embodiment is referred to as an index value (hereinafter, referred to as "product index value") indicating how much the product sold at a certain store and all the customers of the store support the product. ) Is the associated information. FIG. 32 is a block diagram showing a functional configuration of the product appeal degree analysis unit 120. The product appeal degree analysis unit 120 includes an analysis information acquisition unit 122, an analysis target identification unit 124, a cluster index value calculation unit 126, and a product index value calculation unit 128.

The analysis information acquisition unit 122 receives the designated data of the product and the store to be analyzed from the user terminal 500 via the analysis request detection unit 12. The analysis information acquisition unit 122 acquires the cluster sales volume of the analysis target product calculated by the cluster sales volume calculation unit 44 and the cluster customer ratio of the analysis target store calculated by the cluster customer ratio calculation unit 94.

FIG. 33 shows the cluster sales volume of the analysis target product acquired by the analysis information acquisition unit 122. The figure shows the state in which the cluster sales volume of the latest December was acquired for the products to be analyzed, "○○ coffee", "△△ coffee", and "□□ blend". Further, the cluster customer ratios of the analysis target stores acquired by the analysis information acquisition unit 122, in this example, the “Kiba store”, the “Kasai store”, and the “Funabashi store” are shown in FIGS. 26 (a) and 26 (b). The contents shown in FIG. 26 (c) will be used in the following description.

The analysis target identification unit 124 sets the cluster sales volume of the analysis target product in the analysis target cluster as an index value indicating the purchase performance of the product by the customer of the analysis target area cluster (hereinafter referred to as “analysis target cluster”). Identify. For example, in FIG. 33, the cluster sales amount “100” of the analysis target product “○○ coffee” in the analysis target cluster “cluster A” is specified.

The analysis target identification unit 124 also specifies the cluster customer ratio for the analysis target cluster of the analysis target store. For example, in FIG. 26A, the cluster customer ratio “23%” for cluster A in the analysis target store “Kiba store” is specified. Subsequently, the analysis target specifying unit 124 associates the cluster sales volume with the cluster customer ratio for each combination of the analysis target product, the analysis target cluster, and the analysis target store. In the above example, the analysis target specifying unit 124 associates the specified cluster sales volume “100” with the cluster customer ratio “23%”. Return to FIG. 32.

The cluster index value calculation unit 126 is an index showing how much the customer residing in a specific area cluster in the analysis target store supports the analysis target product by a predetermined evaluation function that positively correlates with the cluster sales volume and the cluster customer ratio. The value (hereinafter referred to as "cluster index value") is calculated. Specifically, the calculation result obtained by inputting the cluster sales amount and the cluster customer ratio specified by the analysis target specifying unit 124 into the above evaluation function is used as the cluster index value. The above evaluation function may be a known calculation method, but in the following example, it is assumed that it is a function that outputs the result of integration between the cluster sales volume and the cluster customer ratio.

In the above example, the cluster sales volume "100" in "cluster A" of "○○ coffee" and the cluster customer ratio "23%" for "cluster A" in the analysis target store "Kiba store" are evaluated by the evaluation function. Accumulate and calculate the cluster index value "23". Similarly, the cluster index value calculation unit 126 calculates the cluster index value for each combination of the analysis target product, the analysis target cluster, and the analysis target store. FIG. 34 shows the cluster index value calculated by the cluster index value calculation unit 126. In the figure, the part where the index value is "-" indicates that the customer of the corresponding store does not have a resident of the area cluster. Return to FIG. 32.

The product index value calculation unit 128 aggregates the cluster index values calculated for each area cluster by the cluster index value calculation unit 126 for each analysis target store, and calculates the product index value for the analysis target product sold at the analysis target store. calculate. For example, in FIG. 34, the cluster index values of the clusters A, B, and C of "○○ coffee" in the Kiba store are "23", "16", "270", and "48", respectively. At this time, the product index value calculation unit 128 aggregates these cluster index values and calculates the product index value “357” of “○○ coffee” at the Kiba store.

The product index value calculation unit 128 transmits the product index value at the analysis target store to the user terminal 500 for the analysis target product via the analysis result output unit 14. The product index value calculation unit 128 also accesses the product index value information holding unit 28 and records the calculated product index value in the product index value information. In the product index value column 234 of FIG. 10, the product index value calculated in the above example is recorded.

In the product appeal degree analysis unit 120, the cluster customer ratio is calculated by the store customer segment analysis unit 90 for the product for which the cluster sales volume has been calculated by the cluster sales volume calculation unit 44, regardless of the request from the user terminal 500. You may calculate the product index value in the store. Also in this case, the product index value calculation unit 128 accesses the product index value information holding unit 28 and records the calculated product index value in the product index value information.

