JP4860570B2 - Share prediction apparatus, share prediction method, and share prediction program - Google Patents

Description

  The present invention relates to a technique for predicting the share of the next model product from the actual share of the current model product. The current model product here refers to an existing product that has already been put on the market and sold, and the next model product is the same type of product that has undergone some model changes from the current model product. This is a product that is scheduled to be put on the market next to the current model product.

A product provider may introduce a new model product with a new specification, which is a part of the specification of the current model product, to the market. The product provider may desire to predict the demand amount, that is, the share for the next model product of the new specification before introducing the next model product of the new specification to the market. If we can predict the market share of the next model product with the new specifications, we will be able to establish a product supply system that does not exceed or lack demand.
However, it is difficult to predict the share of the next model product of the new model that has been remodeled in advance. Until it is put on the market, the next model product of the new specification is not known to the consumer, and the reaction of the consumer cannot be known. Therefore, a technology has been developed to investigate the value perceived by consumers for each level of product attributes, and to predict the share of the next model product using the evaluation points for the attributes and levels obtained as a result of the survey. .

In the technique of Patent Document 1, a method such as conjoint analysis is performed on the result of investigating the consumer's preference for the combination of levels of attributes constituting a product. As a result, the share of the existing model product and the existing model product that is in a competitive relationship with the current model product is calculated using the evaluation points that the consumer recognizes for each level of each attribute calculated.
At this time, in the technique of Patent Document 1, the evaluation score of a specific attribute is multiplied by a coefficient so that the share of each existing product calculated from the evaluation score for each level of each attribute matches the sales performance. The share of the next model product is predicted using the evaluation score for each level of each attribute after correction.
Japanese Patent Application Laid-Open No. 2004-206569

In general, in a survey in which a consumer answers a preference for a combination of levels of attributes constituting a product, the number of attributes and levels that can be surveyed is limited. For this reason, in the technique of Patent Document 1, a share attribute of the next model product is predicted by selecting a specific attribute from the attributes to be investigated and correcting the evaluation score of each level of the attribute with a predetermined coefficient. The accuracy of has been improved.
However, in fact, there is a possibility that items that have greatly influenced the value judgment of the consumer at the time of product purchase are not included in the attributes and levels that are the subject of the survey. Alternatively, there are cases where the consumer recognizes the value of the product for factors that are difficult to handle as the object of the survey, such as the product image and the dealer's response.
In the technology of Patent Document 1 in which a part of attributes targeted for the survey is corrected with coefficients, and the share of the next model product is predicted using the revised evaluation score, the elements that are not targeted for the survey as described above There is a possibility that the value recognized by the consumer cannot be accurately reflected in the share prediction of the next model product.

The value that the consumer recognizes for the product is, for example, the value that the consumer recognizes for each attribute level corresponding to the specifications of each product after investigating the value that the consumer recognizes for each level of each attribute. Can be obtained. Furthermore, a share (hereinafter referred to as a calculation share) of each product can be calculated using an evaluation score (hereinafter referred to as an overall evaluation score) added for each product. Based on the calculated share calculated from the above comprehensive evaluation points, the product provider can establish a product supply system that does not have excess or deficiency in demand, and can set up a sales strategy to improve the actual share It becomes. However, it is rare that the calculation share coincides well with the actual product share, and the low accuracy of the calculation share has been a problem.
The present inventors have conducted research on this problem, and the attributes and levels targeted by the questionnaire survey cannot broadly cover the elements that the consumer recognizes when actually purchasing the product. Focusing on the fact that there is a problem, by estimating the evaluation points for attributes that were not included in the survey (unsurveyed attributes) from the share and calculation share of existing products, It was found that the prediction accuracy can be improved.

  The present invention provides a technique for estimating the evaluation point of the unexamined attribute as described above from the actual share of the current model product and predicting the share of the next model product using the evaluation point of the unexamined attribute.

  The present invention can be embodied in an apparatus that predicts the share of the next model product from the actual share of the current model product. This device stores the evaluation score obtained as a result of investigating the value that the consumer recognizes for each level for a part of the attributes of the product for each level of the attribute (surveyed attribute) that is the subject of the survey. Stores the “researched attribute / level / evaluation point” storage means and the level of the surveyed attribute of the existing product group including the current model product and the competitive product that is in a competitive relationship with the current model product for each product. “Investigated attribute / level / existing product group” storage means, “actual share / existing product group” storage means storing the actual share of existing product group for each product, and unexamined for each product in the existing product group “Unresearched attribute evaluation points / existing product group” estimation means for estimating evaluation points for attributes, and evaluation levels for each level of the investigated attributes stored in the “Investigated attributes / levels / evaluation points” storage means; Stored in the “surveyed attribute / level / existing product group” storage means Survey based on the level of the surveyed attribute for each product of the existing product group and the evaluation score for the unsurveyed attribute for each product of the existing product group estimated by the “unstudied attribute score / existing product group” estimation means “Comprehensive evaluation point / existing product group” calculating means for calculating a comprehensive evaluation point for each product of the existing product group, and a comprehensive evaluation point that combines the evaluation point of the level of the completed attribute and the evaluation point for the unexamined attribute, and for each product of the product group “Calculation share / product / product group” calculation means for calculating the calculation share for each product in the product group based on the overall evaluation score, and “inspected attribute / “Level / next model product” storage means, “inspected attribute / level / next model product” storage means, and “inspected attribute / level / next model product” storage means Of the surveyed attribute of the next model product stored in Based on the evaluation points for the unexamined attributes of the current model product estimated by the “Unexamined attribute evaluation points / existing product group” estimation means, the evaluation score of the level of the investigated attribute of the next model product and the current model Calculated by “Comprehensive Evaluation Point / Next Model Product” calculation means that calculates the overall evaluation point of the next model product by combining the evaluation points for the unexamined attributes of the product, and “Comprehensive Evaluation Point / Existing Product Group” calculation means. The next model product in the new product group that includes the next model product and the competitor product based on the overall score of the competitor product and the overall score of the next model product calculated by the “Comprehensive score / next model product” calculation means The calculation share is calculated using the “calculation share / product / product group” calculation means, and a “prediction share / next model product / new product group” output means is provided to output as the forecast share of the next model product. It is. In the above apparatus, the “unstudied attribute evaluation point / existing product group” estimation means calculates the calculation share for each product in the existing product group calculated using the “calculation share / product / product group” calculation means, The unexamined attribute evaluation point for each product of the existing product group is estimated so as to approximate the actual share for each product in the existing product group stored in the “share / existing product group” storage means.

  The evaluation score that the consumer recognizes for the attribute and the level is, for example, for a questionnaire result that is implemented by a method such as presenting the consumer with a combination of the levels of each of the attributes and having their preference answered. It can be obtained by performing analysis processing such as conjoint analysis. When conducting a questionnaire for consumers, it is necessary to narrow down the attributes and levels to be surveyed in order to obtain valid answers. It is practically impossible to get answers for all attributes and levels of products. For this reason, the “surveyed attribute / level / evaluation score” storage means stores the attribute (surveyed attribute) that is the object of the survey and the evaluation score for that level.

