JP2003030537A - Ad response prediction system and method - Google Patents

Abstract

(57) [Summary] An object of the present invention is to provide an advertisement response prediction system and a method thereof capable of predicting the number of responses with high accuracy. SOLUTION: A target number of persons is calculated by multiplying the number of subscribers of a newspaper scheduled to publish an advertisement by a population ratio (target number ratio) of recruitment targets of the recruitment advertisement (symbol 1). Next, the target number of persons is calculated by multiplying the target number of persons by the advertisement attention rate for each placement mode (symbol 2). For this ad audience,
By multiplying the decay rate of the attention rate for each placement, the number of advertisement attention per placement is calculated (symbol 3). Finally, the number of predicted responses is calculated by multiplying the number of advertisers by placement times by the reference response rate (symbol 4).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system and method for predicting the number of responses to recruitment advertisements contained in media such as newspapers.

[0002]

2. Description of the Related Art When planning an advertisement placement plan such as recruitment of human resources or mail order, it is extremely important to predict the response rate of how many people actually respond (apply, apply, etc.) to the number of circulations. . This is because the advertiser can reduce unnecessary time and cost for advertisement placement by obtaining a highly accurate predicted response number in advance.

A conventional general method for predicting the number of responses is to predict the accuracy of the number of responses by continuously accumulating data on the number of responses for the number of copies of the newspaper that issues the advertisement, for example, in mail order in newspaper advertisements. Is used.

[0004]

However, in the conventional method of predicting the number of responses, it is premised that the similar advertisement placement is continued for a long period of time to improve the accuracy. If the data of the number of responses is small, it is difficult to make a highly accurate prediction.

[0005] One of the advertisement types in which the data of the number of responses is less accumulated is a clinical trial advertisement recently approved by the Ministry of Health and Welfare. This clinical trial advertisement is mainly
To recruit subjects for clinical trials in newspaper advertisements, etc.,
It is attracting attention as an attempt to speed up clinical trials by eliminating the difficulty of recruiting subjects, which had been the cause of delaying clinical trial progress. Other advertisements include, for example, recruitment advertisements for occupations having special skills. In any case, due to the nature of the content that requests a response, it is temporally difficult to accumulate similar data and improve the accuracy of predicting the number of responses in such an advertisement, and it will be helpful. It is rare that similar data is available.

The present invention solves the above problems, and an advertisement response prediction system capable of highly accurately predicting the number of responses even when the data of the number of responses is less accumulated, and the same. The purpose is to provide a method.

[0007]

[Means for Solving the Problems and Effects of the Invention] 1) The advertisement response prediction system of the present invention is recorded in association with the media user number information acquisition means for acquiring the media user number information and the advertisement attribute information. Also, an advertisement attention number estimation information acquisition means for acquiring the advertisement attention number estimation information regarding the ratio between the number of media users and the estimated attention attention number, and the response number estimation regarding the ratio between the advertisement attention number and the advertisement estimation response number When the number-of-responses estimation information acquisition unit for acquiring information and the advertisement attribute information are input, the media user number recording unit and the advertisement attention number estimation information recording unit are referred to, thereby responding to the advertisement attribute information. The number of advertisement attention persons calculating means for calculating the number of advertisement attention persons, the calculated number of advertisement attention persons and the response recorded in the response number estimation information recording means By referring to the Nsu number estimation information, and comprising the predicted response number calculating means for calculating a predicted number of responses corresponding to the advertisement attribute information.

As a result, when calculating the predicted number of responses, the predicted number of responses is calculated after extracting only the number of advertisement attention users who may respond from the number of media users. It Therefore, the advertisement response prediction system takes a step of extracting only the estimated number of noticed advertisements, which directly affects the number of responses, before finally calculating the number of responses. It is possible to perform prediction with higher accuracy than when directly predicting.

Also, the advertisement attribute information as a factor that determines the estimated number of noticed advertisements is used in the processing by the number of noticed advertisements calculation means. Therefore, the response number estimation information relating to the ratio between the number of advertisement attention persons and the advertisement estimated response number can be standardized as a fixed value that does not depend on the advertisement attribute information.

Further, according to the advertisement response prediction system, even when the advertisement attributes are different, the number of advertisement attention persons corresponding to each advertisement attribute is calculated, and based on the standardized response number estimation information. Then, the number of responses is calculated. Therefore, it is possible to accurately predict the number of responses even for various variations of advertisement attributes.

5) The advertisement attribute information of the present invention is characterized by including advertisement placement form information.

Thus, the advertisement attention number calculation means calculates the estimated advertisement attention number with high individual correspondence according to the advertisement attention degree of the media user due to the difference in advertisement placement form. Therefore, it is possible to accurately predict the number of responses even for various variations in the advertisement placement form.

