Advertising effectiveness estimation device

Abstract

To provide an advertisement placement effectiveness estimation device for estimating effectiveness of advertising placements with easy operations compared to conventional technologies.SOLUTION: In an advertisement placement effectiveness estimation device 100, a selection unit 130 selects a multiple regression analysis formula and / or a simple regression analysis formula for which a goodness of fit index for the multiple regression analysis formula and / or the simple regression analysis formula is equal to or greater than a prescribed goodness of fit. A path route generation unit 140 generates path routes between each item such that explanatory variables of the selected multiple regression analysis formula and / or simple regression analysis formula are items of an indirect effect source and an objective variable is an item of an indirect effect destination, and further that the explanatory variables of the multiple regression analysis formula and / or simple regression analysis formula selected by the selection unit are items of a direct effect source and the objective variable is an item of a direct effect destination.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an advertising effectiveness estimation device. [Background technology]

[0002] Patent Document 1 (Republished Publication No. 2005-64511) describes an embodiment in which one or more analytical engines, either a Bayesian network or MTS, are applied to individual consumer data stored in a first storage means and a second storage means that store media plans, the dependency of each advertising plan on response results is quantified, advertising plans that are estimated to be effective for the responses based on the quantified dependency are extracted, and the media plans stored in the first storage means are updated to match the advertising plans determined to be effective. In other words, Patent Document 1 describes that an effective media plan is determined by performing a regression analysis based on performance information.

[0003] Patent Document 2 (JP 2021-89485 A) describes an embodiment in which attribute information of users of a first user group and information regarding exposure to advertisements for a product are used as explanatory variables, and values ​​based on the actions taken by the users of the first user group regarding the product are used as objective variables to generate a trained model through machine learning. The trained model then inputs attribute information of users of a second user group and information regarding exposure to advertisements for the product, outputs an index value that evaluates the likelihood that the users of the second user group will take the action of the objective variable regarding the product, associates the index value with the users of the second user group, and notifies at least some of the users of the second user group of an advertisement for the product based on the index value. That is, Patent Document 2 describes generating a trained model for predicting user responses to advertisements through regression analysis, and using this model to determine an index value.

[0004] Patent Document 3 (Japanese Patent Publication No. 7186316) describes an estimation device that estimates the effectiveness of placing an advertisement for a target product in an information transmission medium, identifies the degree of contact with the information transmission medium, calculates the appeal of the information transmission medium to the contactee's awareness of products of the same type as the target product based on the results of a questionnaire from survey subjects including those who have contacted the information transmission medium, and when multiple awareness levels are set according to the stages leading up to a predetermined behavior for products of the same type as the target product, sets a weighting coefficient for regression analysis for the awareness of each level, calculates the product of the appeal of the awareness of each level and the weighting coefficient set for the awareness of each level, and estimates the effectiveness for each medium based on the product calculated for the degree of contact and the awareness of each level. That is, Patent Document 3 describes that the advertising effectiveness for each medium is estimated by performing regression analysis.

[0005] Patent Document 4 (JP 2002-24692 A) describes a mode in which, for advertisements that have already been published or distributed, information indicating individual advertising media indicating the media in which the advertisements have been specifically published or distributed, the content of the advertisements, advertising costs, and user responses to the advertisements are stored in a database, and an advertising plan including advertising periods and budgets for selected individual advertising media is created so as to maximize user responses based on manuscripts from advertisers and / or advertising agencies that are provided with information including advertising periods and budgets. In other words, Patent Document 4 describes that the advertising effectiveness of each medium is estimated based on performance information.

[0006] Patent Document 5 (JP 2005-190340 A) describes a mode in which multiple advertisements are placed on multiple interactive media, reaction parameters related to the target's reaction are measured for each of the multiple advertisements, cognitive parameters related to the recognition of each advertisement are measured for the target, the correlation between the reaction parameters expressed as a function that is a power regression equation and the cognitive parameters is analyzed, and the effectiveness of the advertisement to be placed is predicted based on the correlation obtained in advance. In other words, Patent Document 5 describes that the effectiveness of placing advertisements on interactive media is estimated by performing a regression analysis based on performance information. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] Republished Publication No. 2005-64511 [Patent Document 2] Patent Publication No. 2021-89485 [Patent Document 3] Patent No. 7186316 [Patent Document 4] Japanese Patent Application Laid-Open No. 2002-24692 [Patent Document 5] Japanese Patent Application Laid-Open No. 2005-190340 Summary of the Invention [Problem to be solved by the invention]

[0008] An object of the present invention is to make it possible to estimate the effectiveness of advertising placement with easier operations than conventional techniques. [Means for solving the problem]

[0009] The first aspect includes a performance value acquisition unit that acquires a performance value of an amount of advertising by advertising medium indicating the amount of advertising by advertising medium, a performance value of a response by advertising medium indicating the response by advertising medium, and a performance value of a total response obtained by adding up the performance values ​​of the responses by advertising medium; a regression analysis processing unit that performs analysis processing of a multiple regression analysis formula and / or a simple regression analysis formula that uses either the total response or the response by advertising medium as a target variable and one or more amounts of advertising by advertising medium as an explanatory variable based on the performance value acquired by the performance value acquisition unit; and a regression analysis processing unit that performs analysis processing of a multiple regression analysis formula and / or a simple regression analysis formula that uses an index of goodness of fit of the multiple regression analysis formula and / or the simple regression analysis formula as a predetermined goodness of fit or higher. a path route generation unit that generates path routes between each item so that the explanatory variables of the multiple regression analysis formula and / or the simple regression analysis formula selected by the selection unit become items of indirect effect sources and the objective variable becomes an item of indirect effect destination, and the explanatory variables of the multiple regression analysis formula and / or the simple regression analysis formula selected by the selection unit become items of direct effect sources and the objective variable becomes an item of direct effect destination, and a path coefficient calculation unit that calculates path coefficients between each item of the path routes based on the actual values ​​acquired by the actual value acquisition unit.

[0010] The second aspect is an advertising effectiveness estimation device in which, in the first aspect, the regression analysis processing unit analyzes and processes a multiple regression analysis formula and / or a simple regression analysis formula using a number of explanatory variables limited to a predetermined number that is less than the number of all advertising media.

