Sales prediction using client value represented by three index axes as criteron

• US 20040138958A1
• Filed: 11/21/2003
• Published: 07/15/2004
• Est. Priority Date: 05/31/2001
• Status: Active Grant

• Current Assignee: Dentsu Tec Incorporated (Dentsu Group, Inc.)
• Original Assignee: Dentsu Tec Incorporated (Dentsu Group, Inc.)
• Inventors: Kawai, Tetsuya, Watarai, Koji
• Granted Patent: US 7,577,579 B2
• Time in Patent Office: Days
• US Class Current: 705/26
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities

Abstract

One object of the present invention is to provide a method for determining the current value and future value of customers who purchase specific merchandise, and the resources thereof, in order to provide data by which to select effective sales-promotion investments suitable for such customers, and for predicting sales according to the target and conditions of the investments. A sales-predicting method that classifies customers are into customer-value cells that determine the magnitude of current and future customer value and resources for future customer value, based on purchase data for specific merchandise and using three axes, including a purchase-amount index (first axis), a user-type index (second axis), and a customer-purchase-relevance index (third axis), and that measures changes in the customer-asset cells over time and changes due to sales-promotion investments, and that simulates sales by quantifying causal relationships between sales-promotion investments and sales.

Claims

1. (amended) A sales-predicting method based on triple-axis mapping of customer value, employing a customer-value-analyzing computer system of an institute that collects and analyzes purchase-history data in a database;

said computer system comprising a processing server having a group of analyzing programs for analyzing customer value, a database server for managing various databases, a web server, and a data input/output terminal connected to a communication line;

said database server comprising at least a customer-purchase-history database for accumulating purchase-history data including at least (1) a customer name or code, (2) a product code, (3) the quantity of items purchased, (4) the amount expended on purchases, and (5) the time of purchase, with said purchase-history data collected through transactions of sales outlets for the merchandise, electronic transactions conducted via the Internet, and direct transactions including transactions by telephone and mail when customers purchase products of a specific manufacturer or brand in a specific market; and

a total manufacturer/brand customer-purchase-history database for accumulating purchase-history data including product-purchase history of membership credit cards having a common ID and affiliated with a plurality of businesses in various industries, as well as data obtained through questionnaires and/or marketing approaches at sales outlets, questionnaires and/or marketing approaches in electronic transactions via said web server, questionnaires and/or marketing approaches by direct mail, e-mail, and telephone, and customer data reported by sales clerks;

the sales-predicting method comprising the steps of the processing server;

creating a purchase-amount index (first axis) of cells by searching said customer-purchase-history database at specified periods and classifying customers into a plurality of classifications in order of the amounts expended on purchases or the quantity of items purchased, based on purchase-history data extracted during the search;

creating a user-type index (second axis) of cells by searching said customer-purchase-history database at said specified periods and classifying customers into a plurality of classifications according to user type as determined by combinations of merchandise from customers who purchase the majority of types of the merchandise to customers who purchase 0 or 1 type of the merchandise, by combining a plurality of types of merchandise by specific manufacturers or brands purchased in each specified period;

creating a current customer-value map by dividing customers into cells according to a product of the plurality of classifications in said first index and said second index for analyzing the current value of customers in each cell;

classifying customers by relevance by searching said total manufacturer/brand customer-purchase-history database at said specified periods and classifying customers by purchase amount into a plurality of categories according to purchase monetary sums or quantities of items purchased based on data abstracted from said total manufacturer/brand customer-purchase-history database;

classifying customers by relevance by searching said total manufacturer/brand customer-purchase-history database at said specified periods and classifying customers by user type into a plurality of categories according to combinations of merchandise purchased during said specified periods, from customers who purchase the majority of the types of merchandise to customers who purchase 0 or 1 type of the merchandise, based on data extracted during the search;

creating a customer-purchase-relevance index (third axis) based on a product of the plurality of classifications from each of the two relevance classifying steps; and

creating a future customer-value map by dividing customers into cells formed by a product of each plurality of classifications in said user-type index (second axis) and said customer-purchase-relevance index (third axis) to analyze the future value for customers in each cell;

the sales-predicting method further comprising the steps of;

sorting customers into customer-value cells that determine the magnitude of current and future customer value and resources for future customer value based on specific-merchandise purchase data collected in said database and using a combination of said three axes, including said purchase-amount index (first axis), said user-type index (second axis), and said customer-purchase-relevance index (third axis);

measuring changes in the customer-value cells over time and changes due to sales-promotion investments; and

simulating sales by quantifying causal relationships between sales-promotion investments and sales.