6) The advertisement attribute information of the present invention is characterized in that it includes advertisement placement frequency information.

Thus, the advertisement attention number calculation means calculates the advertisement estimated attention number corresponding to the advertisement placement frequency information. Therefore, when submitting a plurality of times, it is possible to accurately predict the number of responses as a whole or for each submitting time.

7) The advertisement attribute information of the present invention is characterized in that the advertisement includes response target person category information indicating a category of a target person who requests a response.

Accordingly, the advertisement attention number calculating means calculates the advertisement estimated attention number after extracting the number of people who may be candidates for the response to the advertisement. Therefore, it is possible to accurately predict the number of responses even with respect to variations in advertisements that require responses from target persons in various categories.

8) The response number estimation information recording means of the present invention records the response number estimation information relating to the ratio of the advertisement attention number and the advertisement estimated response number in association with the response target person category information. It is characterized by that.

As a result, the predicted response number calculation means calculates the predicted response number in consideration of the tendency of the number of responses for each category of the response candidates for the advertisement. Therefore, it is possible to accurately predict the number of responses even with respect to variations in advertisements that require responses from target persons in various categories.

[0019]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of an advertisement response prediction system according to the present invention will be described. The present embodiment exemplifies the calculation of the number of predicted responses in clinical trial advertisements, which makes it possible to calculate the number of highly accurate responses according to individual clinical trial advertisement plans and the like, and also in terms of time and cost. You can also design the optimal clinical trial advertising plan.

Hereinafter, first, the outline of the predicted response calculation processing according to the embodiment, the hardware configuration, the database configuration, the correspondence between the terms described in the claims and the embodiment will be described, and then the present embodiment. The processing contents and the like will be explained. Table of contents 1. 1. Outline of predicted response number calculation processing Hardware configuration 3. Database configuration 4. 4. Correspondence between terms described in the claims and the embodiments. Description of Calculation Processing of Predicted Response Number According to Embodiment 6. Effects of the embodiment 7. Other Embodiments, etc. -------- Outline of predicted response number calculation processing 1-1. Conventional Prediction Response Number Calculation Method As a premise for explaining the outline of the prediction response number calculation processing according to the present embodiment, first, a conventional prediction response number calculation method will be described. The conventional method of calculating the predicted number of responses is generally performed by the method shown in FIG. 1 when placing an advertisement in a newspaper.
In the conventional calculation of the expected number of responses, the number of applicants to the number of newspaper subscribers, that is, the ratio of the number of responses to the number of advertisements, is calculated by continuously accumulating data on the number of responses to the number of circulations of newspapers that place advertisements. Increase the accuracy of (response rate). If a similar advertisement is scheduled to be submitted, the expected response number is calculated by multiplying the planned advertisement number by the response rate. Therefore, in the conventional response number prediction, accumulating the response rate data a plurality of times plays an important role for improving the prediction accuracy. However, with the conventional method, when the amount of data of past similar advertisements is small,
There is a drawback in that highly accurate prediction cannot be performed.

1-2. Response Number Prediction Method According to Embodiment On the other hand, the response number prediction method according to this embodiment can perform highly accurate prediction even when the amount of data of past similar advertisements is small. The outline of this method is not directly calculating the number of responses from the number of newspaper subscribers, but the factors such as how many of the newspaper subscribers actually see the advertisement and how many of those who saw the advertisement. The final predicted number of responses is calculated after considering the factor of whether or not to respond. Here, regarding the calculation of how many people actually see the advertisement, factors that consider the number of newspaper subscribers applicable to recruitment, and the same advertisement by multiple advertisement placement It will be decomposed into factors that take into consideration the number of people who will not respond.

The method unique to the present invention is as follows.
It was made based on the inventor's original knowledge.

(1) The ratio of the number of notable advertisements to the number of subscribers statistically depends on the advertisement output form. That is, the ratio of the number of people who pay attention to the advertisement (the number of people who think that they have seen the advertisement) to the number of contact persons with the advertisement medium is, Depends on the type of submission.

(2) The ratio of the number of noticed advertisements to the number of subscribers statistically decreases with a similar tendency each time the number of advertisements is repeated.

(3) The number of people who are aware of the advertisement and who have come to the consciousness of deciding whether to apply for it or not to apply for it (hereinafter referred to as "the number of people who understand and judge by paying attention to the advertisement") Of these, the ratio of the number of people who respond to the advertisement (application, application, etc.) can be standardized for each medium in which the advertisement is posted and each category of the target person who views the advertisement (in the present embodiment, the medium and the target disease). Standardized by type).

A specific response number predicting method according to this embodiment, which is made based on the above knowledge, will be described with reference to FIG.