[0011] The third aspect is an advertising effectiveness estimation device in which, in the first aspect, the total response and the response by advertising medium used in the regression analysis processing unit, the path route generation unit, and the path coefficient calculation unit are subjected to correction processing to remove seasonal influences.

[0012] The fourth aspect is an advertising placement effectiveness estimation device in which, in the first aspect, the advertising placement volume by advertising medium used in the regression analysis processing unit, the path route generation unit, and the path coefficient calculation unit has been subjected to a correction process to add residual effects.

[0013] A fifth aspect is a performance value acquisition unit that acquires a performance value of the amount of advertisements for each advertising medium, a performance value of the response for each advertising medium, and a total performance value of the response obtained by adding up the performance values ​​of the responses;

[0014] The advertising placement effectiveness estimation device includes: a path coefficient output unit that outputs path coefficients between items, each of which is the placement volume and total response for each advertising placement medium, based on the performance values ​​acquired by the performance value acquisition unit; an advertising volume ratio performance value output unit that outputs an advertising volume ratio performance value indicating the ratio of the placement volume for each advertising placement medium, based on the performance values ​​acquired by the performance value acquisition unit; an advertising volume ratio change receiving unit that receives a change to the advertising volume ratio relative to the advertising volume ratio performance value; and a total response increase / decrease output unit that outputs the amount of increase / decrease or the rate of increase / decrease in the total response, using the change in the placement volume ratio received by the advertising volume ratio change receiving unit and the path coefficient.

[0015] A sixth aspect is the advertising placement effectiveness estimation device according to the fifth aspect, wherein the path coefficient, the actual value of the advertising placement ratio, and the increase / decrease amount or the increase / decrease rate of the total response are displayed on the screen.

[0016] The seventh aspect is an advertising placement effectiveness estimation device in which, in the fifth aspect, the total response used in the path coefficient output unit and the total response increase / decrease output unit has undergone correction processing to remove seasonal influences.

[0017] The eighth aspect is an advertising placement effectiveness estimation device in which, in the fifth aspect, the placement volume of each advertising placement medium used in the placement volume ratio actual value output unit has been subjected to a correction process to add residual effects. [Effects of the Invention]

[0018] According to the first to eighth aspects, the effectiveness of advertising can be estimated with simpler operations than in the prior art. [Brief explanation of the drawings]

[0019] [Figure 1]FIG. 1 is a block diagram showing the functional configuration of an advertising placement effectiveness estimation device according to the first embodiment. [Figure 2] FIG. 2 is a block diagram showing the functional configuration of an advertising effectiveness estimation device according to the second embodiment. [Figure 3] FIG. 3 is a diagram showing the configuration of one or more personal computer terminals, a server, and a network that communicably connects the personal computer terminals and the server according to the embodiment. [Figure 4] FIG. 4 is a diagram showing a hardware configuration for realizing the functional configuration of the embodiment. [Figure 5] FIG. 5 is a flowchart showing the processing steps of a program for estimating the effectiveness of advertising, which realizes the functions of the advertising effectiveness estimation device of the first embodiment. [Figure 6] FIG. 6 is a flowchart showing the processing steps of a program for estimating the effectiveness of advertising, which realizes the functions of the advertising effectiveness estimation device according to the second embodiment. [Figure 7A] FIG. 7A is a diagram illustrating an example of a performance value acquisition screen displayed on the display screen of the display device of the terminal. [Figure 7B] FIG. 7B is a diagram illustrating an example of a performance value acquisition screen displayed on the display screen of the display device of the terminal. [Figure 8] FIG. 8 is a diagram illustrating an example of a path route screen displayed on the display screen of the display device of the terminal. [Figure 9] 9(A) and 9(B) are diagrams illustrating examples of path coefficient screens displayed on the display screen of the display device of the terminal. [Figure 10] 10(A) and 10(B) are diagrams illustrating examples of information contact path evaluation screens displayed on the display screen of the terminal display device. [Figure 11A] FIG. 11A is a diagram illustrating an example of an advertisement submission ratio change acceptance screen displayed on the display screen of the display device of the terminal, showing an initial state in which no change in advertisement submission ratio has been accepted. [Figure 11B] FIG. 11B is a diagram illustrating an example of an advertisement submission ratio change acceptance screen displayed on the display screen of the display device of the terminal, showing a state in which a change in advertisement submission ratio has been accepted. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0021] (Functional configuration of advertising effectiveness estimation device)

[0022] FIG. 1 is a block diagram showing the functional configuration of an advertising placement effectiveness estimation device 100 according to the first embodiment.

[0023] The advertising placement effectiveness estimation device 100 of the first embodiment includes a performance value acquisition unit 110, a regression analysis processing unit 120, a selection unit 130, a path route generation unit 140, and a path coefficient calculation unit 150.

[0024] FIG. 2 is a block diagram showing the functional configuration of an advertisement placement effectiveness estimation device 200 according to the second embodiment.

[0025] The advertising placement effectiveness estimation device 200 of the second embodiment is configured to include an actual value acquisition unit 110, a path coefficient output unit 160, an advertising volume ratio actual value output unit 170, an advertising volume ratio change reception unit 180, and a total response increase / decrease output unit 190.

[0026] 1 and 2 can be realized by a combination of one or more personal computer terminals 60, a server 70, and a network 80 that communicatively connects the personal computer terminals 60 and the server 70, as shown in Fig. 3. Alternatively, the functions can be realized by a single personal computer terminal 60.

[0027] (Hardware configuration)

[0028] Fig. 4 is a diagram showing a hardware configuration for realizing the functional configuration of the embodiment shown in Fig. 1 and Fig. 2. The hardware configuration of the personal computer terminal 60 or the server 70 is shown as an example.

[0029] As shown in FIG. 4, the personal computer terminal 60 or the server 70 includes a CPU (Central Processing Unit) 61, a ROM (Read Only Memory) 62, a RAM (Random Access Memory) 63, storage 64, an input device 66, a display device 67, a communication I / F 68, and an external storage device 69, which are connected to each other so as to be able to communicate with each other via a system bus 65.