The reason why the product index value calculated by the product appeal degree analysis unit 120 is an index value indicating how much all the customers of the analysis target store support the analysis target product will be described. As described above, the input data for calculating the product index value is the cluster sales volume of the analysis target product and the cluster customer ratio of the analysis target store. The cluster sales volume indicates the actual sales of the products to be analyzed in each area cluster, and is an index value indicating the degree of support of customers residing in each area cluster for the products to be analyzed. On the other hand, the cluster customer ratio shows the ratio of customers residing in each area cluster to the total customers of the stores to be analyzed, and shows the magnitude of the influence of the purchasing behavior of the customers in each area cluster on the stores to be analyzed. It is an index value.

The cluster index value calculated by the evaluation function that positively correlates with the cluster sales volume and the cluster customer ratio becomes larger as the cluster sales volume is larger and becomes larger as the cluster customer ratio is larger for one area cluster. From this, it can be said that the cluster index value is an index value indicating the magnitude of the influence of the support of customers residing in one area cluster on the products to be analyzed on the stores to be analyzed. Therefore, the product index value obtained by summarizing the cluster index values is an index value indicating how much all the customers of the analysis target store support the analysis target product.

With the product index value calculated by the product appeal analysis unit 120, the product seller can compare the size of customer support among the products to be analyzed in the store to be analyzed, and the purchase amount and shelving allocation of each product are appropriate. Make decisions are possible. For example, when the product index value is in the state shown in FIG. 10, the product seller should prioritize the purchase amount and shelving allocation for "Kasai store" as "□□ blend", "△△ coffee", and "○". You can make a decision in the order of "○ coffee".

Further, the cluster sales volume, which is the input of the product index value, can be obtained from the sales results of the stores other than the analysis target store because the sales volume in each store is aggregated for each area cluster. Therefore, the product index value can be calculated even when the product to be analyzed is not sold at the store to be analyzed. For example, suppose that the product index value is in the state of FIG. 10 and the Funabashi store does not sell "○○ coffee". At this time, the product seller can understand that by selling "○○ coffee" at the Funabashi store, it is more likely to be supported by customers than "□□ blend", and purchases "○○ coffee". Can make decisions.

The traditional way to identify how much a product sold in a store is instructed by a customer has been to conduct a questionnaire to the customer. However, the cost of conducting a survey to a large number of customers is high, so the number of surveys is generally small, and the survey results tend to strongly reflect the subjectivity of a small number of respondents. there were. In addition, it took a long time because of human intervention. According to the product appeal degree analysis unit 120, highly accurate product index values can be calculated in a short time based on objective data such as cluster sales volume and cluster customer ratio. Return to FIG.

Product appeal correction unit 130:
The product appeal degree correction unit 130 corrects the cluster index value so that the product index value reflecting the future sales tendency of the product is calculated based on the transition of the cluster sales volume of the product to be analyzed. FIG. 35 is a block diagram showing a functional configuration of the product appeal degree correction unit 130. The product appeal degree correction unit 130 includes a sales volume trend calculation unit 132 and an index value correction unit 134.

The sales volume trend calculation unit 132 calculates the rate of change in the cluster sales volume with the passage of time from the release date of the product for each area cluster based on the transition of the cluster sales volume of the product to be analyzed. Specifically, the sales volume trend calculation unit 132 is the cluster sales volume calculated by the cluster sales volume calculation unit 44, and acquires the cluster sales volume for each predetermined period from the release date of the product to be analyzed to the time of analysis. do. FIG. 36 shows the transition of the cluster sales volume acquired by the sales volume trend calculation unit 132. The figure shows the transition of the cluster sales volume from October to December calculated by the cluster sales volume calculation unit 44 for the analysis target products “△△ coffee” and “□□ blend”. It is assumed that both products to be analyzed started to be released in October.

The sales volume trend calculation unit 132 calculates the rate of change in the cluster sales volume for each area cluster from the transition of the cluster sales volume for the product to be analyzed. The rate of change may be determined by a known method, but since the sales volume of a product generally decreases with the passage of time, a value of 0 <rate of change <1 is typically taken. Here, the sales volume trend calculation unit 132 calculates the rate of change in the cluster sales volume shown in FIG. 36 by a recurrence formula. In the case of FIG. 36, since the recurrence formula in cluster A for “△△ coffee” is a geometric progression with a common ratio of “0.9”, the sales volume trend calculation unit 132 is a cluster of “△△ coffee”. The rate of change in A is calculated as "0.9".

FIG. 37A shows the cluster index value and the rate of change in the cluster sales volume. The uncorrected cluster index value in the figure shows the cluster index value of the product to be analyzed in the “Kiba store” calculated by the cluster index value calculation unit 126, and is the same as the value of the Kiba store cluster index value in FIG. be. The rate of change in FIG. 37 (a) shows the rate of change in the cluster index value calculated by the sales volume trend calculation unit 132 based on the transition of the cluster sales volume shown in FIG. 36. Return to FIG. 35.