The target number of people is calculated by multiplying the number of subscribers of the newspaper planning to place the advertisement by the population ratio (target number ratio) of the recruitment target persons of the advertisement for advertisement (symbol 1). Next, the target number of people is multiplied by the advertisement attention rate for each printing form to calculate the advertisement attention number (symbol 2). Multiplying the number of noticed advertisements by the decline rate of the notice rate for each advertisement, the number of advertisement notices for each advertisement (the number of people who newly understand and judge the advertisement for each advertisement) is calculated. Yes (symbol 3). Finally, the number of predicted responses is calculated by multiplying the number of advertisement notices by the number of submissions by the reference response rate (symbol 4). The standard response rate is a standardized ratio of the number of responses to the number of people who understand and judge the advertisement. By these processes, the predicted response number is finally calculated for each advertisement submission time. Details of each process and the contents such as the advertisement attention rate, the attenuation rate, and the reference response rate will be described later.

1-3. Prediction Response Number Calculation Model As described above, the outline of the calculation processing of the prediction response number according to the present embodiment is understood by paying attention to the advertisement from the target number of the population when a certain advertisement is placed. The number of people to be judged is extracted, and this number is multiplied by the reference response rate. Further, in the present embodiment, as described above, the advertisement is focused on
The number of people to be judged is calculated in consideration of the fact that it decreases with the number of times the advertisement is placed. In the following, the prediction response number calculation model is calculated as follows: (1) calculation of the number N _i of people who newly pay attention to and judge the advertisement by the i-th advertisement advertisement, and (2) prediction response at the i-th advertisement advertisement The calculation of the number will be described separately.

(1) The calculation model N _i of the number N _i of people who newly understand and judge the advertisement by the i-th advertisement placement can be expressed by the equation 1.

[0030]

[Equation 1] The contents of each parameter are as follows.

N: Number of target newspaper subscribers (target number of people) in the advertisement placement area a (s): Advertisement attention rate s _i when the publication form is s: i-th publication form λ: Decay parameter (Advertisement target Probability of understanding and deciding the advertisement) p _i : Probability of understanding and deciding by paying attention to the advertisement at the time of the i-th advertisement placement (p _i = λa (s _i )) The concept of having done is a concept that has led to the awareness of having seen an advertisement, while the concept of having “understood and judged an advertisement” means whether to apply (response) or not to see it. It is a concept that has led to the consciousness that determines the reaction.

As shown in Formula 1, the calculation of N _i (the number of people who newly understand and judge the advertisement at the time of the i-th advertisement placement) is calculated from the target number of people until the i−1-th advertisement placement. After excluding the number of people who have already understood and judged the advertisement, the advertisement attention rate is multiplied. in this way,
It is decided to exclude the number of people who understand and judge the advertisement at the (i-1) th time, and the person who has become conscious of deciding whether to respond to it after seeing the advertisement It is based on the psychological assumption that when the same advertisement is touched, it does not respond to the advertisement and is less likely to respond. This psychological assumption is based on the statistical result obtained by the inventor's own investigation that the ratio of the number of applicants (the number of responses) to the advertisement attention person decreases as the number of submissions increases.
This is an assumption made to model the cause of the attenuation.

The above model can be more simply expressed as a model in which N _i is obtained by multiplying N by the advertisement attention rate p _i which is attenuated at each advertisement submission time. In this case, N _i can be expressed by Equation 2.

[0034]

[Equation 2] ω _i : Attenuation rate of the advertisement attention rate at the time of the i-th advertisement placement Here, ω _i can be expressed by Expression 3.

[0035]

[Equation 3] (2) Calculation mode for the number of predicted responses when the i-th advertisement is placed
Dell N calculated by Equation 1_iFrom the i-th advertisement
Predicted response number R _iThe model to calculate is
It can be represented by Equation 4.

[0036]

[Equation 4] Here, the content of each parameter is as follows.

N ₀ λa (s ₀ ): Standardized number of people who understand and judge by paying attention to advertisements R ₀ : Standardized number of responses R ₀ / N ₀ λa (s ₀ ) , Which is the standard response rate, which is a standardized ratio of the number of people who actually responded to the number of people who understood and judged by focusing on the advertisement in the past advertisement cases similar to the planned advertisement plan. Is. The data of the number of people who understand and judge by paying attention to the advertisement can be aggregated by, for example, a telephone survey.

Thus, the response rate r _i = R _i / N
Is

[Equation 5] Can be expressed as

For the calculation model of the number N _i of people who newly noticed and judged the advertisement by the i-th advertisement placement in the above (1), either Expression 1 or Expression 2 is adopted. However, in the embodiment, the model of Expression 2 is adopted for description.