[0030] The CPU 61 is a central processing unit that executes various programs and controls each device connected to a system bus 65. That is, the CPU 61 reads a program from the ROM 62 or the storage 64 and executes the program using the RAM 63 as a work area. The CPU 61 controls each device connected to the system bus 65 and performs various arithmetic processing in accordance with the program recorded in the ROM 62 or the storage 64. The ROM 62 or the storage 64 stores a BIOS (Basic Input / Output System) and an OS (Operating System), which are control programs executed by the CPU 61, as well as various computer-readable and executable programs and various necessary data for realizing this embodiment.

[0031] The ROM 62 stores various control programs and various data. The RAM 63 functions as the main memory, work area, etc. of the CPU 61 and temporarily stores programs or data as a working area. The storage 64 is configured with an HDD (Hard Disk Drive) or SSD (Solid State Drive) and stores various programs including the BIOS and OS, and various data.

[0032] The input device 66 includes a pointing device such as a mouse, a keyboard, a reading device such as a scanner, and is used to perform various inputs.

[0033] The display device 67 is, for example, a liquid crystal display, and displays various information. The display device 67 may also function as the input device 66 by employing a touch panel system.

[0034] The communication interface 68 is an interface for communicating with other devices such as the server 70 and the terminal 60, and uses standards such as Ethernet (registered trademark), FDDI, Wi-Fi (registered trademark), etc. The communication interface 68 connects to a network 80 and controls the transmission and reception of data.

[0035] The external storage device 69 is composed of various types of memory cards such as USB memory, HDD, SSD, or other external storage media that can be detachably connected.

[0036] (Program for estimating advertising effectiveness)

[0037] 5 is a flowchart showing the processing steps of a program PB1 for estimating advertising effectiveness that realizes the functions of the advertising effectiveness estimation device 100 of the first embodiment. The program PB1 for estimating advertising effectiveness causes a computer to execute a performance value acquisition process S11, a regression analysis process S12, a selection process S13, a path route generation process S14, and a path coefficient calculation process S15. The performance value acquisition process S11 corresponds to the process performed by the performance value acquisition unit 110. The regression analysis process S12 corresponds to the process performed by the regression analysis processing unit 120. The selection process S13 corresponds to the process performed by the selection unit 130. The path route generation process S14 corresponds to the process performed by the path route generation unit 140. The path coefficient calculation process S15 corresponds to the process performed by the path coefficient calculation unit 150.

[0038] 6 is a flowchart showing the processing steps of the program PB2 for estimating advertising effectiveness, which realizes the functions of the advertising effectiveness estimation device 200 of the second embodiment. The program PB2 for estimating advertising effectiveness causes a computer to execute a performance value acquisition process S11, a path coefficient output process S16, an advertising volume ratio performance value output process S17, an advertising volume ratio change acceptance process S18, and a total response increase / decrease output process S19. The performance value acquisition process S11 corresponds to the processing performed by the performance value acquisition unit 110. The path coefficient output process S16 corresponds to the processing performed by the path coefficient output unit 160. The advertising volume ratio performance value output process S17 corresponds to the processing performed by the advertising volume ratio performance value output unit 170. The advertising volume ratio change acceptance process S18 corresponds to the processing performed by the advertising volume ratio change acceptance unit 180. The total response increase / decrease output process S19 corresponds to the processing performed by the total response increase / decrease output unit 190.

[0039] The programs PB1 and PB2 are stored in the ROM 62, the storage 64, or the external storage device 69. The personal computer terminal 60 or the server 70 reads the programs PB1 and PB2 from the ROM 62, the storage 64, or the external storage device 69 and executes the above-mentioned processes.

[0040] Whether the above processes S11 to S19 are performed by the personal computer terminal 60 or the server 70 can be determined arbitrarily depending on the system configuration. For example, the programs PB1 and PB2 may be installed in the personal computer terminal 60, and the programs PB1 and PB2 may be executed by the personal computer terminal 60 alone. Alternatively, for example, the processes of the programs PB1 and PB2 may be executed by both the personal computer terminal 60 and the server 70 by accessing the server 70 from the personal computer terminal 60 via the network 80.

[0041] Each embodiment will be described below.

[0042] (First embodiment)

[0043] The advertising placement effectiveness estimation device 100 of the first embodiment includes a performance value acquisition unit 110, a regression analysis processing unit 120, a selection unit 130, a path route generation unit 140, and a path coefficient calculation unit 150, and executes a program PB1 having the processing procedure shown in Figure 5.

[0044] (Actual value acquisition process S11 in the actual value acquisition unit 110)

[0045] The actual value acquisition unit 110 acquires actual values ​​W1, W2, W3, W4, W5, W6, W7, W8, and W9 of the amount of advertising by advertising medium, which indicate the amount of advertising by advertising medium 1, 2, 3, 4, 5, 6, 7, 8, and 9; an actual value W0 of the total amount of advertising, which is the sum of the actual values ​​W1 to W9 of the amount of advertising by advertising medium; actual values ​​Z1, Z2, Z3, Z4, Z5, Z6, Z7, Z8, and Z9 of the response by advertising medium, which indicate the response by advertising medium; and an actual value Z0 of the total response, which is the sum of the actual values ​​Z1 to Z9 of the response by advertising medium (S11).

[0046] 7A and 7B respectively show examples of performance value acquisition screens 51A and 51B displayed on the display screen of the display device 67 of the personal computer terminal (hereinafter referred to as terminal) 10. Performance value acquisition screens 51A and 51B are configured as interface screens equipped with the functions of performance value acquisition unit 110. For example, by performing operations on performance value acquisition screen 51A shown in FIG. 7A to acquire performance values ​​W1, W2, W3, W4, W5, W6, W7, W8, and W9 of the advertising volume by advertising medium and performance value W0 of the total advertising volume, the acquired performance values ​​W1, W2, W3, W4, W5, W6, W7, W8, and W9 of the advertising volume by advertising medium and the numerical values ​​(indicated by "···" on the screen) of the total advertising volume W0 are displayed by year and month on performance value acquisition screen 51A. For example, the actual values ​​W1, W2, W3, W4, W5, W6, W7, W8, and W9 of monthly advertising volume by advertising medium for the past few years (3 years (36 months)) and the actual value W0 of the total advertising volume are read from a database (not shown) in the server 70 and displayed on the actual value acquisition screen 51A. Note that "..." in the cells in the table in Figure 7A indicates a specific numerical value. The same applies to the other figures.