The index value correction unit 134 adjusts the cluster index value calculated by the cluster index value calculation unit 126 according to the rate of change calculated by the sales volume trend calculation unit 132 for the product to be analyzed. Here, the index value correction unit 134 corrects the cluster index value by integrating the cluster index value and the rate of change in the cluster sales volume. For example, the cluster index value “46” in the cluster A of “△△ coffee” and the rate of change “0.9” in the cluster sales volume are integrated and the cluster index value is corrected to “41”.

The corrected cluster index value column of FIG. 37 (a) shows the cluster index value corrected by the index value correction unit 134. The figure shows that the cluster index value is corrected so that the cluster index value of the area cluster becomes smaller as the area cluster has a larger decrease in the cluster sales volume over time.

The product index value calculation unit 128 of the product appeal degree analysis unit 120 calculates the product index value by aggregating the cluster index values corrected by the index value correction unit 134. FIG. 37B shows a product index value before correction and a product index value after correction. The uncorrected product index value in FIG. 37 is a total of the uncorrected cluster index values in FIG. 37 (a), and the corrected product index value is a total of the corrected cluster index values in FIG. 37 (b). Is.

According to the product appeal degree correction unit 130, the cluster index value can be corrected so that the product index value reflecting the future sales tendency of the product to be analyzed is calculated by the product index value calculation unit 128. This allows merchandise sellers to make more appropriate decisions regarding product purchases and shelving allocations. For example, a product with a large decrease rate at the present time is considered to have a relatively large decrease rate in the future. The product appeal degree correction unit 130 corrects the product index value so that the product index value becomes smaller as the product has a larger decreasing tendency, so that the product index value becomes a more useful index value for the product seller.

When the uncorrected product index value shown in FIG. 37 (b) is shown, it becomes easier for the product seller to make a decision to prioritize “□□ blend” in purchasing and shelving allocation. However, by showing the corrected product index value that reflects the future sales tendency, the product seller can make a decision to give priority to "△△ coffee" in consideration of the future sales tendency. Return to FIG.

Store Sales Volume Forecasting Department 140:
The store sales volume forecasting unit 140 creates store sales volume forecasting information, which is one of the sales activity support information. The store sales volume forecast information in the present embodiment is information indicating the predicted sales volume (hereinafter, referred to as “estimated sales volume”) for a product in a certain store. FIG. 38 is a block diagram showing a functional configuration of the store sales volume prediction unit 140. The store sales volume prediction unit 140 includes a unit sales volume acquisition unit 142, a correlation degree analysis unit 144, and a prospective sales volume calculation unit 146.

FIG. 39 shows the expected sales volume at each store. The figure shows the estimated sales volume of “○○ coffee” at each store and the data related to the calculation, and will be described with reference to the figure in the following example.

The unit sales amount acquisition unit 142 is an index value of the sales amount of the product in each of the plurality of stores, and acquires the sales amount per customer of each store (hereinafter, referred to as “unit customer sales amount”). Specifically, the unit sales volume acquisition unit 142 first accesses the store customer information holding unit 306, refers to the store customer information, and acquires the total number of customers for each store. When the store customer information is in the state shown in FIG. 4, the unit sales volume acquisition unit 142 acquires the total number of customers of each store by adding up the area customer number column 334 for each store. Next, the unit sales volume acquisition unit 142 also accesses the store sales information holding unit 308, refers to the store sales information, and acquires the product sales volume in a predetermined period. Here, regarding the store sales information shown in FIG. 5, the sales volume of "○○ coffee" at each store in December will be acquired.

Finally, the unit sales volume acquisition unit 142 calculates the unit customer sales volume by calculating the ratio between the total number of customers for each store and the product sales volume at each store. For example, the total number of customers at the "Kiba store" is "2550", and the sales volume of "○○ coffee" in December is "255", so the unit customer sales volume in this case is "0.1". Become. The unit customer sales volume in FIG. 39 shows the unit customer sales volume of “○○ coffee” in each store calculated by the unit sales volume acquisition unit 142. The "Makuhari store" and "Akita store", which have no value in unit customer sales volume, indicate that "○○ coffee" is not sold. Return to FIG. 38.

The correlation degree analysis unit 144 accesses the product index value information holding unit 28, refers to the product index value information, acquires the product index value of the product at each store, and also acquires the unit sales volume acquisition unit for the product at each store. Acquire the unit customer sales volume calculated by 142. The product index value in FIG. 39 is the product index value of “○○ coffee” at each store.

Next, the correlation degree analysis unit 144 uses a regression analysis method from the product index value and the unit customer sales volume at each of a plurality of stores, with the product index value and the unit customer sales volume of the product as explanatory variables and objective variables, respectively. , Calculate a regression equation showing the magnitude of the correlation between the product index value of the product and the unit customer sales volume.

For example, the correlation degree analysis unit 144 calculates the following regression equation by performing a regression analysis on the product index values and the unit sales volume in the “Kiba store”, “Kasai store”, and “Funabashi store” in FIG. 39.
Unit sales volume = 0.0002 x product index value + 0.0231
The regression coefficient "0.0002" in this regression equation indicates the magnitude of the influence of the product index value on the unit sales volume. In this example, a simple regression analysis is shown in which only the product index value is used as an explanatory variable, but the correlation degree analysis unit 144 adds a plurality of explanatory variables such as a resident index value and other statistical information to perform a multiple regression analysis. Or the variable selection described above may be carried out. Return to FIG. 38.