2. Hardware Configuration FIG. 3 is an example of a hardware configuration of the clinical trial advertisement response prediction system 100 as the advertisement response prediction system according to the present embodiment. The clinical trial advertisement response prediction system 100 includes a CPU 10, a RAM 20, a hard disk 12, a ROM 14, a keyboard 16, and a display 18. The CPU 10 controls not only the predicted response number calculation process but also the entire system. RAM20
Provides a work area and the like for the CPU 10. The keyboard 16 receives a user's operation input. In addition to the predicted response number calculation processing program, the hard disk 12 records a reference response rate database (hereinafter, “database” is referred to as “DB”) 54, an attenuation rate DB 52, a subscriber DB 50, a disease rate DB 56, and an attention rate DB 58. .

3. Database configuration 3-1. Subscriber DB, Disease Rate DB, Attention Rate DB, Attenuation Rate DB FIG. 4 shows a configuration example of the subscriber DB 50. Subscriber DB5
0 records the number of subscribers for each medium category, and the number of subscribers is classified by area, sex, and age group.

FIG. 5 shows an example of the structure of the disease rate DB 56.
The disease rate DB 56 records the disease rate (disease population ratio) for each target disease type, and the disease rate is the area, sex,
It is divided by age group. This data is, for example,
Data from the Ministry of Health and Welfare may be used.

FIG. 6 shows a configuration example of the attention rate DB 58.
The attention rate DB 58 records the attention rate for each medium category, and the attention rate is classified by newspaper name, area, size, and color.

FIG. 7 shows a configuration example of the attenuation rate DB 52.
The attenuation rate DB 52 records the attenuation rate of the attention rate for each number of submissions.

The structure of each database described above is as follows:
The configuration is not limited to this, and other configurations may be adopted. For example, in order to further improve the accuracy of predicting the number of responses, in the interest rate DB 58, if the interest rate differs depending on the age group, target disease type, family structure, etc., the database configuration should be improved based on such data. It may be subdivided.

Further, the attenuation rate DB 52 may also be recorded not as a separate database from the attention rate DB 58 but as a combination of both. Specifically, the attention rate recorded in the attention rate DB 58 of FIG. 6 may be used as the data recorded as the attention rate for each submission time. Also, the attenuation of the attention rate has a similar tendency statistically as described above, but a slow attenuation pattern,
A plurality of decay patterns, which are fast decay patterns, may be stored in a database to improve the adaptability to individual advertising plans.

3-2. Reference Response Rate DB FIG. 8 shows a configuration example of the reference response rate DB 54. The reference response rate DB 54 records the reference response rate in association with the target disease type and the medium classification. The reference response rate is a standardization of the ratio of the number of people who actually responded to the number of people who understood and judged by paying attention to the advertisement, which was obtained by measuring the response rate in the past, as described in Equation 4. .

This reference response rate is standardized after excluding the factors of the form of submission. The reason that the response rate can be standardized by excluding the factor of the advertisement type is that the advertisement attention rate recorded in the attention rate DB 58 depends on the advertisement type.

The reference response rate is the proportion of the number of people who respond to the advertisement, which is understood and judged by paying attention to a certain advertisement. Therefore, when obtaining the data of the reference response rate, when obtaining the data. It is also necessary to consider the number of submissions of. Specifically, when totaling the number of responses after the second submission, the number of responses is the total of the two submissions, so the “number of people who understand and judge advertisements” is the first. It is necessary to calculate as the total of the number of people who have noticed and understood / determined the advertisement by placing the advertisement, and the number of persons who have newly noticed and understood / determined the advertisement by the second advertisement (in consideration of the decline of the attention rate).

The structure of the database described above is not limited to this, and other structures may be adopted. In the present embodiment, the target disease type and the medium are used as factors for standardizing the reference response rate, but if the response rate depends only on the target disease type and does not depend on the medium, only the target disease type factor is used. May be standardized, or as other factors, new drug to be tested, age group, sex, living area, applicant reward type, etc. may be used.

4. Correspondence between Terms Described in Claims and Embodiments Correspondence between terms described in claims and embodiments is as follows.

The "advertisement response prediction system" is shown in FIG.
It corresponds to the clinical trial advertisement response prediction system 100.
The “media user” corresponds to the newspaper subscriber in the subscriber DB 50 of FIG. 4, and the “response number” corresponds to the predicted response number by submission time in step 919 of FIG. 9.

"Media user number information acquisition means" corresponds to the processing performed by the CPU 10 of the clinical trial response prediction system 100 in steps S903 and S905 of FIG. The "advertisement number estimation information acquisition means" corresponds to the processing performed by the CPU 10 in steps S909 and S913 of FIG. The “response number estimation information acquisition means” corresponds to the processing performed by the CPU 10 in step S917 of FIG.