[0047] Similarly, by performing an operation to acquire the response performance values ​​Z1, Z2, Z3, Z4, Z5, Z6, Z7, Z8, and Z9 by advertising medium and the total response performance value Z0 on the performance value acquisition screen 51B shown in Fig. 7B, the acquired response performance values ​​Z1, Z2, Z3, Z4, Z5, Z6, Z7, Z8, and Z9 by advertising medium and the total response performance value Z0 (indicated by "..." on the screen) are displayed by year and month on the performance value acquisition screen 51B. For example, the response performance values ​​Z1, Z2, Z3, Z4, Z5, Z6, Z7, Z8, and Z9 by advertising medium for each month over the past several years (3 years (36 months)) and the total response performance value Z0 are read from a database (not shown) in the server 70 and displayed on the performance value acquisition screen 51B.

[0048] Here, the variables for the advertising volume of each advertising medium 1, 2, 3, 4, 5, 6, 7, 8, and 9 are defined as X1, X2, X3, X4, X5, X6, X7, X8, and X9, respectively. The response variables for the advertised products and services are defined as Y1, Y2, Y3, Y4, Y5, Y6, Y7, Y8, and Y9. Advertising volume is used to mean the number of advertisements and the amount of advertisements. Response is also used to mean the amount of orders, the number of orders, the amount of orders, the number of site referrals, and the number of searches. Responses may also be set to the number of clicks, the number of registrations, and other user actions that lead to profits from the advertising target, known as conversions. The unit of order amount is, for example, Japanese yen.

[0049] Each of advertising media 1, 2, 3, 4, 5, 7, 8, and 9 is, for example, television, newspaper, radio, magazine, flyer, enclosed flyer, listing advertisement, display advertisement, or other measure. These are only examples, and advertising media can be arbitrarily selected and combined as long as they are media capable of placing advertisements. For example, if the advertising media is newspaper advertisements, advertisements from different newspapers or different regions may be selected and combined, and different publication formats, such as publication size, monochrome advertisements, and color advertisements, may be selected and combined. Furthermore, if the advertising media is television broadcast advertisements, advertisements from different broadcast formats, such as terrestrial television broadcasts, satellite broadcasts, and cable broadcasts, may be selected and combined, or specific or arbitrary television commercials or television programs may be selected and combined. Furthermore, if the advertising media is radio broadcast advertisements, for example, arbitrary radio commercials or radio programs may be selected and combined. Furthermore, if the advertising media is internet advertisements, for example, specific or arbitrary digital advertisements, websites, social networking sites, and application software installed on information terminals such as smartphones and personal computers may be selected and combined. Furthermore, for example, if the advertising medium is outdoor advertising, it may be possible to select and combine billboards, digital signage, and in-vehicle advertising in various types of mobile media such as trains, buses, automobiles, etc. Any one of newspaper advertising, television advertising, radio advertising, internet advertising, and outdoor advertising may be selected, or two or more may be selected and combined.

[0050] The actual advertising volume values ​​W1, W2, W3, W4, W5, W6, W7, W8, and W9 for each advertising medium are obtained by substituting past advertising volumes into the advertising volume variables X1, X2, X3, X4, X5, X6, X7, X8, and X9 for advertising medium 1, 2, 3, 4, 5, 6, 7, 8, and 9, respectively. The actual response values ​​Z1, Z2, Z3, Z4, Z5, Z6, Z7, Z8, and Z9 for each advertising medium are obtained by substituting past response counts and volumes into response variables Y1, Y2, Y3, Y4, Y5, Y6, Y7, Y8, and Y9 for advertising medium 1, 2, 3, 4, 5, 6, 7, 8, and 9, respectively. The actual response total value Z0 is obtained by substituting the total response count and volume into the total response variable Y0.

[0051] (Regression analysis processing S12 in regression analysis processing unit 120)

[0052] Based on the actual values ​​acquired by the actual value acquisition unit 110, the regression analysis processing unit 120 performs analytical processing using a multiple regression analysis formula and / or a simple regression analysis formula with either the total response or the response by advertising medium as the objective variable and one or more advertising volumes by advertising medium as the explanatory variables (S12).

[0053] For example, the multiple regression analysis formula and the simple regression analysis formula can be analyzed under the following conditions.

[0054] Condition 1) The explanatory variables are the volume of advertising by advertising medium corresponding to up to three of advertising media 1 to 9, and multiple regression analysis equations and simple regression analysis equations are calculated for all combinations where the total response or the response by advertising medium corresponding to any advertising medium is used as the objective variable.

[0055] Although the number of explanatory variables used in the multiple regression analysis formula is limited to three or less, the number of explanatory variables used in the multiple regression analysis formula may be limited to any predetermined number that is less than the total number of advertising media (9).

[0056] The multiple regression analysis formula is generally expressed by the following formula (1) with β as the objective variable and α1, α2, and α3 as explanatory variables.

[0057] β=b0+b1·α1+b2·α2+b3·α3···(1)

[0058] A simple regression analysis formula is generally expressed by the following formula (2) with β as the dependent variable and α1 as the explanatory variable.

[0059] β=b0+b1·α1·(2)

[0060] The regression coefficients b0, b1, b2, and b3 are calculated by applying the least squares method to the data for the actual advertising volume values ​​by advertising medium W1, W2, W3, W4, W5, W6, W7, W8, and W9, the actual response values ​​by advertising medium Z1, Z2, Z3, Z4, Z5, Z6, Z7, Z8, and Z9, and the total actual response value Z0. For example, the regression coefficients b0, b1, b2, and b3 that minimize the mean squared error between the actually observed value of the objective variable and the estimated value (theoretical value) calculated by applying the multiple regression analysis formula are calculated.