2. (amended) A sales-predicting method based on triple-axis mapping of customer value as recited in claim 1, and whereby:

said database further comprises a marketing-approach response database for accumulating responses to said marketing approaches; and

said analyzing programs of said processing server comprise the steps of;

performing various marketing approaches and/or sales correspondence including direct mail or e-mail sent to customers of each cell a plurality of times over an interval equal to or greater than said specified period, in addition to said two relevance classifying steps for creating said customer-purchase-relevance index (third axis);

accumulating responses to sales correspondence in said marketing-approach response database and generating the customer-purchase-relevance index (third axis) for sorting customers according to cell by quantifying the responses; and

creating a future customer-value map according to cells in which customers are sorted using the three axes for analyzing the future value of customers in each cell.

10. (amended) A sales-predicting method based on triple-axis mapping of customer value as recited in any of claims 1, 2, 5, or 6, whereby the specific markets include at least cosmetics, soaps, clothing and apparel, decorations, handbags and travel luggage, home appliances, computers and peripherals, services including distribution and transport, and alcoholic beverages such as beer, wine, brandy, and whiskey.

3. (deleted)

4. (deleted)

5. (amended) A sales-predicting method based on triple-axis mapping of customer value employing a customer value-analyzing computer system of an institute for collecting and analyzing purchase-history data in a database, with the purchase-history data including at least (1) a customer name or customer code, (2) a product code, (3) the quantity of items purchased, (4) the amount expended on a purchase, and (5) the time of purchase, with said data collected through transactions of sales outlets for the merchandise, electronic transactions conducted via the Internet, and direct transactions including transactions by telephone and mail when customers purchase products of a specific manufacturer or brand in a specific market;

said computer system comprising a processing server having a group of analyzing programs for analyzing customer value, a database server for managing various databases, a web server, and a data input/output terminal connected to a communication line;

said database server comprising a customer-purchase-history database for accumulating said purchase-history data concerning customers that purchase merchandise of a specific manufacturer or brand; and

a total manufacturer/brand customer-purchase-history database for accumulating customer-purchase-history data extrapolated from amounts expended on purchases, quantities of items purchased, and other purchase details for products of all manufacturers and brands, including those of other manufacturers and brands, in a specific market, with the extrapolated purchase-history data including product-purchase history of membership credit cards having a common ID and affiliated with a plurality of businesses in various industries, as well as data obtained through questionnaires and/or marketing approaches at sales outlets, questionnaires and/or marketing approaches in electronic transactions via the web server and direct transactions via the data input/output terminal, and questionnaires and/or marketing approaches by direct mail, e-mail, and telephone;

the sales-predicting method comprising the steps, according to said analyzing programs, of said processing server;

searching said customer-purchase-history database, dividing customers into m×

n cells according to two axes, including a purchase-amount index for classifying customers into a plurality of categories m in order of purchase sum or quantity based on data stored in the customer purchase-history database for each specified period, and a user-type index for classifying customers by user type into a plurality of categories n according to combinations of merchandise purchased during each specified period, with customers being classified in a range from customers who purchase a majority of the types of merchandise to customers who purchase 0 or 1 type of the merchandise, and generating a current customer-value map based on these cells so as to determine the current customer value in each cell; and

searching the total manufacturer/brand customer-purchase-history database, dividing customers into cells according to a customer-purchase-relevance index (third axis) for cells classified according to a product of categories for customer-purchase-amount relevance and said user-type relevance in the specified periods, and creating a future customer-value map based on said third axis to determine the future value of customers in each cell;

the sales-predicting method further comprising the steps of;

determining the magnitude of current and future customer value and resources thereof for products of specific manufacturers or brands based on data in the customer-purchase-history database and total manufacturer/brand customer-purchase-history database, providing data for selecting effective sales-promotion investments suited to the customers, and providing a method for predicting sales suited to the target and conditions of the investments.

6. (amended) A sales-predicting method based on triple-axis mapping of customer value capable of finding effective marketing approaches and improving overall sales by selecting not only customers having current value, but also customers having high future-value potential, using a customer value-analyzing computer system of an institute for overseeing purchase-history data, with said data including at least (1) a customer name or customer code, (2) a product code, (3) the quantity of items purchased, (4) the amount expended on purchases, and (5) the time of purchases, with all said data collected through transactions of sales outlets for the merchandise, electronic transactions conducted via the Internet, or direct transactions including transactions by telephone and mail when customers purchase products of a specific manufacturer or brand in a specific market;

said computer system comprising at least a processing server having a group of analyzing programs for analyzing customer value, a customer-purchase-history database for accumulating purchase-history data on products of specific manufacturers or brands required for the analyzing programs, and a total manufacturer/brand customer-purchase-history database for accumulating data for quantifying the purchase behavior of customers for merchandise of all manufacturers or brands in a specific market based on questionnaires and/or marketing approaches at sales outlets, in electronic transactions, and in direct transactions, customer data reported by sales clerks, and product-purchase history of membership credit cards with a common ID that are affiliated with businesses in various industries;

the sales-predicting method comprising the steps of said processing server;

classifying customers according to a purchase-amount index (first axis), whereby the amount of money expended on merchandise of specific manufacturers or brands in specific markets is determined by the amount expended on purchases, the quantity of items purchased, the volume of items purchased, or the like of merchandise purchased by customers or members of household units in each specified period, and whereby customers are classified into a plurality of categories m based on this data, which is stored in the customer-purchase-history database, and with the categories including at least a heavy-purchase classification, a medium-purchase classification, and a light-purchase classification;

classifying customers according to a user-type index (second axis), whereby customers who purchase products of specific manufacturers or brands are classified into a plurality of categories n based on data in the purchase-history database, with the categories including an upper category for customers who purchase all types of merchandise of the manufacturers or brands in the specified period, an upper-middle category for customers who purchase a majority of the types of merchandise, a middle category for customers who purchase one-half or more of the types of merchandise, an average category for customers who purchase a one-half or an average amount of the merchandise, and a lower category for customers who purchase 0 or 1 type of products;

creating a current customer-value map by dividing the customers into m×

n cells according to the customer-amount index and the user-type index and generating numerical data indicating the structure and purchasing status of customer groups for each cell, based on the customer-purchase-history database;

classifying customers by relevance, whereby customers are classified into a plurality of categories m according to purchase amounts ordered by amounts expended or quantities of purchases within a specified period for merchandise of all manufacturers or brands, based on data abstracted from the total manufacturer/brand customer-purchase-history database, in order to determine the future value of customers in each cell;

classifying customers by relevance, whereby customers are classified by user type into at least n categories according to combinations of merchandise purchased during the same period, from customers who purchase a majority of the types of the merchandise to customers who purchase 0 or 1 type of the merchandise, based on data stored in the total manufacturer/brand customer-purchase-history database, and further dividing the n categories into at least m×

n cells according to purchase-amount classifications for each category;

creating a customer-purchase-relevance index (third axis) by conducting such marketing approaches and/or sales correspondence as questionnaires and/or marketing approaches at sales outlets, questionnaires and/or marketing approaches in electronic transactions and direct transactions, questionnaires and/or marketing approaches by direct mail, e-mail, and telephone, and customer data reported by sales clerks, quantifying responses to these marketing approaches, and updating the content of the cells for each specified period;

creating a future customer-value map by dividing customers into cells formed by a product of each plurality of categories in said third axis and said second axis; and

determining a relationship between the magnitude of current and future customer value and resources thereof for specific products and the effects of various marketing approaches, providing data for selecting effective sales-promotion investments suited to the customers, and providing a method for predicting sales suited to the target and conditions of the investments.

7. (amended) A sales-predicting method based on triple-axis mapping of customer value as recited in claim 6, whereby customers are arranged in order from largest purchase amount to smallest in the step for classifying customers according to purchase amount, and whereby customers are placed in one of three categories, including a heavy-purchase classification (H) for customers who account for about 50% of the total amount expended on purchases, a medium-purchase classification (M) for customers who account for about 30% of the total, and a light-purchase classification (L) for customers accounting for the remaining 20%.

8. (amended) A sales-predicting method based on triple-axis mapping of customer value as recited in claim 6 or claim 7, whereby numerical data indicating the structure and purchase condition of customer groups in each cell includes the customer or household component percentage, the component percentage of purchase sums, quantities, or volume, the amount expended per person, the number of transactions per person, and the number of customers, and whereby sales for the near future can be predicted by creating maps for consecutive time periods in the current customer-value map-creation step and comparing data for like cells in each map.

9. (amended) A sales-predicting method based on triple-axis mapping of customer value as recited in any of claims 6, 7, or 8, with response data for marketing approaches targeting each customer in each cell, with said response data including inquiries from customers, requests for information materials, store visits, purchases, purchase sums, ongoing purchases, and data regarding ongoing-purchase amounts, with said data stored in the marketing-approach-response database, and with the results of the marketing approaches for each cell analyzed based on data in the marketing-approach-response database, and subsequent marketing approaches conducted only for cells determined to be highly effective.