The "medium user number information recording means" corresponds to the subscriber DB 50 of FIG.

The "advertisement number estimation information" is the attention rate recorded in the attention rate DB 58 of FIG. 6 and the attenuation rate DB5 of FIG.
2 corresponds to the decay rate recorded. The “advertisement number estimation information recording means” corresponds to the attention rate DB 58 in FIG. 6 and the attenuation rate DB 52 in FIG. 7.

The "response number estimation information" corresponds to the reference response rate recorded in the reference response rate DB 54 of FIG. 8, and the "response number estimation information recording means" corresponds to the reference response rate DB 54 of FIG.

The "advertising attribute information" is the step S90 in FIG.
The target disease type, advertisement placement area, and advertisement plan in 1 correspond. The “advertising person number calculating means” corresponds to the processing performed by the CPU 10 of the clinical trial advertisement response prediction system 100 in steps S909, S911, S913, and S915 in FIG. "Prediction response number calculation means"
Corresponds to the processing performed by the CPU 10 in steps S917, S919, and S921 in FIG.

The "advertising form information" is the step S in FIG.
The advertisement plan in 901 corresponds, and specifically, FIG.
The information such as medium classification, newspaper name, area, size, color, etc. corresponds to. The “advertisement number of times information” corresponds to the advertisement plan in step S901 of FIG. 9, and specifically, the information of the number of times of advertisement placement in FIG. “Response target person category information” is shown in step S901 in FIG.
And the information of the target disease type in FIG. 5 corresponds.

The "advertising information recording section" is a column in the attention rate DB 58 of FIG. 6 in which each information of medium category, newspaper name, area, size and color is recorded, or FIG.
The column in which the information on the number of submissions is recorded in the attenuation rate DB 52 corresponds to the “advertisement number estimation information recording unit”
Corresponds to the column in the attention rate DB 58 of FIG. 6 where the attention rate information is recorded or the column in the attenuation rate DB 52 of FIG. 7 where the attenuation rate information is recorded.

The "advertising attribute information recording section" is a target disease type in the reference response rate DB 54 of FIG. 8, or
The column in which each information of the medium classification is recorded corresponds to, and the “response number estimation information recording unit” is the reference response rate DB5.
The column in which the information on the reference response rate in 4 is recorded corresponds to.

5. Description of Predicted Response Number Calculation Processing According to Embodiment The predicted response number calculation processing according to this embodiment will be described with reference to FIG.
A description will be given based on the flowchart. Here, an example of calculating the predicted number of responses from male diabetic patients aged 35 to 49 when a clinical trial advertisement is placed in the Kanto area will be described.

As a premise of the calculation processing of the predicted response number by the CPU 10, the user of the system needs to input the “target disease type”, the “advertising area”, and the “advertising plan”. An example of 18 screen displays is shown in FIG. In FIG. 10, "target disease type" indicates the disease, sex, and age group of the subject to be recruited, "advertising area" indicates the area name in which newspaper advertisements are placed, and "advertising plan" “” Means the form of advertisement when a newspaper advertisement is submitted (either monochrome or color advertisement, the number of steps on the page of the advertisement, the number of advertisements, etc.)
Is shown.

The CPU 10 of the clinical trial advertisement response prediction system 100 determines whether or not the user inputs "target disease type", "advertising area", and "advertisement plan" (step S901). If there is such input, the CPU 10 refers to the newspaper subscriber DB 50 and extracts information on the number of target area persons (the number of newspaper subscribers of "male aged 35 to 49 years in the Kanto area", see FIG. 4) (step 4). S9
03). The CPU 10 refers to the disease rate DB 56 and extracts information on the target disease rate (diabetic morbidity of males aged 35 to 49 in the Kanto area, see FIG. 5) (step S90).
5). The following processing is performed for each prefecture in the Kanto area, and the total of these is output as the final predicted number of responses.

The CPU 10 calculates the target number of people (target area number of people x control disease rate (%)) (step S9).
07). By this calculation, the number of persons who are subject to clinical trial, that is, persons who are subject to response and who are likely to come into contact with the clinical trial advertisement to be submitted is calculated.

The CPU 10 refers to the attention rate DB 58 and extracts the information of the attention rate (attention rate of A newspaper color in 5 steps, see FIG. 6) (step S909). The CPU 10 calculates the number of advertisers (target number of people × attention rate (%)) (step S911). By this calculation, the number of people who pay attention to the clinical trial advertisement scheduled to be placed is calculated from the target number of people.