[0061] For example, as shown in equation (3) below, a multiple regression analysis equation can be obtained in which the amount of advertising by advertising medium X4 corresponding to advertising medium 4 (magazines) and the amount of advertising by advertising medium X8 corresponding to advertising medium 8 (display advertising) are used as explanatory variables, and the response by advertising medium Y1 corresponding to advertising medium 1 (television) is used as the objective variable.

[0062] Y1=b0+b1·X4+b2·Y8 (3)

[0063] Furthermore, for example, as shown in the following equation (4), a multiple regression analysis equation can be obtained in which the amount of advertising by advertising medium X1 corresponding to advertising medium 1 (television), the amount of advertising by advertising medium X2 corresponding to advertising medium 2 (newspaper), and the amount of advertising by advertising medium X3 corresponding to advertising medium 3 (radio) are used as explanatory variables, and the total response Y0 is used as the objective variable.

[0064] Y0=b0+b1·X1+b2·X2+b3·X3 (4)

[0065] Also, for example, as in the following equation (5), a simple regression analysis equation can be obtained in which the amount of advertising by advertising medium X3 corresponding to advertising medium 3 (radio) is used as an explanatory variable and the total response Y0 is used as a target variable.

[0066] Y0=b0+b1·X3 (5)

[0067] (Selection process S13 in the selection unit 130)

[0068] The selection unit 130 selects the multiple regression analysis formula and / or the simple regression analysis formula whose fitness index is equal to or greater than a predetermined fitness index (S13).

[0069] Here, the index of goodness of fit is an index that evaluates the degree of agreement between the predicted value of the dependent variable obtained by regression analysis and the actual value of the dependent variable, and for example, the coefficient of determination R2 can be used.

[0070] For example, among the multiple regression analysis formulas and the simple regression analysis formulas obtained in the regression analysis process S12,

[0071] Coefficient of determination R2 ≥ 0.45 (6)

[0072] The multiple regression analysis formula and simple regression analysis formula are extracted, and advertising media corresponding to the explanatory variables in the extracted multiple regression analysis formula and simple regression analysis formula are selected. An R2 value of 0.45 or greater indicates that the explanatory variables in the multiple regression analysis formula and simple regression analysis formula can explain 45% or more of the target variable. The R2 coefficient is calculated by dividing the regression variation by the total variation. In addition to the R2 coefficient, the adjusted coefficient of determination (corrected coefficient of determination) R2f or the multiple correlation coefficient R can also be used as an indicator of goodness of fit.

[0073] For example, if the multiple regression analysis formulas shown in equations (3) and (4) and the simple regression analysis formula shown in equation (5) satisfy the conditions shown in equation (6), then these multiple regression analysis formulas shown in equations (3), (4) and (5) will be selected for generating the path route.

[0074] (Path route generation process S14 in the path route generation unit 140)

[0075] The path route generation unit 140 generates path routes between each item so that the explanatory variables of the multiple regression analysis formula and / or simple regression analysis formula selected by the selection unit 130 become items that are the source of indirect effects and the objective variable becomes an item that is the destination of indirect effects, and the explanatory variables of the multiple regression analysis formula and / or simple regression analysis formula selected by the selection unit 130 become items that are the source of direct effects and the objective variable becomes an item that is the destination of direct effects (S14).

[0076] FIG. 8 shows an example of a path route screen 52 displayed on the display screen of the display device 67 of the terminal 10. As shown in FIG.

[0077] The path route screen 52 displays the path route PA generated by the path route generation processing S14.

[0078] In FIG. 8, paths in the path route PA are indicated by arrows "→" and items (variables) are indicated by "□".

[0079] The multiple regression analysis formulas shown in equations (3) and (4) and the simple regression analysis formula shown in equation (5) will be used as examples.

[0080] The advertising volume by advertising medium corresponding to the explanatory variables in equation (3) (Y1 = b0 + b1 · X4 + b2 · Y8) is X4 (magazines) and X8 (display advertising), and the response by advertising medium corresponding to the objective variable is Y1 (television). Therefore, an indirect path P1 is generated with advertising volume by advertising medium X4 (magazines) and X8 (display advertising) as the indirect effect source items, and advertising volume by advertising medium X1 (television) as the indirect effect destination item. As shown in Figure 8, an indirect path P1 is generated with the starting point of the path arrow "→" being the indirect effect source items X4 (magazines) and X8 (display advertising), and the ending point of the path arrow "→" being the indirect effect destination item X1 (television).

[0081] Similarly, the advertising volumes by advertising medium corresponding to the explanatory variables in equation (4) (Y0 = b0 + b1 · X1 + b2 · X2 + b3 · X3) are X1 (television), X2 (newspapers), and X3 (radio), and the total response Y0 corresponds to the objective variable. Therefore, a direct path P2 is generated, with the advertising volumes by advertising medium X1 (television), X2 (newspapers), and X3 (radio) as the direct effect source items and the total response Y0 as the direct effect destination item. As shown in Figure 8, a direct path P2 is generated, with the starting point of the path arrow "→" being the direct effect source items X1 (television), X2 (newspapers), and X3 (radio), and the end point of the path arrow "→" being the direct effect destination item Y0 (total response).

[0082] Similarly, the advertising volume by advertising medium corresponding to the explanatory variable in equation (5) (Y0 = b0 + b1 · X3) is X3 (radio), and the total response Y0 corresponds to the objective variable. Therefore, a direct path P3 is generated in which the advertising volume by advertising medium X3 (radio) is the direct effect source item and the total response Y0 is the direct effect destination item. As shown in Figure 8, a direct path P3 is generated in which the starting point of the path arrow "→" is the direct effect source item X3 (radio) and the ending point of the path arrow "→" is the direct effect destination item Y0 (total response).

[0083] Similarly, paths are formed according to the multiple regression analysis formula and / or the simple regression analysis formula selected by the selection unit 130 .

[0084] Furthermore, in this embodiment, a path is generated on the assumption that the path route includes an item for unknown factor U. Note that it is also possible to omit the item for unknown factor U.

[0085] (Path coefficient calculation process S15 in the path coefficient calculation unit 150)

[0086] The path coefficient calculation unit 150 calculates the path coefficient for the path between each item of the path route PA based on the performance values ​​acquired by the performance value acquisition unit 110 (S15).