The attribute for which the evaluation score for each level is stored in the “investigated attribute / level / evaluation score” storage means is an attribute representing the specification of the product, such as “displacement” and “air conditioner”. A plurality of levels are set for each of these attributes. For example, a level such as “2000 cc”, “2500 cc”, and “3000 cc” is set in the attribute “displacement”. The attribute level includes specifications that are not quantified. For example, in the attribute “air conditioner”, there are a level “Yes” in which an air conditioner is equipped and an attribute “none” in which no air conditioner is equipped.
The evaluation point represents the value that the consumer recognizes for each level of each attribute. For example, when the evaluation points for the respective levels “2000 cc”, “2500 cc”, and “3000 cc” of the attribute “displacement” are “5”, “20”, and “15”, respectively, Means that it is preferable (worth worth) in the order of “2500 cc”, “3000 cc”, and “2000 cc”. On the other hand, when the evaluation points for the respective levels “present” and “none” of the attribute “air conditioner” are “10” and “0”, respectively, this is more than when the consumer is not equipped with an air conditioner. Means that it is recognized that the equipment is preferable (worthy) as much as the above-mentioned evaluation points.

Each product in the existing product group including the current model product and the competitive product that is in a competitive relationship with the current model product has different specifications due to the different levels of the above-mentioned attributes. The “investigated attribute / level / existing merchandise group” storage means stores the level of the inspected attribute for each product in the existing merchandise group having different levels of some or all of the attributes.
Each product in the above existing product group is a product that has already been put on the market and sold. In the “actual share / existing product group” storage means, for example, the actual share for each product of the existing product group for each month is stored.

The “uninvestigated attribute evaluation point / existing product group” estimation means estimates an evaluation point for an uninvestigated attribute for each product in the existing product group. For example, if there is an existing product from product 1 to product n, the “uninvestigated attribute evaluation score / existing product group” estimating means evaluates the unstudied attribute evaluation score α _{1 for} each of n existing products. to estimate the ~α _n.

  The “comprehensive evaluation point / existing product group” calculation means calculates an overall evaluation point that combines the evaluation point of the level of the surveyed attribute and the evaluation point of the unexamined attribute for each product of the existing product group. For example, if the current model product has a level of “2000 cc” for the attribute “displacement” and “Yes” for the attribute “air conditioner”, the overall evaluation score of the current model product is the level of the above attribute. On the other hand, it is possible to calculate by adding the evaluation points for the unexamined attributes of the current model to the evaluation points “5” and “10” recognized by the consumer. Similarly, for the products in other existing product groups, the overall score can be calculated by adding the score corresponding to each level of the surveyed attribute and the score of the unsurveyed attribute.

The “calculation share / product / product group” calculation means can calculate a calculation share for each product in the existing product group based on the overall evaluation score for each product of the existing product group. For example, the calculation share of the current model product can be calculated using a total logit model method based on the overall evaluation score of each product.
At this time, if the evaluation score for the uninvestigated attribute of each product is increased or decreased, the value of the calculated share changes. That is, the value of the calculation share can be changed according to the evaluation point value of the uninvestigated attribute estimated by the “uninvestigated attribute evaluation point / existing product group” estimation means.
Therefore, in the above-mentioned “uninvestigated attribute evaluation point / existing product group” estimating means, the calculated share for each product in the existing product group is stored in the “actual share / existing product group” storage means. The unexamined attribute evaluation score for each product of the existing product group is estimated so as to be the closest to the actual share for each product in

  The “surveyed attribute / level / next model product” storage means stores the level of the surveyed attribute of the next model product. The next model product is a product that has not been sold yet. The next model product is composed of specifications and image concepts common to the current model product, although some specifications are different from the current model product. In addition, the next model product is expected to be put on the market through the same sales base and sales method as the current model product. From these facts, it is expected that the evaluation points for the unexamined attributes of the next model product are also common to the current model product. Based on this knowledge, the “comprehensive evaluation product / next model product” calculation means combines the evaluation score of the level of the surveyed attribute of the next model product with the evaluation score of the unexamined attribute of the current model product, Calculate the overall evaluation score of the model product. For example, if the next model product has a level of “3000 cc” for the attribute “displacement” and “Yes” for the attribute “air conditioner”, the overall evaluation score of the next model product is relative to the level of each attribute. Thus, the evaluation points “15” and “10” recognized by the consumer can be added to the evaluation points for the unexamined attributes of the current model product.

In addition, the next model product is expected to be in a competitive relationship with existing products that are common to the current model product. Hereinafter, the product group including the next model product and the competitor product having a competitive relationship with the next model product is referred to as a new product group. In addition, when continuing sales of the current model product even after the next model product is released, the current model product may be included in the new product group as a competitive product for the next model product.
The calculated share of each product in the new product group including the next model product is calculated using the “calculation share / product / product group” calculating means described above. The overall evaluation score for each product of the new product group including the next model product and the product in a competitive relationship with the next model product is calculated using the unexamined attribute evaluation points after correction. The “predicted share / next model product / new product group” output means outputs the calculated calculation share of the next model product as the predicted share of the next model product.
According to the present invention, the predicted share of the next model product can be accurately calculated.

The above-mentioned “uninvestigated attribute evaluation point / existing product group” estimating means assumes “evaluation points for unexamined attributes for each product of the existing product group” and “calculation share / product / product group” calculating means. Existing product group by correcting the evaluation score for the unexamined attribute of the existing product with the largest difference between the calculated share calculated by using and the actual share stored in the “actual share / existing product group” storage means It is preferable to estimate the uninvestigated attribute evaluation score for each product in the existing product group so that the calculated share for each product in the product is the closest to the actual share.
According to the present invention, corrections are made in order from the evaluation points for the uninvestigated attributes of existing products where the difference between the calculated share and the actual share is large. By using the evaluation score for the unexamined attribute after correction, the predicted share of the next model product can be accurately calculated.

  The present invention can be embodied in a method for predicting the share of the next model product from the actual share of the current model product. In this method, the “researched attribute / level / evaluation score” storage means stores the evaluation score obtained as a result of investigating the value that the consumer recognizes for each level for a part of the product attributes. The process of storing for each level of the target attribute (investigated attribute) and the existing model product and competing products that have a competitive relationship with the current model product in the “investigated attribute / level / existing product group” storage means Storing the surveyed attribute level of the product group for each product, storing the actual share of the existing product group for each product in the “actual share / existing product group” storage means, / "Existing product group" estimating means for estimating the evaluation score for the uninvestigated attributes for each product of the existing product group, and "total evaluation point / existing product group" calculating means for "examined attribute / level / evaluation" "Point" storage means for each level of the investigated attribute And the level of the surveyed attribute stored for each product of the existing product group in the “surveyed attribute / level / existing product group” storage means and the “unsearched attribute evaluation point / existing product group” Based on the evaluation points for the unexamined attributes estimated for each product in the existing product group using the estimation means, there is an existing comprehensive evaluation score that combines the evaluation points for the level of the investigated attributes and the evaluation points for the unexamined attributes Comprehensive evaluation calculated for each product in the existing product group using the “comprehensive evaluation point / existing product group” calculating means by the process of calculating for each product in the product group and the “calculation share / product / product group” calculating means. A step of calculating a calculation share for each product in the existing product group based on the points, a step of storing the level of the surveyed attribute of the next model product in the “surveyed attribute / level / next model product” storage means, "Comprehensive evaluation point / next model product" calculation Means for storing each level of the surveyed attribute in the “surveyed attribute / level / evaluation score” storage means, and the next model product stored in the “surveyed attribute / level / next model product” storage means. The level of the surveyed attribute of the next model product based on the level of the surveyed attribute of the current model product and the evaluation score estimated for the unsurveyed attribute of the current model product using the “unstudied attribute score / existing product group” estimation means The process of calculating the overall evaluation score of the next model product by combining the evaluation score of the current model product and the unexamined attribute of the current model product, and the “computation share / product / product group” calculation means, Comprehensive evaluation score calculated for competing products in the existing product group using the “existing product group” calculation means and comprehensive evaluation score calculated for the next model product using the “comprehensive evaluation point / next model product” calculation means Based on the above, the calculation share of the next model product in the new product group including the next model product and the competing product is calculated, and “calculated share / product” is output by the “predicted share / next model product / new product group” output means. / Commodity group "The calculation share of the next model product calculated using the calculating means is provided as a predicted share of the next model product. In the above-described method, in the step of estimating the evaluation score for the uninvestigated attribute for each product of the existing product group by the “uninvestigated attribute evaluation point / existing product group” estimating unit, the “calculation share / product / product group” calculating unit The existing product group so that the calculated share calculated for each product in the existing product group is the closest to the actual share stored for each product in the existing product group in the “actual share / existing product group” storage means. It is characterized by estimating an unexamined attribute evaluation score for each product.

  According to the method of the present invention, the calculated share of each product in the new product group including the next model product is calculated using the unexamined attribute evaluation points after correction. The forecast share of the next model product can be calculated accurately.

In the process of estimating the evaluation points for the uninvestigated attributes for each product of the existing product group by the above-mentioned “uninvestigated attribute evaluation points / existing product group” estimation means, the evaluation points for the unexamined attributes of the existing product group are assumed. However, the existing product with the largest difference between the calculated share calculated using the “calculated share / product / product group” calculating means and the actual share stored in the “actual share / existing product group” storage means is not yet determined. It is preferable to estimate the unexamined attribute evaluation score for each product in the existing product group so that the calculation share for each product in the existing product group is the closest to the actual share by correcting the evaluation score for the survey attribute .
According to the present invention, corrections are made in order from the evaluation points for the uninvestigated attributes of existing products where the difference between the calculated share and the actual share is large. By using the evaluation score for the unexamined attribute after correction, the predicted share of the next model product can be accurately calculated.

  The present invention further provides an evaluation score obtained as a result of investigating the value recognized by the consumer for each level for a part of the attributes of the product for each level of the attribute (surveyed attribute) to be investigated. Stores the “researched attribute / level / evaluation point” storage means stored for each product and the level of the surveyed attribute of the existing product group including the current model product and the competitive product that is in a competitive relationship with the current model product. The “investigated attribute / level / existing product group” storage means, the “actual share / existing product group” storage means storing the actual share of the existing product group for each product, and the inspected attribute of the next model product The computer having the “researched attribute / level / next model product” storage means for storing the level may be embodied in a program that functions as a device for predicting the share of the next model product from the actual share of the current model product. so That. This program uses the above-described computer to estimate an evaluation point for an uninvestigated attribute for each product of an existing product group, “uninvestigated attribute evaluation point / existing product group” estimation means, “inspected attribute / level / evaluation point” The evaluation score of each level of the surveyed attribute stored in the storage means, the level of the surveyed attribute for each product of the existing product group stored in the “searched attribute / level / existing product group” storage means, Based on the evaluation points for the unexamined attributes of each product of the existing product group estimated by the “survey attribute evaluation points / existing product group” estimation means, the evaluation points for the level of the surveyed attributes and the evaluation points for the unexamined attributes "Comprehensive evaluation point / existing product group" calculation means for calculating the total comprehensive evaluation score for each product in the existing product group, and calculating the calculation share for each product in the product group based on the overall evaluation point for each product in the product group "Calculation share / product / product group" Output means, evaluation points for each level of the surveyed attributes stored in the “examined attribute / level / evaluation score” storage means, and next model products stored in the “examined attribute / level / next model product” storage means Level of the surveyed attribute of the next model product based on the level of the surveyed attribute of the current model product and the evaluation score of the unexamined attribute of the current model product estimated by the “unstudied attribute score / existing product group” estimation means "Comprehensive evaluation point / next model product" calculation means, "Comprehensive evaluation point / existing product group" calculation that calculates the overall evaluation point of the next model product by combining the evaluation points and evaluation points for the unexamined attributes of the current model product The new product group that includes the next model product and the competitor product based on the overall evaluation score of the competitive product calculated by the means and the overall evaluation score of the next model product calculated by the "comprehensive evaluation point / next model product" calculation means Next term in Functions as a “prediction share / next model product / new product group” output means that calculates the calculated share of Dell products using the “calculation share / product / product group” calculation means and outputs it as the predicted share of the next model product Let Furthermore, in the program of the present invention, the “unsearched attribute evaluation point / existing product group” estimation means calculates the calculation share for each product in the existing product group calculated using the “calculation share / product / product group” calculation means. , Estimating the unexamined attribute evaluation score for each product of the existing product group so as to most closely match the actual share for each product in the existing product group stored in the “actual share / existing product group” storage means And

  Further, the “unstudied attribute evaluation score / existing product group” estimation means assumes the evaluation score for the unstudied attribute for each product of the existing product group, and calculates the “calculation share / product / product group”. By correcting the evaluation score for the unexamined attribute of the existing product with the largest difference between the calculated share calculated using the calculating means and the actual share stored in the “actual share / existing product group” storage means. It is preferable to estimate the uninvestigated attribute evaluation score for each product in the existing product group so that the calculated share for each product in the existing product group is the closest to the actual share.

  According to the present invention, even when the element that has greatly influenced the value judgment of the consumer at the time of actual product purchase is not included in the surveyed attribute, the actual share of the current model product is used as described above. The value that the consumer recognizes for the element can be quantified as an evaluation point for the unexamined attribute. The predicted share of the next model product can be calculated using the comprehensive evaluation value corrected with the evaluation score for the uninvestigated attribute. It is possible to accurately reflect the value that the consumer recognizes about the elements that were not subject to the survey in the share prediction of the next model product.