The CPU 10 refers to the attenuation rate DB 52 and extracts the information of the attenuation rate (the attenuation rate for the number of times of submission 1 to 2 times, see FIG. 7) (step S913). The CPU 10 calculates the number of advertisement noticers by advertisement submission number (the number of advertisement noticers × the attenuation rate (%)) (step S915). By this calculation, the number of people who newly pay attention to the advertisement and understand / determine (the number of advertisement attention by advertisement number) is calculated every time the advertisement number of the clinical trial advertisement to be submitted is increased.

The CPU 10 uses the reference response rate DB 54.
Information of the reference response rate (reference response rate when the diabetic patient understands and judges by paying attention to the newspaper clinical trial advertisement, see FIG. 8) is extracted (step S917). CP
U10 calculates the number of predicted responses for each submission number (the number of advertisement noticers for each submission × reference response rate (%)) (step S919). By this calculation, the expected number of responses for each submission of the clinical trial advertisement scheduled to be submitted is calculated.
The CPU 10 causes the display 18 to display the predicted response number for each submission count, and ends the predicted response number calculation process.

Here, FIG. 11 shows an example of the screen of the display 18 when the predicted response number for each submission number is output.
In the table of FIG. 11, "target number of people" represents the number of target diseased persons among the newspaper subscribers who publish the targeted advertisement, and "response index" is the attention rate x attenuation rate x standard. Obtained by the response rate. “Number of responses” represents the number of predicted responses for each number of submissions,
The “cumulative number of responses” represents the number of persons who have accumulated the number of responses for each submission, and the tendency is graphically displayed in a “response chart”. Although the total of the Kanto area is displayed in the figure, it may be displayed for each prefecture.

6. Effects of the Embodiment In the present embodiment, when calculating the number of predicted responses,
The predicted number of responses is calculated after extracting only the number of advertisement noticers who may respond from the number of newspaper subscribers (see step S911 in FIG. 9). Therefore, the clinical trial advertisement response prediction system 100 takes the step of extracting only the number of advertisement noticers who directly influence the number of responses before finally calculating the number of responses, and thus the number of responses from the total number of newspaper subscribers is calculated. It is possible to perform prediction with higher accuracy than when directly predicting.

In the present embodiment, the advertisement plan (see step S901 in FIG. 9) as a factor that determines the number of advertisement attention is referred to in the calculation of the number of advertisement attention. Therefore, the reference response rate can be standardized as a fixed value that does not depend on the advertisement plan (see FIG. 8).

In the present embodiment, even when different advertisement plans are input, the number of noticed advertisements corresponding to each advertisement plan is calculated, and then the number of responses is calculated based on the standardized reference response rate. It Therefore, it is possible to accurately predict the number of responses even for various variations of the advertising plan.

In particular, the number of noticed advertisements, which affects the final number of responses, is calculated independently as depending on the advertisement plan. Therefore, when the data accumulation of the number of responses of similar clinical trial advertisements is small. However, it is possible to simulate the number of responses to various advertising plans.

In this embodiment, the predicted number of responses is calculated for each number of advertisement placements (step S in FIG. 9).
919). Therefore, when submitting a plurality of times, it is possible to accurately predict the number of responses as a whole or for each submitting time.

The present embodiment is characterized in that the reference response rate is recorded in association with the target disease type and the medium classification (see FIG. 8). As a result, the number of predicted responses is calculated in consideration of the tendency of the response due to the disease type and the medium. Therefore, it is possible to accurately predict the number of responses even for variations in clinical trial advertisements relating to various diseases.

7. Other Embodiments In this embodiment, the final number of applications for clinical trial advertisements published in newspapers is predicted, but the present invention is not limited to this. As another embodiment, the number of responses can be predicted even when a magazine is used as a medium or a recruitment advertisement such as a general job advertisement is used as an advertisement. For example, when recruiting a job for a job that requires special skills, a database of the population ratio of the number of persons having such special skills (for example, the number of qualified persons) should be created as in the example illustrated in FIG. Good. In addition, the same one as the attention rate database exemplified in FIG.
If the ratings for M viewers are sorted, it is possible to accurately predict the number of responses to the solicited CM.

In the present embodiment, as a premise for predicting the final number of responses, the number of advertisement attention persons is calculated (see step S911 in FIG. 9). Therefore, the present system is not limited to the case of predicting the final number of responses, but can also be used for predicting the advertisement attention rate or the advertisement viewing rate by a specific target person.

In the present embodiment, the number of predicted responses for each submission time in the target area is calculated, but if the required number of responses is input in advance, the required number of responses and the predicted number of responses will be calculated. A comparison can be made. For example, if the acceptable number of medical institutions participating in the clinical trial is entered for each target area as the required number of responses, it is possible to compare the acceptable number of people and the expected number of responses for each target area. You can design the optimal clinical trial advertising plan for the target area.