[0087] For example, the path coefficients are calculated by structural analysis of covariance (SEM). The path coefficients may be standard partial regression coefficients or partial regression coefficients from multiple regression analysis.

[0088] 9(A) and (B) show examples of the pass coefficient screen 53 displayed on the display screen of the display device 67 of the terminal 10.

[0089] The path coefficient screen 53 includes a path coefficient display section 53A and a path coefficient detail display section 53B.

[0090] As shown in Figure 9(A), the path coefficient display section 53A displays, for each advertising medium 1 (television), 2 (newspaper), 3 (radio), 4 (magazine), 5 (flyer), 6 (enclosed), 7 (listing advertisement), 8 (display advertisement), 9 (other measures), and for each unknown factor U, the numerical value of the path coefficient of the direct path (direct effect), the numerical value of the path coefficient of the indirect path (indirect effect), and the numerical value of the path coefficient (total effect) obtained by adding the path coefficient of the direct path and the path coefficient of the indirect path.

[0091] As shown in Fig. 9(B), the path coefficient detail display section 53B displays the numerical values ​​of the path coefficients for the paths between the items of the path route PA. The column items in Fig. 9(B) correspond to the items at the start points of the path arrows "→", and the row items in Fig. 9(B) correspond to the items at the end points of the path arrows "→".

[0092] For example, the path coefficient (direct effect) of direct path P4 of advertising medium 4 (magazine) is EN1 (-0.077). The path coefficient (indirect effect) of indirect path P5 of advertising medium 4 (magazine) is EN2 (0.320) for indirect path P41 from the item "Advertising volume by advertising medium X4 (magazine)" to the item "Advertising volume by advertising medium X1 (TV)," and the path coefficient of direct path P10 from the item "Advertising volume by advertising medium X1 (TV)" to the item "Total response Y0" is EN3 (0.148). Therefore, the path coefficients EN2 (0.320) and EN3 (0.148) are multiplied together to obtain EN4 (0.047). The direct and indirect effects of each of the other advertising media are calculated in the same way and displayed in path coefficient display section 53A.

[0093] 10(A) and (B) show examples of the information contact path evaluation screen 54 displayed on the display screen of the display device 67 of the terminal 10. FIG.

[0094] The information contact path evaluation screen 54 includes an influence degree upper path display section 54A and an influence degree lower path display section 54B.

[0095] As shown in Figure 10(A), the top influence path display section 54A displays the top information contact paths (e.g., the top 10 paths) with the highest influence (direct effect or indirect effect) among the information contact paths leading to the items with the total response Y0.

[0096] As shown in Figure 10(B), the influence lower path display section 54B displays the top information contact paths (e.g., the bottom 10 paths) with the lowest influence (direct effect or indirect effect) among the information contact paths leading to the items with the total response Y0.

[0097] For example, as shown in Figure 10(A), the influence (indirect effect) of the information contact path P5 from the item of advertising volume by advertising medium X4 (magazines) through the item of advertising volume by advertising medium X1 (television) to the item of total response Y0 is EN4 (0.047), and the influence is evaluated to be in the top five.

[0098] Also, as shown in Figure 10(B), the influence (direct effect) of the information contact path P4 from the item of advertising volume by advertising medium X4 (magazines) to the item of total response Y0 is EN1 (-0.077), and the influence is evaluated to be third lowest.

[0099] Generally, when performing path analysis, it is necessary to make a hypothesis about the causal relationship between each item (variable) of the path route. However, according to the first embodiment, it is possible to estimate the direct effect, indirect effect, and total effect of each advertising medium simply by performing an operation to acquire the past advertising volume and actual response values ​​of each advertising medium.

[0100] Since the total response Y0 and the responses by advertising medium Y1 to Y9 are seasonal, seasonal adjustments may be made.

[0101] That is, the total response Y0 and the responses Y1 to Y9 by advertising medium used in the regression analysis processing unit 120, the path route generating unit 140, and the path coefficient calculation unit 15 may be corrected to remove seasonal influences.

[0102] For example, the seasonal index Si can be calculated using an analytical method based on the EPA (Economic Planning Agency) method, and the seasonal influence can be removed by multiplying this seasonal index Si by the total response Y0 and the responses Y1 to Y9 by advertising medium.

[0103] The total response Y0 and the responses Y1 to Y9 by advertising medium may be influenced more by the residual effect of the advertising medium-specific advertising volumes X1 to X9 than by the advertising medium-specific advertising volumes X1 to X9 themselves. Therefore, a correction process may be performed to add the residual effect to the advertising medium-specific advertising volumes X1 to X9 used by the regression analysis processing unit 120, the path route generation unit 140, and the path coefficient calculation unit 15.

[0104] The residual effect Amt for period t of advertising media m (= 1, 2, 3, 4, 5, 6, 7, 8, 9) is expressed by the following equation (7).

[0105] Amt= λ·Xmt +(1-λ)·Am(t-1) ···(7)

[0106] In the above equation (7), t is

[0107] t=1, 2,..., n...(8)

[0108] and indicates the number of periods. Xmt indicates the amount of advertising placed by advertising medium m in period t. λ indicates the survival rate (forgetting rate, decay rate) and takes the following range:

[0109] 0≦λ≦1 (9)

[0110] The residual rate λ can be calculated as a value that maximizes the time-series correlation between the total response Y0, the responses Y1 to Y9 by advertising medium, and the residual effect amount Amt, for example.

[0111] When performing multiple regression analysis taking residual effects into account, for example, the above equation (3) (Y1=b0+b1·X4+b2·Y8) can be rewritten as the following equation (3)':

[0112] Y1t=b0+b1·A4t +b2·A8t ···(3)´

[0113] In equation (3), Y1t indicates the response by advertising medium 1 (television) in period t. The same applies to simple regression analysis.

[0114] (Second embodiment)

[0115] The advertising placement effectiveness estimation device 200 of the second embodiment includes an actual value acquisition unit 110, a path coefficient output unit 160, an advertising volume ratio actual value output unit 170, an advertising volume ratio change reception unit 180, and a total response increase / decrease output unit 190, and executes a program PB2 having the processing procedure shown in Figure 6.