The estimated sales volume calculation unit 146 is a store (hereinafter, "prediction target store") for which the sales volume of a specific product (hereinafter, referred to as "prediction target product") is predicted via the analysis request detection unit 12. The designated data of) is received from the user terminal 500. The expected sales volume calculation unit 146 accesses the product index value information holding unit 28, refers to the product index value information, and acquires the product index value for the forecast target product in the forecast target store.

Next, the forecast sales volume calculation unit 146 inputs the product index value in the forecast target store into the regression equation calculated by the correlation degree analysis unit 144, and the unit customer sales volume of the forecast target product in the forecast target store (hereinafter, It is called "estimated unit sales volume"). The estimated unit sales volume in FIG. 39 shows the estimated unit sales volume obtained by inputting the product index value of XX coffee in each store into the above regression equation.

The forecast sales volume calculation unit 146 calculates the sales volume of the forecast target product (hereinafter referred to as “estimated sales volume”) at the forecast target store based on the forecast unit sales volume. Here, the estimated sales volume is calculated by integrating the total number of customers of the forecast target store and the expected unit sales volume for the forecast target product. The estimated sales volume in FIG. 39 shows the integrated result of the estimated unit sales volume of XX coffee in each store and the total number of customers in each store in FIG. The total number of customers at the "Makuhari store" is "5000", and the total number of customers at the "Akita store" is "1000".

The estimated sales amount calculation unit 146 transmits the estimated sales amount of the forecasted product in the forecasted store to the user terminal 500 via the analysis result output unit 14. FIG. 40 shows the actual sales volume and the forecast of the sales volume at each store in association with each other. The plain sales volume in the figure shows the actual sales volume of "○○ coffee" at each store, and the shaded sales volume shows the expected sales volume of "○○ coffee" at each store.

The conventional method of predicting the sales volume of a product is to actually sell the product to be predicted at a sample store and predict the sales volume at another store from the sales tendency. However, if the number of sample stores is small and the sales period is short, the accuracy of the analysis is low, and if the number of sample stores is large and the sales period is long to improve the accuracy, the time and research cost will rise. There was a problem that it would end up. According to the store sales volume forecasting unit 140, a reasonable sales forecast can be realized in a short time by calculating the expected sales volume based on the product index value.

The reason why a reasonable expected sales volume can be calculated by the store sales volume forecasting unit 140 will be described. As described above, the store sales volume prediction unit 140 first calculates a regression equation showing the correlation between the product index value and the unit customer sales volume. The product index value is an index value indicating how much the customer of each store supports the forecast target product. On the other hand, the unit customer sales volume is the ratio of the sales volume of the forecasted product in each store to the total number of customers in each store. Since the set of product index value and unit sales volume for each store that sold the forecast target product can be treated equally, the correlation between the product index value as the index value and the unit customer sales volume as the actual value by regression analysis. The regression equation showing is can be calculated.

Next, the store sales volume prediction unit 140 calculates the expected unit sales volume by inputting the product index value of the forecast target store into the regression equation calculated for the forecast target product. Since the estimated unit sales volume is a value per customer, the estimated sales volume can be calculated by adding up with the total number of customers in the forecast target store. This estimated sales volume reflects the cluster sales volume, which indicates the sales performance at other stores, based on the cluster customer ratio of the forecast target stores. In addition, residents with similar living environments and ways of thinking live in the same area cluster. Therefore, it can be said that the estimated sales volume is a sales volume that can be realized even at the forecast target stores, based on the sales results at other stores.

Further, as described above, the product index value is an index value calculated from objective data based on actual results, and the unit customer sales volume is also objective data based on actual results. Therefore, the calculation is highly accurate because it is based on the sales results at a plurality of stores, and the calculation can be realized in a short time.

Furthermore, according to the store sales volume forecasting unit 140, the product seller compares the actual sales record with the estimated sales volume, as shown in the "Kiba store", "Kasai store", and "Funabashi store" in FIG. can. As a result, the product seller can understand that the sales volume of "○○ coffee" at the "Funabashi store" can be increased, and can make a decision to execute sales activities such as sales promotion measures for the "Funabashi store".

Furthermore, as shown in FIG. 40, the estimated sales volume can be provided to the product sellers at the “Makuhari store” and the “Akita store” that do not sell “○○ coffee”. As a result, the product seller can decide to sell "○○ coffee" even at the "Makuhari store", for example. That is, the product seller can specify an appropriate store to sell the product and purchase an appropriate amount of the product in the store.

The present invention has been described above based on the first embodiment. It is understood by those skilled in the art that this embodiment is an example, and that various modifications are possible for each of these components and combinations of each processing process, and that such modifications are also within the scope of the present invention. be.

[Second Embodiment]
Similarly to the first embodiment, the second embodiment also proposes a technique for creating information for supporting the sale of products in stores and the like. The following "area type" corresponds to the area cluster in the first embodiment.

FIG. 41 shows the configuration of the IDPOS analysis system 600 of the second embodiment. The IDPOS analysis system 600 is an information processing system including an IDPOS-DB602, an area DB604, and an analysis device 606. The IDPOS-DB602, the area DB604, and the analyzer 606 are connected via a communication network including a LAN, a WAN, the Internet, and the like.

IDPOS-DB602 is a database that holds IDPOS data and updates IDPOS data on a daily basis. IDPOS data is data in which ID information for identifying a customer is added to POS (Point of Sales) data indicating what, when, how many, and how much was sold, and can be said to be "sales record data". FIG. 42 shows an example of IDPOS data. The IDPOS data of the embodiment includes the date, the product for sale, and the selling price as general POS data, and further includes the customer ID and address as customer identification information (for example, information at the town street level indicating the customer's residential area). include. The IDPOS data may further include an item of sales quantity.

Returning to FIG. 41, the area DB 604 defines an area type in which a plurality of geographically divided areas (town chome level) are grouped by areas having similar inhabitant attributes (hereinafter, also referred to as “area data”). ). FIG. 43 shows an example of area data. The area data of the embodiment includes a plurality of addresses (for example, a town chome level area on a map), an area type (for example, about 20), and the number of households existing at each address (area). The number of households includes residents (households) who are not given a customer ID, and includes, for example, residents (households) who do not hold a point card or a credit card. The number of households may be set based on the results of the national census and the like.

The area DB 604 further holds area type data in which a plurality of area types are associated with attribute information and index values (for example, the resident attribute and the resident index value shown in FIG. 7) relating to the inhabitants of each area type. Further, the area type data includes the total number of customer IDs in each area type. The total number of customer IDs can be said to be the number of residents who have customer IDs, for example, the number of residents who have been given customer IDs based on point cards, credit cards, and the like.

Returning to FIG. 41, the user terminal 608 is an information terminal operated by a user of the IDPOS analysis system 600. The user terminal 608 may be, for example, a PC, a tablet terminal, or a smartphone. The user terminal 608 is connected to the analyzer 606 via a communication network 610 including LAN, WAN, the Internet, and the like.

The analysis device 606 is an information processing device that analyzes IDPOS data and provides the analysis result to the user terminal 608. FIG. 44 is a block diagram showing a functional configuration of the analyzer 606 of FIG. 41. The analysis device 606 includes an area data acquisition unit 620, a sales record acquisition unit 622, an analysis unit 624, and an analysis result output unit 626.

A computer program including a plurality of modules corresponding to the plurality of functional blocks of FIG. 44 may be stored in a recording medium, distributed, and installed in the analyzer 606. Further, the analyzer 606 may download and install the computer program from an external server. The CPU of the analyzer 606 may exert the function of each functional block by executing the computer program. Further, the analyzer 606 may include a communication unit (not shown) that communicates with the external device using a predetermined communication protocol, and the functional block of FIG. 44 may transmit / receive data to / from the external device via the communication unit. ..

The area data acquisition unit 620 acquires area data and area type data from the area DB 604. The sales record acquisition unit 622 acquires IDPOS data from IDPOS-DB602. At least one of the area data, the area type data, and the IDPOS data may be stored in a storage (not shown) in the analyzer 606 instead of the external device of the analyzer 606. At least one of the area data acquisition unit 620 and the sales record acquisition unit 622 may read the area data, area type data, or IDPOS data stored in the local storage from the storage into the main memory.

The analysis unit 624 aggregates the sales performance indicated by the IDPOS data acquired by the sales performance acquisition unit 622 for each area type, and generates analysis information about the product as sales activity support information based on the aggregation result for each area type. .. The analysis result output unit 626 is the analysis information generated by the analysis unit 624, and in the embodiment, the product profile information or the sales potential information is provided to the user terminal 608.

The analysis unit 624 includes an area type addition unit 628, an area type addition IDPOS holding unit 630, a request reception unit 632, an aggregation unit 634, a product profile generation unit 636, and a potential information generation unit 638.

The area type assigning unit 628 specifies the area type corresponding to the address for each record of the IDPOS data acquired by the sales record acquisition unit 622, and the data in which the area type is added to each record of the IDPOS data is the area type addition IDPOS data. Generate as. The area type addition unit 628 stores the generated area type addition IDPOS data in the area type addition IDPOS holding unit 630.

FIG. 45 shows an example of the area type additional IDPOS data held in the area type additional IDPOS holding unit 630. The area type addition IDPOS data in the figure is obtained by adding the area type in FIG. 43 to each record of the IDPOS data in FIG. 42 with the address as a key.