In order to design a more cost-effective advertisement placement plan, the cost of the advertisement plan may be calculated in the calculation processing of the advertisement response prediction system according to the present invention. Specifically, the advertisement posting fee for each advertisement placement form is stored in a database in advance, and the advertisement posting fee for the entire number of submitted advertisements is calculated in accordance with the advertisement plan input in step S901 in FIG.
The result is output together with the number of responses. According to this, when simulating advertisement plans of various variations, it is possible to easily determine the optimum advertisement plan from the viewpoint of both the number of predicted responses and cost efficiency.

In this embodiment, the algorithm of the predicted response number calculation process performed by the CPU 10 of the clinical trial advertisement response prediction system 100 is shown by the flowchart of FIG. 9, but the algorithm is not limited to this. As another embodiment, for example, the process of multiplying the disease rate (step S907 in FIG. 9) may be performed after the process of calculating the number of advertising attention. In any case, what can be considered as an independent event, such as the step of multiplying the disease rate, the step of multiplying the attention rate, the step of multiplying the reference response rate, etc., is the final number of responses regardless of the order of the calculation algorithm. Can be determined.

As another embodiment, the advertisement response number prediction system according to the present invention may be connected to a network. Specifically, the advertisement response number prediction server
It suffices to receive the advertisement plan or the like input by the client and transmit the result of the predicted response number according to the plan. At this time, the subscriber D in the present embodiment
B50, attenuation rate DB52, reference response rate DB54,
The disease rate DB 56 and the attention rate DB 58 may be recorded in a server different from the advertisement response number prediction server.

In this embodiment, the program for the operation of the CPU 10 is stored in each hard disk 12. However, the program of the hard disk 12 is read from the CD-ROM in which the program is stored and installed in the hard disk or the like. Good. Also, CD-
In addition to ROM, floppy disk (F
A program such as D) or an IC card may be installed from a computer-readable recording medium. Further, the program can be downloaded using a communication line. In addition, by installing the program from the CD-ROM, the program stored in the CD-ROM is not directly executed by the computer, but the program stored in the CD-ROM is directly executed. You may do it.

It should be noted that the programs executable by the computer include programs that can be directly executed by just installing them, but programs that need to be converted to another form (for example, decompressed data compressed programs). Etc.), and those that can be executed in combination with other module parts are also included.

The hard disk 12 shown in this embodiment.
For each database recorded in
The data may be recorded via a recording medium such as an FD or an IC card, or may be downloaded using a communication line.

The calculation of the number of predicted responses according to the present invention can be performed by the calculation algorithm illustrated in FIG. 9, but this can be executed without using a computer system.

[Brief description of drawings]

FIG. 1 is a diagram showing a conventional advertisement response prediction method.

FIG. 2 is a diagram showing a concept of an advertisement response number prediction method of the present invention.

FIG. 3 is a diagram showing an overall configuration of a clinical trial advertisement response prediction system 100 according to an embodiment.

FIG. 4 is a diagram showing a subscriber database 50.

FIG. 5 is a diagram showing a disease rate database 56.

FIG. 6 is a diagram showing an attention rate database 58.

FIG. 7 is a diagram showing an attenuation rate database 52.

FIG. 8 is a diagram showing a reference response rate database 54.

FIG. 9 is a diagram showing a flowchart of a predicted response number calculation process.

FIG. 10 is a diagram showing a screen for inputting a clinical trial advertisement plan.

FIG. 11 is a diagram showing a screen on which a result of predicting the number of clinical trial advertisement responses is output.

[Explanation of symbols]

100 ... Clinical trial advertisement response prediction system 10 ... CPU 50 ... Subscriber DB 52 ... Attenuation rate DB 54 ... Standard response rate DB 56 ... Disease rate DB 58 ・ ・ ・ Attention rate DB

Claims (14)