[0116] In the following description, the description of processes that overlap with the processes in the first embodiment will be omitted as appropriate.

[0117] (Actual value acquisition process S11 in the actual value acquisition unit 110)

[0118] The same process as the performance value acquisition process S11 in the first embodiment is executed.

[0119] That is, as shown in Figures 7A and 7B, when an operation is performed on the display screen of the display device 67 of the terminal 10 to obtain the past advertising volume and actual response values ​​of each advertising medium, the past advertising volume and actual response values ​​of each advertising medium are displayed.

[0120] (Path coefficient output process S16 in the path coefficient output unit 160)

[0121] The path coefficient output unit 160 outputs the path coefficient between each item, where the items are the amount of advertising and the total number of responses for each advertising medium, based on the performance values ​​acquired by the performance value acquisition unit 110 (S16).

[0122] For example, the same processes as the regression analysis process S12, selection process S13, path route generation process S14, and path coefficient calculation process S15 of the first embodiment are executed to calculate the path coefficients between each item, where the items are the amount of advertising for each advertising medium and the total number of responses. The calculated path coefficients are displayed on the display screen of the display device 67 of the terminal 10.

[0123] For example, as shown in FIGS. 9(A) and 9(B), the path coefficients between each item, where each item is the amount of advertising for each advertising medium and the total number of responses, are displayed on the display screen of the display device 67 of the terminal 10.

[0124] (Advertisement volume ratio result value output process S17 in the advertisement volume ratio result value output unit 170)

[0125] The advertisement volume ratio result value output unit 170 outputs an advertisement volume ratio result value indicating the ratio of advertisement volumes of each advertisement medium based on the result values ​​acquired by the result value acquisition unit 110 (S17).

[0126] FIG. 11A illustrates an example of an advertising volume ratio change acceptance screen 55 displayed on the display screen of the display device 67 of the terminal 10. As shown in FIG.

[0127] The advertising volume ratio change reception screen 55 includes an advertising volume ratio performance value display section 55A, an advertising volume ratio change reception section 55B, and a total response increase / decrease display section 55C.

[0128] The advertisement volume ratio performance value display unit 55A functions as an advertisement volume ratio performance value output unit 170.

[0129] The advertising volume ratio performance value display section 55A displays the current advertising volume (performance value) ratios MR1, MR2, MR3, MR4, MR5, MR6, MR7, MR8, and MR9 for each advertising medium 1 (television), 2 (newspapers), 3 (radio), 4 (magazines), 5 (flyers), 6 (enclosed mail), 7 (listing ads), 8 (display ads), and 9 (other measures). The current advertising volume (performance value) ratios MR1 ​​to MR9 are calculated as the ratio to the total advertising volume M1+M2+···+M9, for example, by calculating the average advertising volume M1 to M9 over the past three years for each advertising medium 1 to 9, as shown in the following formula (7).

[0130] MRi=100%×Mi / (M1+M2+···+M9)···(10)

[0131] Here, i corresponds to each advertising medium 1, 2, 3, 4, 5, 6, 7, 8, and 9.

[0132] The current advertising volume (actual value) ratios MR1 ​​to MR9 can be used as a reference when changing the advertising volume ratios.

[0133] (Advertisement volume ratio change acceptance process S18 in the advertisement volume ratio change acceptance unit 180)

[0134] The advertising volume ratio change receiving unit 180 receives a change in the advertising volume ratio relative to the advertising volume ratio performance value (S18).

[0135] The advertisement volume ratio change receiving unit 55B functions as an advertisement volume ratio change receiving unit 180.

[0136] The advertising volume ratio change receiving unit 55B receives input operations to change the advertising volume ratios NR1, NR2, NR3, NR4, NR5, NR6, NR7, NR8, and NR9 for each advertising medium (1 (television), 2 (newspaper), 3 (radio), 4 (magazine), 5 (flyer), 6 (enclosure), 7 (listing advertisement), 8 (display advertisement), and 9 (other measures). The advertising volume ratio change receiving unit 55B may be configured with text boxes that allow the input of numerical values ​​in units of %, for example. The operator can input realistic change ratios (%) into the advertising volume ratio change receiving unit 55B as text boxes, using the current advertising volume (actual value) ratios MR1 ​​to MR9 as reference values ​​so that the sum of the advertising volume ratios NR1 to NR9 becomes 100 (%). The operator can determine the advertising volume ratio change values ​​by referring to the information contact path evaluation screen 54 shown in Figures 10(A) and 10(B).

[0137] In the advertisement volume ratio change receiving section 55B, the current advertisement volume (actual value) ratios MR1, MR2, MR3, MR4, MR5, MR6, MR7, MR8, and MR9 are displayed as initial values ​​for reference.

[0138] 11A shows an initial state in which no change in advertising volume ratio has been accepted. The advertising volume ratio change accepting section 55B displays the current advertising volume (actual value) ratios MR1, MR2, MR3, MR4, MR5, MR6, MR7, MR8, and MR9.

[0139] 11B shows a state in which a change to the advertising volume ratio has been accepted. The advertising volume ratio change accepting section 55B displays the changed advertising volume ratios NR1, NR2, NR3, NR4, NR5, NR6, NR7, NR8, and NR9.

[0140] (Response total increase / decrease output process S19 in response total increase / decrease output unit 190)

[0141] The total response increase / decrease output unit 190 outputs the increase / decrease amount or the increase / decrease rate of the total response using the change in the advertising volume ratio accepted by the advertising volume ratio change acceptance unit 180 and the path coefficient (S19).

[0142] The response total increase / decrease display unit 55C functions as a response total increase / decrease output unit 190.

[0143] When the advertising volume ratio change receiving unit 55B receives a change to the advertising volume ratios NR1, NR2, NR3, NR4, NR5, NR6, NR7, NR8, and NR9, the total response increase / decrease display unit 55C displays the advertising volume ratios NR1, NR2, NR3, NR4, NR5, NR6, NR7, NR8, and NR9 and the increase / decrease rate SN of the total response using the path coefficients. The unit of the total response increase / decrease rate SN is %, and it is expressed as a ratio (%) of the initial total response value SM, where SM is set to 100%. Therefore, if there is no increase or decrease in the total response, the total response increase / decrease rate SN will be 100 (%), the same as the initial total response value SM.

[0144] 11A shows the initial state where no change in the advertising volume ratio has been accepted. The total response increase / decrease display section 55C displays the initial value SM (100%) of the total response according to the current advertising volume (actual value) ratios MR1, MR2, MR3, MR4, MR5, MR6, MR7, MR8, and MR9.

[0145] 11B shows a state in which a change in the advertising volume ratio has been accepted. The total response increase / decrease display section 55C displays the increase / decrease rate SN (e.g., 150%) of the total response according to the changed advertising volume ratios NR1, NR2, NR3, NR4, NR5, NR6, NR7, NR8, and NR9.

[0146] In the second embodiment, the rate of increase or decrease of the total response is displayed, but the amount of increase or decrease of the total response may be displayed instead of the rate of increase or decrease of the total response.

[0147] According to the second embodiment, the rate or amount of increase or decrease in the total response can be estimated simply by performing an operation to obtain the past advertising volume and actual response values ​​of each advertising medium and then performing an operation to change the advertising volume ratio based on the actual advertising volume ratio value.

[0148] In the second embodiment, similarly to the first embodiment, the total response Y0 used in the path coefficient output unit 160 and the total response increase / decrease output unit 190 may be subjected to correction processing to remove seasonal influences.

[0149] Also in the second embodiment, similarly to the first embodiment, a correction process may be performed to add a residual effect to the advertisement volumes X1 to X9 of the advertisement media used in the advertisement volume ratio actual value output unit 170. [Explanation of symbols]

[0150] 100 Advertising effectiveness estimation device 110 Actual value acquisition unit 120 Regression analysis processing unit 130 Selection Section 140 Path Route Generation Unit 150 Path coefficient calculation section 200 Advertising effectiveness estimation device 160 Path coefficient output section 170 Advertising volume ratio actual value output section 180 Advertising Volume Ratio Change Reception Department 190 Total response increase / decrease output section

Claims

1. a performance value acquisition unit that acquires a performance value of the amount of advertising by advertising medium, which indicates the amount of advertising by advertising medium, a performance value of the response by advertising medium, which indicates the response by advertising medium, and a total performance value of the response by advertising medium; a regression analysis processing unit that performs analysis processing using a multiple regression analysis formula and / or a simple regression analysis formula with either the total response or the response by advertising medium as a target variable and one or more advertising amounts by advertising medium as explanatory variables based on the performance values ​​acquired by the performance value acquisition unit; a selection unit that selects a multiple regression analysis formula and / or a simple regression analysis formula having a fitness index of the multiple regression analysis formula and / or the simple regression analysis formula that is equal to or greater than a predetermined fitness index; When the objective variable of the multiple regression analysis formula and / or simple regression analysis formula selected by the selection unit is a response by advertising medium, a first advertising medium corresponding to the advertising volume by advertising medium, which is an explanatory variable of the multiple regression analysis formula and / or simple regression analysis formula, becomes the advertising medium that is the source of the indirect effect, and a second advertising medium corresponding to the response by advertising medium, which is the objective variable of the multiple regression analysis formula and / or simple regression analysis formula, becomes the advertising medium that is the destination of the indirect effect, by generating an indirect path from the advertising volume by advertising medium corresponding to the first advertising medium corresponding to the advertising volume by advertising medium corresponding to the second advertising medium among the path routes between a plurality of advertising volumes by advertising medium and total responses; a path route generation unit that generates a direct path from the advertising volume by advertising medium corresponding to the third advertising medium to the total response among the path routes, so that when the objective variable of the multiple regression analysis formula and / or the simple regression analysis formula selected by the selection unit is the total response, the third advertising medium corresponding to the advertising volume by advertising medium, which is the explanatory variable of the multiple regression analysis formula and / or the simple regression analysis formula, becomes the advertising medium that is the direct effect source, and the total response, which is the objective variable of the multiple regression analysis formula and / or the simple regression analysis formula, becomes the direct effect destination; a path coefficient calculation unit that calculates path coefficients for the indirect path and the direct path among the path routes based on the performance values ​​acquired by the performance value acquisition unit; An advertising placement effectiveness estimation device comprising:

2. The regression analysis processing unit performs analytical processing using a multiple regression analysis formula and / or a simple regression analysis formula using explanatory variables limited to a predetermined number less than the number of all advertising media. The advertising effectiveness estimation device according to claim 1 .

3. The total response and the response by advertising medium used in the regression analysis processing unit, the path route generation unit, and the path coefficient calculation unit are corrected to remove seasonal influences. The advertising effectiveness estimation device according to claim 1 .

4. The amount of advertising by advertising medium used in the regression analysis processing unit, the path route generation unit, and the path coefficient calculation unit has been subjected to correction processing to add residual effects. The advertising effectiveness estimation device according to claim 1 .

5. An advertising volume ratio actual value output unit that outputs an advertising volume ratio actual value indicating the ratio of advertising volumes of each advertising medium based on the actual values ​​acquired by the actual value acquisition unit; an advertising volume ratio change receiving unit that receives a change in the advertising volume ratio with respect to the actual advertising volume ratio value; a total response increase / decrease output unit that outputs the amount of increase / decrease or the rate of increase / decrease of the total response using the change in the advertising volume ratio accepted by the advertising volume ratio change acceptance unit and the path coefficient; The advertisement placement effectiveness estimation device according to claim 1, comprising:

6. 6. The advertising effectiveness estimation device according to claim 5, wherein the path coefficient, the actual value of the advertising volume ratio, and the increase / decrease amount or the increase / decrease rate of the total response are displayed on a screen.

7. The total response used in the total response increase / decrease output unit is subjected to a correction process to remove seasonal influences. The advertisement placement effectiveness estimation device according to claim 5 .

8. The advertisement volume of each advertising medium used in the advertisement volume ratio actual value output unit is corrected by adding a residual effect. The advertisement placement effectiveness estimation device according to claim 5 .

Images