First, the main features of the embodiments described below are listed.
(Characteristic 1) The share prediction apparatus includes means for calculating a divergence index representing the degree of divergence between the calculated share and the actual share of the existing product group.
(Characteristic 2) The uninvestigated attribute evaluation score for each product in the existing product group is estimated so that the difference between the calculated share and the actual share is minimized.
(Characteristic 3) “Uninvestigated attribute evaluation point / existing product group” “Uninvestigated attribute assumed evaluation point / existing product group” that stores evaluation points for the uninvestigated attribute of each product of the existing product group estimated by the estimating means A storage means is provided.
(Characteristic 4) Of the unexamined attribute evaluation points of the existing product group stored in the “uninvestigated attribute assumed evaluation point / existing product group” storage means, the “calculation share / product / product group” calculation means is used for calculation. The calculated share for each product in the existing product group is the closest to the actual share for each product in the existing product group stored in the “actual share / existing product group” storage means. Identify unexamined attribute evaluation points.

Embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a diagram illustrating a schematic configuration of a share prediction apparatus 2 according to an embodiment of the present invention.
As shown in FIG. 1, the share prediction apparatus 2 includes a “researched attribute / level / evaluation point / demand” storage unit 16, a “researched attribute / level / existing product group” storage unit 18, and a “researched attribute”. Attribute / level / next model product ”storage means 20,“ actual share / existing product group ”storage means 22,“ uninvestigated attribute assumed evaluation point / existing product group ”storage means 23,“ uninvestigated attribute evaluation point / "Existing product group" storage means 24, "Unexamined attribute evaluation point / existing product group" estimation means 4, "Comprehensive evaluation point / existing product group" calculation means 6, and "Calculation share / product / product group" calculation means 8, "Comprehensive evaluation point / next model product" calculating means 10, "predicted share / next model product / new product group" output means 12, share prediction program 26, display means 14, and input means 28 I have.

  "Surveyed attribute / level / evaluation point / customer" storage means 16, "Surveyed attribute / level / existing product group" storage means 18, "Surveyed attribute / level / next model product" storage means 20, The “actual share / existing product group” storage means 22, the “uninvestigated attribute assumed evaluation point / existing product group” storage means 23, and the “uninvestigated attribute evaluation point / existing product group” storage means 24 include a hard disk or the like. This is a general storage means configured by (shown). Each of the various storage means described above may be configured as separate storage media, or may be configured by appropriately dividing the storage area of one storage medium.

  The “investigated attribute / level / evaluation score / customer” storage means 16 uses the evaluation score obtained as a result of investigating the value perceived by the consumer for each level for a part of the product attributes as a consumer. Remember every time. The evaluation score that the consumer recognizes for each attribute and each level is, for example, the result of a questionnaire conducted by a method such as presenting the consumer with a combination of multiple attributes and levels and answering their preferences. Thus, it can be obtained by performing analysis processing such as conjoint analysis. FIG. 4 illustrates information stored in the “surveyed attribute / level / evaluation point / demand” storage unit 16. The “investigated attribute / level / evaluation point / customer” storage means 16 is a part of attributes constituting the product, such as “displacement”, “fuel consumption”, “theft prevention device”, etc. The attributes (investigated attributes) and the evaluation points regarding the level are stored for each consumer. The evaluation point represents the relative value that the consumer recognizes for each level. For example, in the case of the consumer 1, there is a large difference between the level “16 km / l” and the level “8 km / l” of the attribute “fuel consumption”, whereas the attribute “displacement” is There is not much difference between the levels. This means that the fuel efficiency is emphasized in the evaluation of the product of the consumer 1 as the automobile. On the other hand, the customer 2 attaches great importance to the “displacement” in the evaluation of the automobile product, and does not attach much importance to other attributes such as “fuel consumption” and “theft prevention device”. Understood.

The “investigated attribute / level / existing product group” storage means 18 stores, for each product, the level of the inspected attribute of the existing product group including the current model product and competing products in a competitive relationship with the current model product. Yes. FIG. 5 illustrates information stored in the “surveyed attribute / level / existing product group” storage unit 18.
Each product in the existing product group is a product that has already been put on the market and sold. The “actual share / existing product group” storage means 22 stores the actual share of the existing product group for each product and each sales month. FIG. 7 illustrates information stored in the “actual share / existing product group” storage unit 22. Although FIG. 7 shows the case where the sales volume of each product is stored as the actual share, the calculated sales share of each product is calculated by dividing the sales volume of each product by the total sales volume of all products. It may be stored as a share.
The “researched attribute / level / next model product” storage means 20 stores the level of the surveyed attribute of the next model product. The next model product is a product of the same type in which some of the specifications have been changed from the current model product, and is a product that is scheduled to be put on the market next to the current model product. FIG. 6 illustrates information stored in the “surveyed attribute / level / next model product” storage unit 20. The next model product X 'is different from the current model product X' shown in Fig. 5 in terms of "displacement", "fuel consumption", and "anti-theft device", but the specifications and image concepts common to the current model product X It is constituted by.

The “uninvestigated attribute assumed evaluation point / existing product group” storage means 23 stores an evaluation point (hereinafter referred to as an uninvestigated attribute evaluation point) for an uninvestigated attribute estimated for each product in the existing product group. The unexamined attribute evaluation score represents the value that the consumer recognizes for each product in the existing product group with respect to an element that has not been subjected to the questionnaire survey as the surveyed attribute.
The “unstudied attribute evaluation point / existing product group” storage unit 24 is an unstudied attribute evaluation point estimated for each product of the existing product group, and stores the “calculation share / product / product group” calculation unit 8 described later. The existing product group when the calculated share for each product in the existing product group calculated using the closest approximation to the actual share for each product in the existing product group stored in the “actual share / existing product group” storage unit 22 The unexamined attribute evaluation score for each product is stored. The unexamined attribute evaluation point estimation process will be described with reference to a flowchart described later.

  The “uninvestigated attribute evaluation point / existing product group” estimation means 4 estimates an evaluation point for an uninvestigated attribute for each product in the existing product group. The estimated uninvestigated attribute evaluation score for each product of the existing product group is stored in the “uninvestigated attribute assumed evaluation score / existing product group” storage means 23. Furthermore, the unexamined attribute evaluation point / existing product group ”estimating means 4 calculates the calculation share for each product in the existing product group calculated using the“ calculation share / product / product group ”calculating means 8 described later. The unexamined attribute evaluation score for each product in the existing product group is estimated so as to approximate the actual share for each product in the existing product group stored in the “share / product” storage unit 22. Regarding the uninvestigated attribute evaluation score for each product in the existing product group when the calculated share for each product in the existing product group is the closest to the actual share for each product, the unexamined attribute evaluation score / existing product It is stored in the “group” storage means 24.

  The “total evaluation score / existing product group” calculation means 6 is provided with an evaluation score for each level of the surveyed attribute for each consumer stored in the “surveyed attribute / level / evaluation score / customer” storage means 16. The level of the surveyed attribute for each product of the existing product group stored in the “surveyed attribute / level / existing product group” storage unit 18 and the existing level estimated by the “unsearched attribute evaluation point / existing product group” estimation unit 4 Based on the evaluation points for the uninvestigated attributes for each product in the product group, calculate the overall evaluation score for the products in the existing product group, combining the evaluation points for the level of the investigated attributes and the evaluation points for the uninvestigated attributes. .

The “comprehensive evaluation point / next model product” calculation means 10 includes an evaluation point for each level of the investigated attribute for each consumer stored in the “investigated attribute / level / evaluation point / demand” storage means 16, and “ The level of the surveyed attribute of the next model product stored in the “surveyed attribute / level / next model product” storage means 20 and the current model product estimated by the “unsearched attribute evaluation point / existing product group” estimation means 4. Based on the evaluation score for the unexamined attribute, the evaluation score for the level of the investigated attribute of the next model product and the evaluation score for the unexamined attribute of the current model product are combined to calculate the overall evaluation score of the next model product.
As described above, the next model product is composed of specifications and image concepts common to the current model product, although the specifications of some attributes are different from those of the current model product. In addition, the next model product is expected to be put on the market through the same sales base and sales method as the current model product. From these facts, it is expected that the evaluation points for the unexamined attributes of the next model product are also common to the current model product. Based on such knowledge, the “comprehensive evaluation point / next model product” calculation means 10 combines the evaluation point of the level of the surveyed attribute of the next model product and the evaluation point of the unexamined attribute of the current model product, Calculate the overall evaluation score for the next model product.

  The “calculation share / product / product group” calculation means 8 calculates a calculation share for each product in the product group based on the comprehensive evaluation score for each product in the product group. The calculation share of the merchandise group can be calculated using a technique such as an aggregate logit model, for example. For example, the “calculation share / product / product group” calculation means 8 can calculate the calculation share for each product in the existing product group based on the comprehensive evaluation score for each product in the existing product group. The above-mentioned “uninvestigated attribute evaluation point / existing product group” estimation means 4 is configured such that the calculation share of each product of the existing product group calculated by the “calculation share / product / product group” calculation means 8 is “actual share / existing The uninvestigated attribute evaluation score for each product in the existing product group is estimated so as to approximate the actual share of each product in the existing product group stored in the “product group” storage unit 22.

Furthermore, since the next model product is the same type of product as the current model product, it is expected that the next model product is in a competitive relationship with the existing product in common with the current model product. From this, the “calculation share / product / product group” calculation means 8 calculates each product in the new product group based on the overall evaluation score for each product in the new product group including the next model product and the products in the existing product group. Calculation share can be calculated.
The “predictive share / next model product / new product group” output means 12 includes the overall score of the existing product group calculated by the “total evaluation point / existing product group” calculation means 6, and the “total evaluation score / next model product”. The next model in the new product group including the next model product and the competitor product calculated by using the “calculation share / product / product group” calculation unit 8 based on the overall evaluation score of the next model product calculated by the calculation unit 10 The calculated share of the product is output as the predicted share of the next model product. FIG. 10 illustrates the output prediction share.

“Unexamined attribute evaluation point / existing product group” estimation means 4, “total evaluation point / existing product group” calculation means 6, “calculation share / product / product group” calculation means 8, “comprehensive evaluation score” The “/ next model product” calculating means 10 and the “predictive share / next model product / new product group” output means 12 read the data stored in the various storage means and perform various processes using the read data. Execute. In other words, the various means described above are realized by a CPU (not shown) executing the share prediction program 26 stored in a hard disk or the like.
“Unexamined attribute evaluation point / existing product group” estimation means 4, “Comprehensive evaluation point / existing product group” calculation means 6, “Calculation share / product / product group” calculation means 8, “Comprehensive evaluation point / next period” The processing executed by the “model product” calculating means 10 and the “predicted share / next model product / new product group” output means 12 will be described in detail in the flowcharts described later.

Next, processing executed by the share prediction program 26 will be described with reference to the flowcharts of FIGS.
When an operator (product planner) executes a predetermined operation with the input unit 28, a CPU (not shown) reads and executes the share prediction program 26. Thereby, the CPU starts the processing shown in the flowchart of FIG. When the following processing is started, “investigated attribute / level / evaluation point / customer” storage means 16, “investigated attribute / level / existing product group” storage means 18, and “inspected attribute” It is assumed that the predetermined data described above is already stored in the attribute / level / next model product ”storage unit 20 and the“ actual share / existing product group ”storage unit 22.

In step S2, an evaluation score for an uninvestigated attribute for each product in the existing product group is set as an initial value. For example, when the existing product group is configured by N products including the current model product X, the uninvestigated attribute evaluation point α _n of the product n is set to the initial value in this step. In this embodiment, the initial values of the uninvestigated attribute evaluation points α ₁ ,..., Α _N of each product are all set to “0”.
Next, in step S4, the step size Δα of the unexamined attribute evaluation point is set. In this embodiment, a common step size Δα is used for all unexamined attribute evaluation points. The value of the step size Δα of the unexamined attribute evaluation point may be arbitrarily set by the operator. For example, a predetermined step size such as 0.5 or 1.0 is set in the share prediction program 26 by default. May be programmed. In this embodiment, the step size Δα is set to 1.0. The value of the step width Δα is held by a storage unit (not shown) while the CPU is processing.
In step S6, a deviation degree storage variable RMSE _min representing the degree of deviation between the calculated share and the actual share of each product in the existing product group is set to an initial value. FIG. 9 illustrates values taken by the deviation degree storage variable RMSE _min . As shown in FIG. 9, the divergence degree storage variable RMSE _min converges to the minimum value through repeated calculations, and therefore the initial value of the divergence degree storage variable RMSE _min is set to a reasonably large value. The initial value may be arbitrarily set by the operator, or may be set by default in the share prediction program 26. The value of the deviation degree storage variable RMSE _min is held by a storage unit (not shown) while the CPU is processing.

In step S8, the calculated overall score U _n of each product existing product lines. Overall rating point U _n of each product of existing products group, and the evaluation point of "investigation already attribute / level / evaluation point / consumers" every level and demand's been examined attributes that are stored in the storage means 16, "Survey "Computed attribute / level / existing product group" is calculated based on the surveyed attribute level for each product of the existing product group stored in the storage means 18 and the set unexamined attribute evaluation score for each product of the existing product group. The In a present Example, the comprehensive evaluation score of each product is calculated about each consumer, and the consumer evaluation of each product of the existing product group is calculated by calculating the consumer average of the comprehensive evaluation score of each product.
In this step, first, an overall evaluation score U _in of the product n of the consumer i is calculated from the following equation.

In the above equation, J is the total number of surveyed attributes, l (n, j) is the level in surveyed attribute j of product n, and β _{il (n, j)} is in surveyed attribute j of product n. It is the evaluation point of the consumer i for the level l, and α _n is the unexamined attribute evaluation point for the product n. By processing the above formula, a comprehensive evaluation score U _in for each consumer of each product in the existing product group including the current model product is calculated. For example, if the overall evaluation score of the consumer 1 for the current model product X in FIG. 5, the evaluation score “−2” of the attribute “displacement” level “2000cc” and the attribute “fuel consumption” shown in FIG. An evaluation point “3” for the level “12 km / l”, an evaluation point “2” for the attribute “theft prevention device”, an evaluation point “2”, and the current model product X Calculated by adding the survey attribute evaluation point α. Note that if the uninvestigated attribute evaluation points α _n remain at the initial values (that is, when the process of step S8 is executed for the first time), all the uninvestigated attribute evaluation points are 0. In this step, a comprehensive evaluation score for each consumer based on the evaluation score of the surveyed attribute is calculated.
Overall evaluation point U _in the above is calculated for each consumer, to calculate the overall score U _n of each product existing product lines from the following equation.

In the above formula, U _in is the overall evaluation score of the product n for each consumer, and I is the total number of consumers. By processing the above formula, overall score U _n of each product existing product lines including the current model item is calculated. For example, in the case of the current model product X, by processing the above formula, the overall evaluation score of the consumer 1, the overall evaluation score of the customer 2,. Is calculated.

Next, in step S10, a calculation share S ′ _n of the existing product group is calculated. Calculated share _S'n of each product of an existing product group is calculated based on the overall score U _n for each product in existing product lines. The calculation share S ′ _n of each product can be calculated from the following formula, for example.

In the above formula, N is the total number of products in the existing product group. Calculated share _S'n items n existing product lines, using the aggregate logit model described above, each product value after exponential conversion overall score U _n items n exp a (U _n) in existing product lines can be calculated by dividing the sum of the overall score U _n value exp after exponential conversion of (U _n) of the.

Next, in step S12, the degree of divergence between the calculated share S ′ _n and the actual share S _n of each product in the existing product group is calculated. The degree of divergence can be calculated using a technique such as RMSE (Root Mean Square Error). For example, the following formula is processed to calculate the divergence degree RMSE between the calculated share S ′ _n and the actual share S _n of each product in the existing product group.

In the above formula, S ′ _n is the calculated share of the product n, S _n is the actual share of the product n, and N is the total number of products in the existing product group.
In step S14, the RMSE calculated by processing the above equation is compared with the deviation degree storage variable RMSE _min set in step S6. If RMSE <RMSE _min (YES in step S14), the value of RMSE _min is updated to the minimum RMSE, that is, the value of RMSE calculated in step S12 and stored (step S16).
When the value of RMSE _min is updated, the unexamined attribute evaluation point α _n of the set existing product group is then stored in the “uninvestigated attribute assumed evaluation point / existing product group” storage means 23 (step S17). Immediately after starting the process from step S2 (that is, when the process of step S17 is executed for the first time), in this step, the initial value of the unexamined attribute evaluation point α _n of the existing product group is set to “assumed unexamined attribute”. Stored in the “evaluation point / existing product group” storage means 23.

Next, in step S18, from the existing product lines, the absolute value of the difference between the calculated share _S'k and actual share S _k identifies the maximum of the product k. The product k specified in this step means that the evaluation indicated by the calculation share is most deviated from the actual evaluation of the consumer indicated by the actual share. For this reason, it is necessary to correct the unexamined attribute evaluation point of the product k so that the calculated evaluation of the product k is close to the actual evaluation.
Therefore, in step S20, the value of the uninvestigated attribute evaluation point α _k of the product k specified in step S18 is updated. For example, when the value of the calculated share S ′ _k of the product k is larger than the value of the actual share S _k , this means that the product k is overestimated in the calculation evaluation. On the other hand, if the value of the calculated share S ′ _k is smaller than the value of the actual share S _k , this means that the product k is underestimated in the calculation evaluation. Therefore, in this step, the uninvestigated attribute evaluation point α _k of the product _k is corrected using the following equation depending on whether the value of the calculated share S ′ _k is larger than the value of the actual share S _k .

For example, if the unexamined attribute evaluation score α _{k of} the product k is the initial value (ie, “0”), and the value of the calculated share S ′ _k of the product k is larger than the value of the actual share S _k , The unexamined attribute evaluation point α _k of _k is reduced by Δα = 1.0. On the other hand, when the value of the calculated share S ′ _k of the product k is smaller than the value of the actual share S _k , the unexamined attribute evaluation point α _k of the product k is increased by Δα = 1.0. When the process of step S20 is executed for the first time, initial values are set for the unexamined attribute evaluation points of existing products other than the updated product k.

When the value of the uninvestigated attribute evaluation point α _k of the product k is updated, the processing from step S8 to step S14 is repeated using the unexamined attribute evaluation point of the updated existing product group. When a value smaller than the stored RMSE _min is calculated in step S12 (YES in step S14), the processing from step S16 to step S20 is repeatedly executed. That is, the unexamined attribute evaluation score of the existing product group updated by the first processing in step S20 is stored in the “uninvestigated attribute assumed evaluation score / existing product group” storage means 23, and the difference between the calculated share and the actual share is calculated. At that time, the updated value of the unexamined attribute evaluation score for the largest existing product is further corrected.
By repeating the above processing, the deviation degree storage variable RMSE _min is gradually updated to a smaller value. FIG. 9 shows a change in the deviation degree storage variable RMSE _min with respect to the number of processes. When the processes from step S8 to step S20 are repeatedly executed, the RMSE _min is minimized when a certain number of processes have been performed. Therefore, when the calculated RMSE value is equal to or greater than the stored RMSE _min value (NO in step S14), the process proceeds to step S21.

In step S 21, the unexamined attribute evaluation point α _n of the existing product group stored in the “uninvestigated attribute assumed evaluation point / existing product group” storage unit 23 is used as the “uninvestigated attribute evaluation point / existing product group” storage unit 24. To remember. In the following process, the process of calculating the share of the next model product is executed using the unexamined attribute evaluation point of the existing product group stored in the “uninvestigated attribute evaluation point / existing product group” storage unit 24.
In the step S22, to calculate the overall score _{U n} of the next model goods. Overall rating point U _n of the next model product, and the evaluation point of each consumer about the level of investigation already attributes stored in the "investigation already attribute / level / evaluation point / consumer" storage means 16, "investigation already attribute / Level / next model product "storage means 20 and the level of the surveyed attribute of the next model product and the unexamined current model product stored in the" unsearched attribute / evaluation point / existing product group "means 24 Calculated based on attribute evaluation points. For the overall evaluation score of the next model product, calculate the overall evaluation score of the next model product for each consumer and calculate the consumer average in the same way as calculating the overall evaluation score of each product in the existing product group described above. Can be obtained.

Next, in step S24, a calculation share S ′ _n of the new product group is calculated. As described above, the new product group in this embodiment refers to a product group including the next model product and the existing product group. The calculation share S ′ _n of each product in the new product group is the same as when calculating the calculation share of each product in the existing product group. with, calculated on the basis of the overall score U _n for each item in the new product group. Note that the overall evaluation score of the existing product group including the current model product has already been calculated by the process of step S8.
The calculated share of the calculated new product group is output to the display means 14 and displayed as the predicted share of the next model product (see FIG. 10).

  In the series of processes described above, the overall evaluation score for each product of the new product group including the next model product and the product in a competitive relationship with the next model product is calculated using the unexamined attribute evaluation point α after correction. . The predicted share of the next model product reflects the value that the consumer recognizes for the elements that are not included in the questionnaire survey, which are represented by the above-mentioned corrected unsurveyed attribute evaluation point α. By using the share prediction device 2, the prediction share of the next model product can be accurately calculated.

  The “actual share / existing product group” storage means 22 stores the actual share of the existing product group for each product and each sales month. For this reason, the “unstudied attribute evaluation score / existing product group” estimation unit 4 determines that the calculation share calculated by the “calculation share / product / product group” calculation unit 8 approximates the actual share of each month. It is possible to estimate an uninvestigated attribute evaluation point of each product in the existing product group. By using the unexamined attribute evaluation points estimated in this way, it is possible to obtain a predicted share of the next model product for each month throughout the year. Or you may estimate an unexamined attribute evaluation score so that a calculation share may approximate the average performance share of a predetermined number of months.

  In the above-described embodiment, the degree of divergence between the calculated share and the actual share of each product in the existing product group is calculated using RMSE, but the method for calculating the degree of divergence between the calculated share and the actual share It is not limited. For example, the calculation may be performed using a technique such as MAPE (Mean of Absolute Percentage Error).

Specific examples of the present invention have been described in detail above, but these are merely examples and do not limit the scope of the claims. The technology described in the claims includes various modifications and changes of the specific examples illustrated above.
The technical elements described in this specification or the drawings exhibit technical usefulness alone or in various combinations, and are not limited to the combinations described in the claims at the time of filing. In addition, the technology illustrated in the present specification or the drawings achieves a plurality of objects at the same time, and has technical utility by achieving one of the objects.

It is a figure showing the schematic structure of a share prediction apparatus. It is a flowchart showing the process which the program for share prediction performs. It is a flowchart showing the process which the program for share prediction performs. It is a figure showing the information which a "surveyed attribute / level / evaluation point / customer" memory | storage means memorize | stores. It is a figure showing the information which a "surveyed attribute / level / existing product group" memory | storage means memorize | stores. It is a figure showing the information which a "surveyed attribute / level / next model product" memory | storage means memorize | stores. It is a figure which shows the information which a "result share / existing merchandise group" memory | storage means memorize | stores. It is a figure showing the unexamined attribute evaluation point of the calculated existing goods group. It is a figure showing the change of a deviation degree memory variable. It is a figure showing the prediction share of the calculated next model goods.

Explanation of symbols

2: Share prediction device 4: “Uninvestigated attribute evaluation point / existing product group” estimation means 6: “Comprehensive evaluation point / existing product group” calculation means 8: “Calculation share / product / product group” calculation means 10: “General "Evaluation point / next model product" calculating means 12: "Predicted share / next model product / new product group" output means 14: Display means 16: "Researched attribute / level / evaluation point / customer" storage means 18: "Research" Completed attribute / level / existing product group "storage means 20:" examined attribute / level / next model product "storage means 22:" actual share / existing product group "storage means 23:" uninvestigated attribute assumed evaluation point / existing product " Group "storage means 24:" Uninvestigated attribute evaluation point / existing product group "storage means 26: Share prediction program 28: Input means

Claims (6)

A device that predicts the share of the next model product from the actual share of the current model product,
The evaluation score obtained as a result of investigating the value that the consumer recognizes for each level for a part of the product attribute is stored for each level of the attribute (inspected attribute) targeted for the survey. "Investigated attribute / level / evaluation score" storage means,
“Investigated attribute / level / existing product group” storage means for storing the level of the surveyed attribute of the existing product group including the current model product and the competitive product having a competitive relationship with the current model product for each product;
“Actual share / existing product group” storage means for storing the actual share of the existing product group for each product;
“Uninvestigated attribute evaluation point / existing product group” estimating means for estimating an evaluation point for an uninvestigated attribute for each product of the existing product group,
Each level of the surveyed attribute stored in the “surveyed attribute / level / evaluation score” storage means and each product of the existing product group stored in the “searched attribute / level / existing product group” storage means The level of the surveyed attribute is evaluated based on the level of the surveyed attribute and the evaluation score of the unsurveyed attribute for each product of the existing product group estimated by the “unsurveyed attribute evaluation point / existing product group” estimation means "Comprehensive evaluation point / existing product group" calculating means for calculating an overall evaluation point for each product of the existing product group, by combining the points and evaluation points for unexamined attributes,
“Calculation share / product / product group” calculating means for calculating a calculation share for each product in the product group based on the overall evaluation score for each product in the product group;
"Surveyed attribute / level / next model product" storage means for storing the level of the surveyed attribute of the next model product,
The evaluation score of each level of the surveyed attribute stored in the “surveyed attribute / level / evaluation score” storage means, and the survey of the next model product stored in the “surveyed attribute / level / next model product” storage means Based on the attribute level and the evaluation point of the unexamined attribute of the current model product estimated by the “unexamined attribute evaluation point / existing product group” estimation means, "Comprehensive evaluation point / next model product" calculation means for calculating the overall evaluation point of the next model product by combining the evaluation points for the unexamined attributes of the current model product,
Based on the overall evaluation score of the competitive product calculated by the “comprehensive evaluation score / existing product group” calculation means and the overall evaluation score of the next model product calculated by the “comprehensive evaluation score / next model product” calculation means Calculate the calculated share of the next model product in the new product group including the model product and the competitor product using the “calculation share / product / product group” calculating means, and output it as the predicted share of the next model product. Model product / new product group "output means,
The “unsurveyed attribute evaluation score / existing product group” estimation means calculates the calculation share of each product in the existing product group calculated using the “calculation share / product / product group” calculation means as “actual share / existing share”. A share prediction apparatus that estimates an uninvestigated attribute evaluation score for each product of the existing product group so as to approximate the actual share for each product in the existing product group stored in the “product group” storage unit.   The “uninvestigated attribute evaluation point / existing product group” estimating means assumes the evaluation score for the uninvestigated attribute for each product of the existing product group, and calculates the “calculation share / product / product group” calculating means. By correcting the evaluation score for the unexamined attribute of the existing product with the largest difference between the calculated share calculated using the actual share and the actual share stored in the “actual share / existing product group” storage means, The apparatus according to claim 1, wherein an uninvestigated attribute evaluation score for each product of the existing product group is estimated so that a calculated share for each product in the group is closest to the actual share. A method for predicting the share of the next model product from the actual share of the current model product,
“Surveyed attributes / levels / evaluation points” The attribute for which the evaluation points obtained as a result of investigating the value perceived by consumers for each level for a part of product attributes are stored in the storage means Storing each level of (inspected attribute);
Storing in the “investigated attribute / level / existing product group” storage means for each product the level of the inspected attribute of the existing product group including the current model product and the competing product that is in a competitive relationship with the current model product;
In the “actual share / existing product group” storage means, a process of storing the actual share of the existing product group for each product;
A step of estimating an evaluation score for an uninvestigated attribute for each product of an existing product group by an “uninvestigated attribute evaluation point / existing product group” estimating means;
The “total evaluation score / existing product group” calculating means stores the evaluation score stored for each level of the investigated attribute in the “examined attribute / level / evaluation score” storage means, and the “examined attribute / level / existing” The level of the surveyed attribute stored for each product in the existing product group in the “commodity group” storage means and the “unsurveyed attribute evaluation point / existing product group” estimation means are used to estimate each product in the existing product group. Calculating a total evaluation score for each product of the existing product group based on the evaluation score for the unexamined attribute,
Products in the existing product group based on the overall rating score calculated for each product in the existing product group by the “computation share / product / product group” calculating unit using the “total score / existing product group” calculating unit. The process of calculating the calculation share for each;
Storing the level of the surveyed attribute of the next model product in the “surveyed attribute / level / next model product” storage means;
The "total evaluation point / next model product" calculation means stores the evaluation points stored for each level of the investigated attribute in the "examined attribute / level / evaluation point" storage means, and the "examined attribute / level / next period" The level of the surveyed attribute of the next model product stored in the “model product” storage means and the evaluation score estimated for the unexamined attribute of the current model product using the “unsearched attribute evaluation point / existing product group” estimation means To calculate the overall evaluation score of the next model product by combining the evaluation score of the level of the surveyed attribute of the next model product and the evaluation score of the unexamined attribute of the current model product,
The “computation share / product / product group” calculation means uses the “total evaluation point / existing product group” calculation means to calculate the overall evaluation score for the competing products of the existing product group, and the “total evaluation score / A step of calculating a calculation share of the next model product in the new product group including the next model product and the competing product based on the overall evaluation score calculated for the next model product using the “next model product” calculating means;
The "predicted share / next model product / new product group" output means outputs the calculated share of the next model product calculated using the "calculated share / product / product group" calculating means as the predicted share of the next model product. It has a process,
In the step of estimating the evaluation score for the uninvestigated attribute for each product of the existing product group by the “uninvestigated attribute evaluation point / existing product group” estimating means, the “calculation share / product / product group” calculating means is used. Products in the existing product group so that the calculated share calculated for each product in the existing product group is closest to the actual share stored for each product in the existing product group in the “actual share / existing product group” storage means. A share prediction method characterized by estimating an unexamined attribute evaluation score for each.   In the step of estimating the evaluation point for the uninvestigated attribute of each product of the existing product group by the “uninvestigated attribute evaluation point / existing product group” estimation means, the evaluation point for the unexamined attribute of the existing product group is assumed. Of the existing product having the largest difference between the calculated share calculated using the “calculated share / product / product group” calculating means and the actual share stored in the “actual share / existing product group” storage means. By correcting the evaluation points for unexamined attributes, it is possible to estimate the unexamined attribute evaluation points for each product in the existing product group so that the calculated share for each product in the existing product group is closest to the actual share. 4. A method according to claim 3, characterized in that The evaluation score obtained as a result of investigating the value that the consumer recognizes for each level for a part of the product attribute is stored for each level of the attribute (inspected attribute) targeted for the survey. “Investigated attribute / level / evaluation point” storage means, and “examined attribute” storing the level of the investigated attribute of the existing product group including the current model product and the competitive product that is in a competitive relationship with the current model product for each product. / Level / existing product group "storage means," actual share / existing product group "storage means storing the actual share of the existing product group for each product, and the level of the surveyed attribute of the next model product A program that stores a computer having a “surveyed attribute / level / next model product” storage means as a device for predicting the share of the next model product from the actual share of the current model product, the computer comprising:
“Uninvestigated attribute evaluation point / existing product group” estimation means for estimating an evaluation point for an unexamined attribute for each product of the existing product group,
Each level of the surveyed attribute stored in the “surveyed attribute / level / evaluation score” storage means and each product of the existing product group stored in the “searched attribute / level / existing product group” storage means The level of the surveyed attribute is evaluated based on the level of the surveyed attribute and the evaluation score of the unsurveyed attribute for each product of the existing product group estimated by the “unsurveyed attribute evaluation point / existing product group” estimation means "Comprehensive evaluation point / existing product group" calculating means for calculating an overall evaluation point for each item of the existing product group, which combines the points and evaluation points for unexamined attributes,
“Calculation share / product / product group” calculating means for calculating a calculation share for each product in the product group based on the overall evaluation score for each product in the product group,
The evaluation score of each level of the surveyed attribute stored in the “surveyed attribute / level / evaluation score” storage means, and the survey of the next model product stored in the “surveyed attribute / level / next model product” storage means Based on the attribute level and the evaluation point of the unexamined attribute of the current model product estimated by the “unexamined attribute evaluation point / existing product group” estimation means, "Comprehensive evaluation point / next model product" calculation means for calculating the overall evaluation point of the next model product by integrating the evaluation points for the unexamined attributes of the current model product,
Based on the overall evaluation score of the competitive product calculated by the “comprehensive evaluation score / existing product group” calculation means and the overall evaluation score of the next model product calculated by the “comprehensive evaluation score / next model product” calculation means Calculate the calculated share of the next model product in the new product group including the model product and the competitor product using the “calculation share / product / product group” calculating means, and output it as the predicted share of the next model product. Function as "model product / new product group" output means,
The “unsearched attribute evaluation point / existing product group” estimation means calculates the calculation share for each product in the existing product group calculated using the “calculation share / product / product group” calculation means as “actual share / existing product”. A share prediction program characterized by estimating an uninvestigated attribute evaluation score for each product of an existing product group so as to approximate the actual share for each product in the existing product group stored in the “group” storage means .   The “uninvestigated attribute evaluation point / existing product group” estimating means assumes the evaluation score for the uninvestigated attribute for each product of the existing product group, and calculates the “calculation share / product / product group” calculating means. By correcting the evaluation score for the unexamined attribute of the existing product with the largest difference between the calculated share calculated using the actual share and the actual share stored in the “actual share / existing product group” storage means, 6. The share prediction program according to claim 5, wherein an uninvestigated attribute evaluation score for each product in the existing product group is estimated so that a calculated share for each product in the group is most approximate to the actual share.

Images