Returning to FIG. 44, the request receiving unit 632 receives the data (hereinafter referred to as “analysis request”) for requesting the creation of analysis information transmitted from the user terminal 608. The analysis request includes the specification of the product to be analyzed, the period to be analyzed, and the analysis type (product profile analysis or sales potential analysis in the embodiment). The aggregation unit 634 aggregates the area type additional IDPOS data related to the target product specified in the analysis request, and the area type additional IDPOS data including the date during the target period specified in the analysis request for each area type.

FIG. 46 shows an example of the aggregation result data by the aggregation unit 634. Each record of the aggregation result data includes an area type, a total number of customer IDs, a sales quantity, an average sales quantity, an average sales price, and a repeat rate. The "sales quantity" and "sales price" in FIG. 46 can also be referred to as "purchase quantity" and "purchase price" from the customer's point of view. Although FIG. 46 shows up to area type 10, actually, records for the number of area types (for example, area types 1 to 20) are generated.

The aggregation unit 634 sets the total number of customer IDs of each area type based on the area type information acquired by the sales record acquisition unit 622. The aggregation unit 634 may count customers whose product purchase frequency is equal to or higher than a predetermined threshold value as the target of the total number of customer IDs. For example, the aggregation unit 634 targets customers who have purchased some product more than a predetermined number of times (for example, once or twice or more) within the target period (or within a certain period including outside the target period) for the total number of customer IDs. May be counted as. In this way, by setting a lower limit of the product purchase frequency and limiting the customers to be included in the analysis population, it becomes easier to obtain a more realistic index value.

The aggregation unit 634 specifies a record (hereinafter referred to as “target record”) indicating the sale of the target product among the records of the area type additional IDPOS data within the target period. The aggregation unit 634 derives the number of target records for each area type (the total value of the sales quantity when each record includes the sales quantity) as the sales quantity. Further, the aggregation unit 634 derives the average value of the sales quantity per household (in other words, the purchase quantity) in the unit period as the average sales quantity. For example, the average sales quantity may be calculated by the following formula.
Average sales volume (pieces / household / month) = Sales volume (for one year) ÷ Total number of customer IDs ÷ 12

The aggregation unit 634 derives the average value of the selling price indicated by the target record for each area type, and sets the average value as the average selling price of each area type. The average selling price may be, for example, a weighted average value obtained by weighting the selling price by the selling quantity. Further, the aggregation unit 634 refers to the target record for each area type, and when the same customer (same ID) purchases the same product on different dates, the number of repeats is set to 1 and the number of repeats is set for each customer (ID). Count. The aggregation unit 634 calculates the average value of the number of repeats for each customer, and sets the average value as the repeat rate. If the date of the target record includes the time, the number of repeats may be incremented when the same product is purchased at different times.

The product profile generation unit 636 generates product profile information which is information that visualizes the purchaser characteristics of the product when the analysis type specified in the analysis request is the product profile analysis. The product profile generation unit 636 outputs the generated product profile information to the analysis result output unit 626.

FIG. 47 shows an example of product profile information. In the example of the figure, the horizontal axis represents the area type, and the vertical axis represents the product profile information with the average sales volume in FIG. 46. The product profile generation unit 636 may generate product profile information in which the horizontal axis is the area type and the vertical axis is the average selling price or the repeat rate in FIG. 46, according to the designation in the analysis request. The product profile information makes it easier for the user to understand the product purchase amount for each cluster type, the purchase price for each cluster type, and the repeat rate for each cluster type. In this way, the analyzer 606 can provide the user with useful information for purchasing products, setting prices, allocating shelves, and the like.

Further, a plurality of target products may be specified in the analysis request. In this case, the aggregation unit 634 aggregates the area type additional IDPOS data in units of each of the plurality of target products, and generates a plurality of aggregation result data. The product profile generation unit 636 generates product profile information including a plurality of graphs corresponding to a plurality of target products based on a plurality of aggregation result data. For example, by designating a plurality of target products having different prices from each other, the user can grasp the influence of the price on the sales quantity of the goods, and can grasp the cluster type having a large influence of the price. Further, the user can grasp the influence of the capacity on the sales quantity of the product by designating a plurality of target products having different capacities, and can grasp the cluster type having a large influence of the capacity.

Further, a plurality of target periods may be specified in the analysis request. In this case, the aggregation unit 634 aggregates the area type additional IDPOS data corresponding to each of the plurality of target periods, and generates a plurality of aggregation result data. The product profile generation unit 636 generates product profile information including a plurality of graphs corresponding to a plurality of target periods based on a plurality of aggregation result data. This makes it easier to understand changes in the product profile over time. Such time-series analysis is useful, for example, when analyzing the effects of sales promotion measures and identifying areas with strong product needs at the initial stage.

FIG. 48 shows an example of product profile information. Here, it is assumed that the sales promotion measure was implemented in October, and the user specifies September as the first target period, October as the second target period, and November as the third target period. I will do it. The product profile information in FIG. 48 is shown by superimposing the average sales volume graph (dotted line) in September, the average sales volume graph (solid line) in October, and the average sales volume graph (two-dot chain line) in November. There is. According to the product profile information in FIG. 48, it is easy for the user to strongly appeal to the area type 4 consumers and the area type 8 consumers that the sales promotion measures in October (for example, mass advertisements and in-store campaigns). Can be grasped.

Furthermore, the analysis request may further specify the target area for analysis. For example, the target area may be designated with various particle sizes such as country, prefecture, city, ward, town, village, and town chome. Further, a specific point (for example, a specific point on the map) may be designated, and a range (a circle having a radius of 1 KM, etc.) from that point to the target area may be designated. In this case, the aggregation unit 634 aggregates the area type additional IDPOS data including the address in the target area, and generates the aggregation result data. As a result, the accuracy of the product profile information for the product whose sales area is limited can be improved. In addition, the designation of the target area is suitable for determining the effect of the sales promotion measures by the above time series analysis. When regional sales promotion measures (local CM, special sales at stores, etc.) are implemented, specify the area where the sales promotion measures are implemented as the target area, and compare the product profile information before and after the implementation of the sales promotion measures. This makes it easier to understand the effects of sales promotion measures.

Returning to FIG. 44, the potential information generation unit 638 generates sales potential information which is information that visualizes the potential sales potential of the product when the analysis type specified in the analysis request is the sales potential analysis. Specifically, the potential information generation unit 638 has a sales potential including an estimated sales volume (estimated value) of the target product in each area based on the average sales volume of each area type and the number of households in each area. Generate information. The potential information generation unit 638 outputs the generated sales potential information to the analysis result output unit 626.

FIG. 49 shows an example of sales potential information. Here, it is assumed that "ABC curry sweet" is designated as the target product in the analysis request, and eight areas (town chome level) are designated as the target area. In this case, the potential information generation unit 638 derives the sales potential value in each of the eight areas.

For example, in "Fujimi (1) (1-chome), Chiyoda-ku, Tokyo" in FIG. 43, the number of households is 100 and the area type is 1. Further, in FIG. 46, the average sales volume of the target product in the area type 1 is 0.56. In this case, the potential information generation unit 638 calculates the sales potential value of the target product in "Fujimi (1), Chiyoda-ku, Tokyo" by calculating (the number of households in the area x the average sales volume in the corresponding area type). Estimated to be "56". Estimate the sales potential value for other areas as well.

Sales potential analysis can provide estimates of the originally expected sales volume in various areas, regardless of whether or not they are actually selling products. This can support the judgment of the product lineup and the shelving allocation in the store. In addition, the user can easily determine the necessity of sales promotion measures by comparing the actual sales volume with the estimated value. The estimated value of the sales quantity provided by the analyzer 606 is based on the sales performance in other areas having similar resident attributes to the area concerned, and a certain degree of accuracy is guaranteed.

Further, in the sales potential analysis as well as the product profile analysis, the target area of the analysis may be specified with various particle sizes such as a country, a prefecture, a municipality, and a town chome. The potential information generation unit 638 derives the sales potential value of the target product for each of one or more areas (town chome level) included in the target area, and totals the sales potential values of each area included in the target area. May derive the sales potential value of the target area. This makes it possible to obtain sales potential values in areas of various particle sizes.

Furthermore, in the analysis request, a first region to be aggregated and a second region to be analyzed (different from the first region) may be specified. In this case, the aggregation unit 634 aggregates the area type additional IDPOS data in which the address corresponding to the first area is set, and generates the aggregation result (for example, FIG. 46). The potential information generation unit 638 may derive the sales potential value of the target product for one or more areas corresponding to the second region. As a result, for example, it is possible to obtain a sales potential value in an area where the product has not been released, based on the sales performance in the area where the product is actually sold. For example, it is possible to obtain an estimated value of the sales volume when a product sold only in the Kanto area is newly sold in the Kinki area.

Furthermore, the analysis request may specify the store to be analyzed. In this case, the analyzer 606 holds store information including the correspondence between the store and the trade area (typically one or more areas around the store). For example, the first store "AAA Mart Fujimi Store" may be associated with "Fujimi, Chiyoda-ku, Tokyo (1)" and "Fujimi, Chiyoda-ku, Tokyo (2)" as a commercial area. Further, the second store "AAA Mart Hitotsubashi store" may be associated with "Hitotsubashi, Chiyoda-ku, Tokyo (1)" and "Hitotsubashi, Chiyoda-ku, Tokyo (2)".

FIG. 50 shows an example of sales potential information. Here, the potential information generation unit 638 sets the sales potential values of "ABC Curry Sweet" of "AAA Mart Fujimi Store" as "Fujimi (1), Chiyoda-ku, Tokyo" and "Fujimi, Chiyoda-ku, Tokyo (2)" in FIG. ) ”, Which is the total of the sales potential values, is calculated as“ 123 ”. In addition, as the sales potential value of "ABC Curry Sweet" of "AAA Mart Hitotsubashi Store", the total sales potential value of "Hitotsubashi, Chiyoda-ku, Tokyo (1)" and "Hitotsubashi, Chiyoda-ku, Tokyo (2)" in Fig. 49. Is calculated as "94".

Furthermore, a plurality of target products may be specified in the analysis request. In this case, the aggregation unit 634 aggregates the area type additional IDPOS data in units of each of the plurality of target products, and generates a plurality of aggregation result data. The potential information generation unit 638 generates a plurality of sales potential information corresponding to a plurality of target products based on a plurality of aggregation result data (see, for example, FIG. 50).

The operation of the analyzer 606 with the above configuration will be described.
FIG. 51 is a flowchart showing the operation of the analyzer 606. The area data acquisition unit 620 periodically reads the area data and the area type data from the area DB 604, such as when the analyzer 606 is started (S10). The sales record acquisition unit 622 reads IDPOS data from IDPOS-DB602 periodically, such as once a day (S12). The area type addition unit 628 generates the area type addition IDPOS data in which the area type is added to each record of the IDPOS data, and stores the area type addition IDPOS data in the area type addition IDPOS holding unit 630 (S14).

The user inputs the analysis type, the target product, the target period, the target area, etc. in the analysis of the IDPOS data to the user terminal 608. The user terminal 608 transmits an analysis request including designation of an analysis type, a target product, a target period, a target area, and the like to the analysis device 606. When the request receiving unit 632 of the analyzer 606 receives the analysis request from the user terminal 608 (Y in S16), the aggregation unit 634 corresponds to the target product, the target period, and the target area from the area type additional IDPOS holding unit 630. Area type to be added IDPOS data is extracted, and the extracted data is aggregated by area type (S18).

When the analysis type is product profile analysis (Y in S20), the product profile generation unit 636 generates product profile information based on the aggregation result by the aggregation unit 634 (S22). The analysis result output unit 626 transmits the product profile information to the user terminal 608 that is the source of the analysis request (S24). When the analysis type is sales potential analysis (N in S20, Y in S26), the potential information generation unit 638 estimates the sales quantity of the target product in the target area based on the aggregation result by the aggregation unit 634, and the estimation result. (S28) to generate sales potential information including. The analysis result output unit 626 transmits the sales potential information to the user terminal 608 that is the source of the analysis request (S30).

The product profile information or sales potential information may be stored in a predetermined storage area of the analyzer 606. The user terminal 608 may access the analyzer 606 via the communication network 610 and browse the product profile information or the sales potential information stored in the storage area. If the analysis request has not been received (N in S16) or the analysis type is neither product profile analysis nor sales potential analysis (N in S26), the subsequent processing is skipped and the flow in this figure ends. ..

The present invention has been described above based on the second embodiment. It is understood by those skilled in the art that this embodiment is an example, and that various modifications are possible for each of these components and combinations of each processing process, and that such modifications are also within the scope of the present invention. be.

The analyzer 606 has generated analytical information about the sale of goods, but the subject of analysis by the analyzer 606 includes various tangible and intangibles to be sold, such as services. In addition, the medium for selling goods or services is not limited to actual stores. For example, goods or services sold by mail order or electronic commerce are also subject to analysis by the analyzer 606.

Any combination of the embodiments and modifications described above is also useful as an embodiment of the present invention. The new embodiments resulting from the combination have the effects of the combined embodiments and variants. It is also understood by those skilled in the art that the functions to be fulfilled by each of the constituent elements described in the claims are realized by a single component or a cooperation thereof shown in the embodiments and modifications. ..

600 IDPOS analysis system, 606 analyzer, 620 area data acquisition unit, 622 sales record acquisition unit, 624 analysis unit, 626 analysis result output unit, 634 aggregation unit, 636 product profile generation unit, 638 potential information generation unit.

Claims (1)

The first acquisition unit that acquires area type data in which multiple geographically divided areas are grouped by areas with similar inhabitant attributes, and
The second acquisition unit that acquires the sales performance data including the data of the area where the customer resides, which is the data indicating the sales performance of the product or service to the customer.
An analysis unit that aggregates sales performance data of the product or service for each area type and generates analysis information about the product or service based on the aggregation result for each area type.
Equipped with
The sales performance data includes the ID of the customer who purchased the product or service, the data of the area where the customer resides, and the data indicating the product or service.
The analysis unit counts the total number of customers in each area type for customers whose frequency of purchasing some goods or services is equal to or higher than a predetermined threshold, and determines the sales volume in each area type for the products or services to be analyzed. By dividing by the total number of customers in each area type, the average sales volume in each area type is derived .
The analysis unit calculates the product of the average sales volume of the product or service to be analyzed in the area type to which the area to be analyzed belongs and the number of households in the area to be analyzed to be the product of the analysis target in the area to be analyzed. An analyzer characterized by deriving as a forecast value for the sales volume of a product or service .

Images