[Claims] 1. An advertisement response prediction system for predicting the number of responses of a media user to an advertisement contained in a medium using a computer, wherein the advertisement response prediction system is a medium for acquiring information on the number of media users. User number information acquisition means, advertisement attention number estimation information acquisition means for acquiring advertisement attention number estimation information relating to the ratio between the number of media users and the advertisement estimation attention number recorded in association with the advertisement attribute information , A response number estimation information acquisition means for obtaining response number estimation information relating to the ratio of the advertisement attention number and the advertisement estimated response number, when the advertisement attribute information is input, the media user number acquisition means and the advertisement attention number By referring to the estimated information acquisition means, the advertisement number for which the advertisement attention number corresponding to the advertisement attribute information is calculated. Person number calculation means, a prediction for calculating a predicted response number corresponding to the advertisement attribute information by referring to the calculated advertisement attention number and the response number estimation information acquired by the response number estimation information acquisition means An advertisement response prediction system comprising: a response number calculation means. 2. A recording medium in which a program for causing a computer to function as an advertisement response prediction system is recorded, wherein the recording medium is a medium user number information acquisition means for acquiring the medium user number information. , The advertisement attention number estimation information acquisition means for acquiring the advertisement attention number estimation information regarding the ratio between the number of media users and the advertisement estimation attention number recorded in association with the advertisement attribute information, and the advertisement attention number Response number estimation information acquisition means for acquiring response number estimation information related to the ratio with the advertisement estimated response number, and when the advertisement attribute information is input, the media user number recording means and the advertisement attention number estimation information recording means An advertisement attention number calculation means for calculating the number of advertisement attention corresponding to the advertisement attribute information by referring to: An advertisement provided with a predicted response number calculation means for calculating the predicted response number corresponding to the advertisement attribute information by referring to the number of advertisement attention persons and the response number estimation information recorded in the response number estimation information recording means A recording medium that records a program that causes it to function as a response prediction system. 3. A program for causing a computer to function as an advertisement response prediction system, wherein the program corresponds to a media user number information acquisition unit for acquiring media user number information, and advertisement attribute information. The advertisement attention number estimation information acquisition means for acquiring the advertisement attention number estimation information regarding the ratio between the number of media users and the advertisement estimation attention number, which are recorded, and the advertisement attention number and the advertisement estimation response number. Response number estimation information acquisition means for acquiring response number estimation information regarding the ratio, when the advertisement attribute information is input, by referring to the media user number recording means and the advertisement attention number estimation information recording means, An advertisement attention number calculating means for calculating the advertisement attention number corresponding to the advertisement attribute information, and the calculated advertisement attention number and Function as an advertisement response prediction system including a predicted response number calculation unit that calculates the predicted response number corresponding to the advertisement attribute information by referring to the response number estimation information recorded in the response number estimation information recording unit. A program to let you. 4. The advertisement response prediction system according to claim 1, further comprising: a media user number information recording means for recording the media user number information; and the advertisement attention number estimation information in association with the advertisement attribute information. And a response number estimation information recording unit that records the response number estimation information. 5. The advertisement response prediction system, the recording medium, or the program according to claim 1, wherein the advertisement attribute information includes advertisement placement form information. 6. The advertisement attribute information according to any one of claims 1 to 5, wherein the advertisement attribute information includes advertisement placement frequency information. 7. The advertisement attribute information according to any one of claims 1 to 6, wherein the advertisement attribute information includes response target person category information indicating a category of a target person for which an advertisement requests a response. 8. The response number estimation information recording means according to claim 1, wherein the response number estimation information relating to a ratio of the advertisement attention number and the advertisement estimated response number is the response target category. Characterized by recording in association with information. 9. An advertisement placement information recording unit for recording advertisement placement form information or advertisement placement count information, and an advertisement attention number relating to a ratio between the number of media users and the estimated advertisement attention number in association with the advertisement placement information. Data having a data structure including an estimated information recording unit. 10. An advertisement attribute information recording section for recording advertisement attribute information, and response number estimation information for recording response number estimation information relating to a ratio between the number of advertisement attention persons and the advertisement estimated response number in association with the advertisement attribute information. Data having a data structure including a recording unit. 11. A recording medium on which data having the data structure recorded on the recording medium according to claim 9 is recorded. 12. An advertisement response prediction method for predicting the number of responses of a media user to an advertisement included in a medium by using a computer, wherein the computer receives the input advertisement form information or the input advertisement. Based on the number of advertisements, the advertisement estimated attention number estimated to pay attention to the advertisement is calculated from the number of media users, and based on the response estimation information regarding the ratio between the advertisement attention number and the advertisement estimated response number, An advertisement response prediction method comprising: calculating a predicted response number estimated to respond to an advertisement from the calculated estimated advertisement attention number. 13. A media audience rating prediction method for predicting a media audience rating of a media target of the media among a number of media users using a computer, wherein the computer receives the input. Based on the media advertisement information, the number of media users targeted by the media is calculated from the number of media users, and the advertisement estimated number of users estimated to pay attention to the advertisement is calculated from the number of media users. A method for predicting media audience rating, which is characterized by: 14. An advertisement response prediction method for predicting the number of responses of a media user to an advertisement included in a medium using a computer, wherein the computer uses the medium when calculating the predicted response number. The final predicted number of responses after considering the factors of how many people among the number of people pay attention to the advertisement and the number of people who respond to the advertisement independently. An advertising response prediction method characterized by: