JP4702642B2 - Creative optimization server, creative optimization system, creative optimization method, and program - Google Patents

Description

  The present invention relates to a creative optimization server and a creative optimization system for optimizing a hyperlink creative on a website such as an advertisement on a computer network (telecommunications line) in which a hyperlink is attached to text, a still image, and a moving image. , A creative optimization method, and a program.

  Usually, hyperlinks and advertisements posted on websites on computer networks are created by website creators or advertisement creators and posted on websites to guide users. However, it is unclear whether the expected conversion (final result), click rate (CTR: Click Through Rate), etc. can be derived.

  Content such as advertisements created and produced by combining text, still images, videos, music, etc. are collectively referred to as creative. In order to obtain the expected conversion, click rate, etc. for the creative displayed on the web page, it is necessary to optimize the creative itself having a hyperlink (extract a more effective creative) and prompt the user's action.

  For example, in Patent Document 1, an advertisement is distributed to a client computer via a computer network, the effectiveness of the advertisement is automatically evaluated, and the feature of the advertisement is automatically changed based on the evaluation result. To optimize advertising activity on a computer network.

The feature of the invention described in Patent Document 1 is that the object is simply to optimize from a log of a click rate and a conversion rate. That is, the effectiveness of the advertisement is evaluated based on the results. For example, when evaluating the effectiveness of an advertisement with respect to conversion, the advertisement is evaluated by judging the number of users who actually purchased a product of the advertisement by looking at the advertisement displayed on the web page.
JP 2007-510773 A

  However, the method described in Patent Document 1 is effective when a large number of samples (impressions) can be taken, and can be used when the number of creatives exceeds 100 or when the posting period on a web page is short. Can not do it.

  Furthermore, the method described in Patent Document 1 cannot cope with a diversified network society. It is said that the timing of sales and the speed of sales are important in selling products. The speed of sales is to increase consumers' willingness to purchase by performing appropriate advertising activities in an appropriate period, and to ensure profits by conducting sales activities while consumers are highly motivated to purchase. For example, products that increase in sales or consumption in the spring are more efficient for the seller if they are sold more in the spring, which requires quick and appropriate advertising activities.

  When evaluating the effectiveness of an advertisement based on actual results as described in Patent Document 1, generally, a certain amount of impression (display / exposure) is required as a population in order to increase the reliability of the evaluation. . The greater the number of impressions, the higher the reliability of the evaluation result. However, it takes time to increase the number of impressions, and the appropriate time for selling the product to be sold is lost due to the advertisement.

  The present invention has been made in view of such a situation, and an object thereof is to accurately evaluate the effectiveness of a creative with a small number of exposures so that the creative is efficiently exposed.

The creative optimization server according to one aspect of the present invention includes an index calculation unit, a reliability calculation unit, an expected value calculation unit, a read probability calculation unit, and a creative output unit.
The index calculation unit calculates each evaluation index at the present time for a plurality of creatives in which the relationship between the evaluation index, which is information for evaluating the effectiveness of the creative, and the number of creatives has a normal distribution.
The reliability calculation unit, based on the exposure times to the appropriate creative web pages, and past statistical data corresponding to the number of times out said exposure stored in the database at the time of calculation of the evaluation index, the evaluation index Calculate reliability .
In addition, the expected value calculation unit calculates an expected value of the evaluation index in the future by multiplying the value of the evaluation index by the reliability .
In addition, the read probability calculation unit changes the exposure frequency, which is the number of times the creative is exposed to the web page within an arbitrary time , based on the expected value of the evaluation index calculated by the expected value calculation unit.
The creative output unit extracts a creative high exposure frequency from the plurality of creative, it causes the output to c Ebupeji comes a hyperlink to the creative.
Then, the reliability calculation unit updates the past statistical data associated with the number of exposures of the creative stored in the database based on the performance of the current evaluation index.

The creative optimization system according to one aspect of the present invention includes a database and a creative optimization server that optimizes the creative.
The database stores a plurality of creatives and past statistical data associated with the number of times the creative is exposed to a web page .
The creative optimization server is configured to include a creative optimization server including an index calculation unit, a reliability calculation unit, an expected value calculation unit, a read probability calculation unit, and a creative output unit.
The index calculation unit calculates each evaluation index at the present time for a plurality of creatives in which the relationship between the evaluation index, which is information for evaluating the effectiveness of the creative, and the number of creatives has a normal distribution.
Further, the reliability calculation unit, based on the number of exposures to the web page of the corresponding creative at the time of calculating the evaluation index, and the past statistical data corresponding to the number of exposures stored in the database, the evaluation index The reliability of is calculated.
In addition, the expected value calculation unit calculates an expected value of the evaluation index in the future by multiplying the value of the evaluation index by the reliability .
In addition, the read probability calculation unit changes the exposure frequency, which is the number of times the creative is exposed to the web page within an arbitrary time , based on the expected value of the evaluation index calculated by the expected value calculation unit.
The creative output unit extracts a creative high exposure frequency from the plurality of creative, it causes the output to c Ebupeji comes a hyperlink to the creative.
Then, the reliability calculation unit of the creative optimization server updates the past statistical data associated with the number of exposures of the creative stored in the database based on the performance of the current evaluation index.

Also, the creative optimization method and program according to one aspect of the present invention provides a normal distribution of a relationship between an evaluation index that is information for evaluating the effectiveness of a creative and the number of creatives by an index calculation unit included in the creative optimization server. For each of a plurality of creatives, the current evaluation index is calculated.
Further, the number of exposures to the web page of the corresponding creative at the time when the evaluation index is calculated by the reliability calculation unit included in the creative optimization server, and past statistical data corresponding to the number of exposures stored in the database Based on the above, the reliability of the evaluation index is calculated.
A step of calculating an expected value in the future of the evaluation index by multiplying the reliability of the value of the evaluation index by an expected value calculation unit included in the creative optimization server ;
Then, based on the expected value of the evaluation index calculated by the expected value calculating unit by the read probability calculating unit included in the creative optimization server , the exposure frequency that is the number of times the creative is exposed to the web page within an arbitrary time Step to change the
Thereafter, the creative output unit the creative optimization server comprising extracts creative high exposure frequency from the plurality of creative and step of outputting the U Ebupeji comes a hyperlink to the creative has,
Further, in the step of calculating the reliability of the evaluation index, the past statistical data associated with the number of exposures of the creative stored in the database is updated based on the actual performance of the current evaluation index.

  In one aspect of the present invention, for example, by combining a statistical method with an evaluation index such as a click rate and functionalizing reliability, a future value is predicted, and it is efficient and faster (even with a small number of exposures). Creative can be extracted.

  As described above, according to the present invention, the effectiveness of a creative can be accurately evaluated with a small number of exposures, and a highly effective creative can be efficiently exposed.

  Hereinafter, examples of embodiments of the present invention will be described with reference to the accompanying drawings.

  The embodiment described below is a specific example of a preferred embodiment for carrying out the present invention, and therefore various technically preferable limitations are given. However, the present invention is not limited to these embodiments unless otherwise specified in the following description of the embodiments. Therefore, for example, the materials used in the following description, the amounts used, the processing time, the processing order, and the numerical conditions of each parameter are only suitable examples, and the dimensions, shapes, and arrangements in the drawings used for the description The relationship and the like are also schematic showing an example of the embodiment.

  Hyperlinks of websites (web pages) published on computer networks and their creatives are one of the major gaps in value. In order to improve the transition rate of the hyperlink, that is, the arrival rate to the target web page, the necessity of spending cost on improving the performance (attractiveness) of the advertisement and improving the performance (attractiveness) of the website is extremely high.

  In the first place, Internet advertising has a very high media cost (the cost of placing an advertisement on the media) and the production cost is relatively low, so the value created by the creative is great. In one embodiment of the present invention, candidates for creatives and hyperlinks are prepared to improve performance, and the exposure of high-expected creatives is increased while trying them, and the affiliator (producer) is rewarded for the effect. It is configured to obtain. Note that a banner advertisement (hereinafter simply referred to as “advertisement”) will be described as an example of a creative. However, the present invention is not limited to this example, and can be applied to optimization of any creative.

FIG. 1 is a diagram showing the needs of each stakeholder (advertiser, agency, affiliator) in the advertising industry.
Advertiser 1 wants many clicks on the advertisement on the web page if it costs the same medium. In addition, for the agency 2 that is requested by the advertiser 1 to promote the product etc., the creative production cost of creating a hyperlink to the web page on which the product is displayed is thin and the fixed cost. I would like to outsource production as much as possible and take a margin. Therefore, the agency 2 requests the affiliate 2 to produce a creative. Affiliator 3 wants money (the value of creative production) efficiently, wants to be evaluated, and wants to be praised.

  One of the indexes when the advertiser 1 evaluates an advertisement is a click rate. The click rate is the rate at which an advertisement is clicked, and can be calculated by dividing the number of clicks by the number of impressions (the number of advertisements displayed / delivered).

FIG. 2 is a graph showing the transition of the click rate from 1 impression to 1 million impressions for a plurality of advertisements. The horizontal axis represents the number of impressions (numerical value × 1000), and the vertical axis represents the click rate. In the example shown in FIG. 2, the click rates at 1 million impressions are investigated for 0.063, 0.053, 0.061, and 0.042, respectively.
In general, the click rate is not stable at the beginning when the advertisement is exposed to the website, but at a certain stage, it enters a stable period and gradually decreases. In the example shown in FIG. 2, it is stable after 400,000 impressions (broken line part). Therefore, it can be inferred that there is an advantage at an early stage, rather than a large amount of impressions required to measure the effectiveness of a creative.

  FIG. 3 is an example of a histogram showing the relationship between the number of creatives and the click rate. The example shown in FIG. 3 is an investigation of the number of creatives at each stage from a click rate of 0.03 to a click rate of 0.07. The distribution of the click rate and the number of creatives was found to be almost a normal distribution as shown by curve 4 in FIG. 4 from the average click rate per set and the standard deviation. In this example, since the average click rate (average CTR) is 0.048% and the maximum click rate (maximum CTR) is 0.07%, if the click rate when there is one creative is 0.048% Then, the click rate can be improved by 45% or more by selecting the creative with the highest CTR (the broken line portion in FIG. 4).

  I mentioned earlier that the effectiveness of creatives can be seen at an early stage, but there are several reasons for this. Assuming that 1000 impressions are made for a set of 1000 creatives, the entire set has 1 million impressions. Therefore, it can be calculated that the 1000 creatives are a set of creatives having an average click rate a and a standard deviation b earlier than individual creatives. In that case, the possibility that the click rate is considered to be higher in the stage of the number of impressions is very likely to be included in the broken line portion of FIG.

FIG. 5 is a graph showing an example of a change in normal distribution based on the number of creatives.
If the click rate and the number of creatives are normally distributed, the larger the total number of creatives, the higher the possibility that creatives with higher effectiveness (effectiveness) will be found. In other words, if a good one is selected from more, CTR (click rate) and CVR (conversion rate) can be further improved. For example, in the example shown in FIG. 5, since the absolute number of creatives is larger in the curve 5 than in the normal distribution curve 4 having the same average CTR, the number of creatives having the highest CTR is also larger.

  From the above, the present applicant has come up with the idea of commercializing a mechanism that automatically selects and displays the best creative with the highest click rate from as many creatives as possible. For example, FIG. 6 shows an example of a creative posted on a certain web page. An image creative 12 is displayed on the right side of the web page screen 10 and a text creative 13 is displayed on the lower side. When the present invention is applied to the web page screen 10, for example, among the images in which hyperlinks to the target site are embedded in the first push column 11 of this week, the exposure frequency is high but the exposure value is high. Moreover, although the expected value is high in the text, the exposure frequency becomes high. As a result, it is expected that the user's action such as mouse click becomes active.

  By the way, as explained in the background section, the larger the number of impressions, the higher the reliability of the evaluation result. However, if it takes time to increase the number of impressions, the time that is appropriate for obtaining the final result may be missed. , Which can lead to inexperienced inexperienced creatives. Therefore, it is necessary to evaluate the effectiveness of the creative as early as possible. However, since the reliability of the creative evaluation result is proportional to the number of impressions, the reliability of the creative evaluation result at the stage where the number of impressions is small is a problem. In other words, how to ensure reliability is important when evaluating creatives when there are few impressions.

  Hereinafter, a method for improving the reliability of the evaluation result of the creative will be described by taking a click rate which is one of evaluation indexes as an example.

7 and 8 are graphs showing examples of transitions in the click rate of a banner advertisement with a click rate of 0.063%. The horizontal axis indicates the actually measured number of impressions (100 impression units), and the vertical axis indicates the click rate.
The curve 21 gradually settles down to a click rate of 0.07% from 1 impression to 3098000 impressions. However, immediately after the start of measurement, the fluctuation rate of the click rate is large because the number of samples (impressions) is small. I understand. For example, a click rate of 0.07% is recorded at 1000 impressions (see FIG. 7). In addition, a click rate of 0.07% is recorded at 350,000 impressions (see FIG. 8). However, some of those that record a click rate of 0.07% at 1000 impressions have the following click rate transition.

FIG. 9 is a graph showing an example of transition of the click rate of a banner advertisement having a plurality of click rates of 0.063%, where the horizontal axis indicates the number of impressions (100 impression units) and the vertical axis indicates the click rate.
The curves 21, 22, and 23 all record a click rate of 0.07% at the 1000th impression, but it can be seen that the subsequent fluctuation width is large. When a large number of creatives with a click rate of 0.07% are collected at 1000 impressions, they are represented by a normal distribution with an average CTRa% as shown by the curve 24 in FIG. 10, and the standard deviation b1 is relatively large and widely dispersed. I understand that.

FIG. 11 is a graph showing an example of the click rate transition of a plurality of creatives that record a click rate of 0.07% at the time of 350,000 impressions. The horizontal axis represents the number of impressions (100 impression units), and the vertical axis represents the click rate.
It can be seen that the curves 21, 26 and 27 shown in FIG. 11 do not have a large fluctuation width as compared with the case where the same click rate of 0.07% is recorded at the 1000th impression in FIG. When many creatives with a click rate of 0.07% at the time of 350,000 impressions are collected, they are represented by a normal distribution with an average CTRa% as shown by the curve 28 in FIG. 12, and the standard deviation b2 is relatively small and narrowly distributed. You can see that

  In this specification, the standard deviation of the normal distribution corresponds to the reliability of the evaluation index such as the click rate. In other words, as the number of impressions increases and the number of click rate samples increases, the standard deviation decreases and the reliability increases (the blur width decreases). When the click rate at a certain number of impressions N (natural number) is set to r, a large number of samples of N and r are recorded, and by accumulating data on how they change thereafter, the click rate of The number of samples that serve as the basis for calculating reliability will increase.

  Here, a creative evaluation method using the reliability of the click rate will be described.

  As an example, a banner advertisement A having a click rate of 50% at the second impression time and a banner advertisement B having a click rate of 50% at the second million impression time are assumed. Both click rates are 50%, but the values are quite different. The banner advertisement A has a very high possibility that the click rate will decrease from now on, but it can be said that the click rate of the banner advertisement B is already stable. In other words, it is difficult to judge which one is better in a simple comparison of click rates.

  Therefore, banner advertisement A and banner advertisement B are converted into a function by multiplying the reliability of the impression at that time from the statistics of the past click rate and indexed as “expected click rate (hereinafter referred to as“ expected CTR ”). Can be compared. Actually, it is not a simple multiplication but a function of CTR and reliability. In addition, since the reliability of the reliability varies depending on the impression, in the present invention, calculation is performed by multiplying the reliability by a constant that should be called reliability of reliability.

CTR Reliability Expected CTR
Banner ad A 50% × 0.01 = 0.5
Banner ad B 50% x 0.95 = 47.5

  In this way, it is possible to compare those having different number of impression samples. The same is true for other normal distribution metrics such as conversions. That is, CTR (click rate), CVR (conversion rate), etc. are merely past values or present values, but the future value can be expressed by bringing the concept of expected values into an index.

  In this specification, the reliability of the click rate of the creative is defined as follows. Although the click rate will be described as an example, this calculation method can be applied to other evaluation indexes represented by a normal distribution such as a conversion rate.

  First, r is a click rate based on actual measurement data, and this is called a sample click rate. If N (natural number) is the number of display times (impressions) and c is the number of clicks, then r = c / N.

  Further, P (x) is the probability that the click rate x occurs. When this click rate x is given, the conditional probability that the sample click rate r occurs is expressed as PN (r | x). N is the number of times of display, and the description of N may be omitted in the following description. In that case, it is labeled P (r | x). The click rate x is assumed to be a numerical value obtained when N is maximized.

  As described above, the probability distribution of P (x) is a normal distribution (Gaussian distribution) from experience. Therefore, P (x) can be defined as the following formula (1).

  Also, let P (r, x) be the probability that the sample click rate r and click rate x when the creative is displayed N times are the same, that is, the probability that they will be established simultaneously (simultaneous probability).

Here, in order to calculate the reliability of the sample click rate r, it is necessary to consider what happens to the conditional probability P (x | r) of the click rate x when the value of the sample click rate r has already been obtained. is there. When the click rate x is a continuous random variable, the average μ _r and the standard deviation σ _r of the click rate x under the condition of the sample click rate r are _expressed by the following equations (2) and (3).

The average μ _r is an average of the click rate x when the sample click rate r is given when the sample click rate r is given. It can be said that the standard deviation σ _r is the reliability of the sample click rate r at the time of the display number N.

  Here, Bayes' theorem is used to obtain the conditional probability P (x | r). The conditional probability P (x | r) is expressed by Equation (4) by Bayes' theorem.

  P (r) is a probability that the sample click rate r occurs when the sample click rate r and the click rate x are satisfied at the same time. As shown in the equation (5), P (x) P (r | x) is expressed as follows. It can be expressed as an integral.

Since P (x) is a known number obtained from past statistics or empirical rules, P (r | x) can be calculated from Equation (6) including a binomial coefficient. That is, if the creative is displayed N times and is considered to be displayed with the probability of x, the probability is obtained by the binomial distribution. At this time, μ _r = r.

Assuming that P (x) is empirically known, the probability of P (r | x) is obtained by binomial distribution. Therefore, P (x) in Equation (5) is obtained, and finally P (x | r) is obtained from Bayes' theorem shown in Equation (4). Then, a solution of the standard deviation σ _r of Equation (3) is obtained, and reliability, that is, reliability is defined by the solution.

  As described above, reliability (standard deviation) is calculated from N (number of impressions) and r (CTR: click rate) when a certain creative is exposed, and what click rate is recorded by the creative after that Can be used to calculate the expected CTR.

Here, the relationship between the normal distribution and the standard deviation will be described with reference to the graph shown in FIG.
As shown in FIG. 13, the curve 29 of the normal distribution falls within the standard deviation, and the ratio of the data that falls within the standard deviation can be calculated as follows.
(Range) (Percentage of data included in it)
μ ± σ 0.6827 (about 2/3)
μ ± 2σ 0.9545 (about 19/20)
μ ± 3σ 0.9973
* This means that 68.27% of all data is distributed within the range of the average value ± standard deviation, and 95.45% of all data is distributed within the range of the average value ± (twice the standard deviation).

  By using the above relationship, it is possible to quickly determine the superiority of the click rate within a certain probability (within risk). That is, with this calculation method, it is possible to determine at an early stage which creative has a high advantage in the click rate. Based on these sample data, as the measured impression N and click rate r data are accumulated, the width of the click rate r decreases.

  Next, a creative optimization system using the above-described creative evaluation method will be described.

FIG. 14 is a diagram showing an overview of a creative optimization system according to an embodiment of the present invention.
The creative optimization system 30 of the present embodiment mainly includes a submission / examination interface device 32 for preparing a plurality of creative candidates, a creative optimization server 33 for optimizing the creative, and a web for transmitting information such as advertisements. The server 35 and the client terminal 36 are included.

  The new part of the business model that uses this creative optimization system 30 is that the creation of a large number of creatives itself is not suitable for economic rationality. It is a point that can be collected. Also, it ’s not about people judging these creatives, but trying them all. In addition, the most appropriate one is extracted from the creative in a short period of time. Furthermore, the effect is obtained by introducing the effect into a list page such as an advertisement or a search result of a web page with a high value gap (a value can be improved by induction).

  Further, the new part of the creative optimization system 30 is the one that extracts the one with the highest evaluation index such as CTR and CVR at the fastest speed. That is, instead of simply comparing the current CVR and CTR, calculating the expected value of the creative by combining statistical methods makes it possible to determine an accurate evaluation index even from a small number of samples. Yes.

  Next, processing of the entire creative optimization system 30 will be described with reference to a flowchart shown in FIG. Here, a case where a creative is published on a network for the purpose of guiding a user to a certain website will be described.

  First, in step S 1, a large number of creative candidates that are considered to meet the purpose of a certain hyperlink destination are prepared in the submission / examination interface device 2. In this case, the guidance destination and the intention of guidance are publicized on the network, and a third party such as the users 31a to 31c is allowed to produce a large number of creatives to collect candidates.

  For example, it can be expected that the thread format of the many-to-many community as shown in FIG. Affiliators who contribute to the campaign, both clients and operators, will build a community where participants can participate freely as participants 31a to 31f. In that, a thread is set up for each advertisement (purpose) and posted.

  As for the screening, I think that if you go open as well as the submission, you will be able to provide an exciting place. For example, as shown in FIG. 17, a large number of users post creatives on a posting site and create threads one after another as threads 41, 42. The client sets up his own thread 43 and publishes the examination contents while writing a comment 43a on the posted creative.

  In step S2, for example, the creatives A, B, and C prepared in the submission / examination interface device 32 are examined for appropriateness, and those that pass the examination are stored in a creative database to be described later. In this review, the emphasis is placed on formal reviews, such as the existence of copyright infringement, rather than the aesthetic appearance of the creative itself.

  In step S3, a program 34 including access to the creative optimization server 33 is distributed from the web server 35 that distributes content on the network to the client terminal device 36. For example, the client terminal device 36 connects to the creative optimization server 33 through the network, triggered by the user starting up a browser on the client terminal device 36 or accessing a predetermined homepage. As an example of the program 34, for example, a flash banner in which a script is described is submitted.

  In step S4, the creative optimization server 33 extracts all the creatives from the creative database little by little or with a certain probability, and distributes the creative with the hyperlink to the client terminal device 36 through the network.

  In step S5, the distributed creative with hyperlink is displayed on the client terminal device 36 (display device connected thereto), and the creative optimization server 33 responds to the response of the client terminal device 36 (user) to the creative with hyperlink. Get through the network.

  In step S <b> 6, the creative optimization server 33 calculates the expected value of the creative by adding reliability to the information acquired from the client terminal device 36. This is because the fastest and the highest evaluation index (CTR, CVR, etc.) is extracted.

  In step S7, the exposure probability of the creative when the creative optimization server 33 is accessed next is changed based on the expected value. That is, the exposure of the extracted creative with a high expected value is increased. In this way, many exposures with high expectation values are displayed with an inclination to the exposure probability.

  In step S8, the processes in steps S1 to S7 (particularly steps S4 to S7) are repeated, and the exposure probability is updated as needed for each creative in the creative database. At this time, it is confirmed whether or not those creatives with high expectation values have the expected CTR and CVR. If they are in line with expectations, keep them exposed with a tilt. On the other hand, if the expectations are not met or if other creatives with a higher expected value are found, the exposure of creatives with higher expected values is increased (changes the slope distribution of exposure). By repeating the above process, CTR and CVR are optimized.

  Note that “being in line with expectations” refers to whether or not the product falls within the range of risks specified in the reliability section. That is, whether the click rate falls within the ± σ range (68.27%) of the standard deviation, the click rate within the ± 2σ range (95.45%), or the ± 3σ range (99.73%). It is determined that there is no problem as long as the click rate falls within the range set in advance. The specific range is actually adjusted by overlapping the number of cases, and it is preferable that the amount of risk (that is, error) is appropriately changed through repeated verification.

  In step S9, a reward or incentive for the creator (affiliator) is set according to the performance of the creative (user reaction). After this process is finished, the process of the entire system is finished. In an actual system, a series of processes of step S1 to step S9 or step S3 to step S9 are repeated.

  When extracting creatives that would be good from a large number, it would be very inefficient to calculate from each expected CTR. Therefore, instead of comparing individual creatives, the exposure is maximized by narrowing down from the set to “a small set that will have the highest click rate among them”, and clicking It may be possible to improve the system performance (performance / efficiency) by going through the steps (steps S6 and S7) of verifying and narrowing down the rate and conversion rate.

  In addition, a check stage (steps S5 to S8) is prepared in order to pick up other good creatives mixed in with a large number of creatives. If a creative that does not meet expectations or has a high expected value is found, Increase the exposure of creatives with high expectations (change the exposure slope distribution). At that time, creatives with different degrees of reliability are compared using the concept of expected values. These statistical data can be learned and grown to a more reliable system by repeating correction, and optimum performance (performance / efficiency) can be extracted.

  Next, with reference to FIG. 18, the structure of the creative optimization system 30 in one embodiment of the present invention will be described in detail.

  The creative optimization system 30 includes a submission / examination interface device 32, a creative optimization server 33, a creative database 33A, a web server 35A, an advertisement distribution server 35B, and client terminal devices 36A to 36C. Are connected to each other. In particular, although the submission / examination interface device 32 and the creative optimization server 33 are prepared for this system, the submission / examination interface device 32 may be provided as necessary. The web server 35A, the advertisement distribution server 35B, and the client terminals 36A to 36C use existing devices.

  Note that a wide area network such as the Internet can be applied to the computer networks 50 and 50a. The computer network 50a may be a local area network (LAN). In the present embodiment, transmission / reception of data between the devices is performed via the computer networks 50 and 50a, but in the following description, a description that the data is transmitted via the computer networks 50 and 50a is omitted. In the present embodiment, transmission / reception of data on the computer networks 50 and 50a is performed using TCP / IP (Transmission Control Protocol / Internet Protocol) protocol. However, the present invention is limited to this example. is not. Furthermore, in the following description, the client terminal devices 36A to 36C are collectively referred to as the client terminal device 36.

  The submission / examination interface device 32 is an example of an interface device described in claims, and executes processing related to creative submission and examination.

  The creative optimization server 33 calculates an expected value of a plurality of creatives stored in the creative database 33A by using a statistical method at an early stage with few impressions, and extracts a creative with a high expected value. Then, the extracted creative with a high expected value, that is, the optimized creative is transmitted to the web server 35A, the advertisement distribution server 35B, and the like via the computer network 50 according to the purpose. Thus, the overall system can be optimized by increasing the exposure of creatives with higher expected values.

  The creative database 33A is a non-volatile storage means such as a large-capacity magnetic recording device or a semiconductor memory that can store and store a large amount of creatives. The submission / examination interface device 32 and the creative database 33A are connected via the computer network 50a, but the creative database 33A may be provided in the submission / examination interface device 32. Alternatively, the creative database 33A may be stored in the creative optimization server 33 and configured integrally.

  The web server 35 </ b> A transmits data such as a web page in response to a request from the client terminal device 36 on the computer network 50.

  The advertisement distribution server 35B distributes advertisements displayed on a web page or the like transmitted to the client terminal device 36 on the computer network 50. The advertisement distribution server 35B is also a web server in a broad sense in that it is connected to a computer network and transmits web content.

  As the client terminal devices 36A to 36C, in addition to a general-purpose computer such as a personal computer, a device having a communication function with the computer network 50 such as a mobile phone or a game machine is applied.

Here, the submission / examination interface device 32 will be described in detail.
FIG. 19 is a block diagram showing an example of the internal configuration of the submission / examination interface device 32. The submission / examination interface device 32 mainly includes a submission / examination discrimination unit 51, an input processing unit 52, and an examination processing unit 53.

  The submission / examination discriminating unit 51 discriminates whether to execute the submission process or the examination process when accessed from the client terminal device 36.

  The submission processing unit 52 performs a predetermined submission process when the submission / examination discrimination unit 51 receives an access request for submission processing from the client terminal device 36. The submission processing unit 52 is configured to include a login processing unit 52A, a guidance purpose selection unit 52B, a guidance destination / guidance purpose display screen generation unit 52C, and a creative submission unit 52D.

  The login processing unit 52A performs login processing (authentication processing) of the client terminal device 36 that has accessed the submission / examination interface device 32 for submission. For example, a login screen requesting input of an identification code (ID) and password is displayed on the client terminal devices 36A to 36C, and the input identification code and password are registered in the submission / examination interface device 32. Compare with things to see if you are a legitimate user. This login screen will be described later.

  The guidance purpose selection unit 52B displays a guidance purpose selection screen that prompts the authenticated client terminal device 36 to select a guidance purpose (such as a campaign or a site), and determines which guidance purpose has been selected. This guidance purpose selection screen will be described later.

  The guidance destination / guidance purpose display screen generation unit 52C generates a guidance destination / guidance purpose display screen, which will be described later, based on the determination result of the guidance purpose selection unit 52B.

  The creative submission unit 52D receives the creative uploaded from the client terminal device 36 through the guidance destination / guidance purpose display screen, and temporarily stores it in the creative database 33A in association with predetermined metadata.

  The examination processing unit 53 performs predetermined examination processing when the submission / examination discrimination unit 51 receives an access request for examination processing from the client terminal device 36. The examination processing unit 53 is configured to include a login processing unit 53A, a guidance purpose selection unit 53B, an inappropriate creative selection unit 53C, and an inappropriate creative deletion unit 53D.

  Similar to the login processing unit 52A, the login processing unit 53A performs login processing (authentication processing) of the client terminal devices 36A to 36C that have accessed the submission / examination interface device 32 for examination. For example, a login screen requesting input of an identification code (ID) and a password is displayed on the client terminal devices 36A to 36C, and the input identification code and password are registered in the submission / examination interface device 32. Compare to see if you are a legitimate user. The login screen may have the same appearance as that displayed by the login processing unit 52A.

  The guidance purpose selection unit 53B displays a guidance purpose selection screen that prompts the authenticated client terminal device 36 to select a guidance purpose (such as a campaign or a site), and determines which guidance purpose has been selected. The guidance purpose selection screen may have the same appearance as that displayed by the login processing unit 53A.

  The inappropriate creative selection unit 53C generates a screen showing the submitted creative list based on the determination result of the guidance purpose selection unit 53B, and displays the creative list screen on the client terminal device 36. Then, the user is prompted to select an inappropriate creative, and it is determined which creative has been selected as inappropriate. This creative list screen will be described later.

  The inappropriate creative deletion unit 53D performs a predetermined process on the creative temporarily stored in the creative database 33A based on the determination result of the inappropriate creative selection unit 53C. That is, an appropriate creative is stored in the creative database 33A, and an inappropriate creative is deleted from the creative database.

  Although the login processing units 52A and 53A and the guidance purpose selection units 52B and 53B are provided in the submission processing unit 52 and the screening processing unit 53, only one each is provided in the submission / examination interface device 32. It may be shared between the drafting process and the screening process.

Next, data stored in the creative database 33A will be described.
The creative database 33A stores a creative table that mainly manages each creative and a case table that manages cases for each guidance purpose (campaign, site, etc.).

FIG. 20 is a diagram illustrating an example of a creative table.
The creative table 61 (creative DB) identifies and manages each creative by creative ID, and creative information and producer information are registered corresponding to one creative ID.

  Creative information includes, for example, creative data (text, still image, video, music, etc.), destination data (link destination URL (Uniform Resource Locator), etc.), impressions, clicks, conversions, CTR (clicks), CVR (conversion rate), reliability, expected CTR, expected CVR, and the like are included. The producer information includes at least a producer ID for identifying the producer.

FIG. 21 is a diagram illustrating an example of a case table.
The case table 62 (case DB) identifies and manages each case by the case ID, and basic information and creative information are registered corresponding to one case ID.

  The basic information includes client data, guidance destination data (URL (Uniform Resource Locator) of link destination), deadline date, purpose of the case, and the like. The creative information includes a plurality of creative IDs (those that have been submitted and have been reviewed).

Next, the creative optimization server 33 will be described in detail.
FIG. 22 is a block diagram illustrating an internal configuration example of the creative optimization server 33. The creative optimization server 33 includes a creative input / output unit 71, an event determination unit 72, a log recording unit 73, an index calculation unit 74, a reliability / expected value calculation unit 75, a read probability calculation unit 76, It is comprised so that the calculating part 77 may be included.

  The creative input / output unit 71 calls the link destination information for creative and guidance purposes from candidates in the creative database 33A with a certain probability. Specifically, for example, a creative with a high expected value is read from the creative database 33A according to the probability calculated by the read probability calculation unit 76. The creative input / output unit 71 outputs the creative with the read probability changed from the read probability calculating unit 76 to the web server 35A, the advertisement distribution server 35B, and the like through the computer network 50. The creative input / output unit 71 is an example of a creative output unit described in the claims.

  The event determination unit 72 determines the content of information (events) such as the processing contents of the client terminal device 36, for example, the creative impression, which creative was clicked, or which creative is the action for which creative. It is.

  The log recording unit 73 records the log of the determination result of the event determination unit 72 in a non-illustrated non-volatile memory that is a recording unit.

  The index calculation unit 74 calculates a creative evaluation index represented by a normal distribution, such as a click rate or a conversion rate, based on the log recorded by the log recording unit 73.

The reliability / expected value calculation unit 75 is an example of a reliability calculation unit or an expected value calculation unit described in the claims, and based on the number of impressions at the time of calculating the creative evaluation index, Reliability is calculated, and the expected value is calculated from the evaluation index and reliability. Calculations for reliability and expected values are performed using the above-described equations (3) to (6).

  The read probability calculator 76 changes the creative read probability (exposure frequency) based on the expected value calculated by the reliability / expected value calculator 75. For example, when the expected value of a certain creative is high as a result of calculation, the slope of the probability of exposure is changed so that the exposure of the creative increases. That is, the ratio of the expected value in a certain creative to the sum of the expected values of the entire creative is calculated, and the exposure frequency is determined based on the ratio. Therefore, creatives with a high expectation value account for a higher proportion of the total, and the exposure frequency increases. On the other hand, creatives with low expected values have a lower percentage of the total and exposure frequency decreases.

  The consideration calculation unit 77 calculates a consideration (points, etc.) for creative production based on the creative evaluation index calculated by the index calculation unit 74 and the expected value calculated by the reliability / expectation value calculation unit 75. is there.

  In the creative optimization server 33 according to the present embodiment, the function of transmitting the URL information of the creative and the link destination is not specified, but the creative transmission unit that transmits data to the outside is naturally provided due to the nature of the server. In preparation. In addition, each processing unit in the creative optimization server 33 is configured in the creative optimization server 33 that is one server, but may be stored in a separate server instead of being integrated.

Next, the internal configuration of the client terminal devices 36A to 36C will be described.
FIG. 23 is a block diagram illustrating a configuration example when the client terminal devices 36A to 36C are general-purpose computers such as personal computers.

  A CPU (Central Processing Unit) 81 of the computer 80 executes various processes in addition to a series of processes according to a program recorded in a ROM (Read Only Memory) 82 or a recording unit 88. A RAM (Random Access Memory) 83 appropriately stores programs executed by the CPU 81 and data. The CPU 81, ROM 82, and RAM 83 are connected to each other by a bus 84.

  An input / output interface 85 is also connected to the CPU 81 via the bus 84. Connected to the input / output interface 85 are an input unit 86 made up of a keyboard, mouse, microphone, and the like, and an output unit 87 made up of a display, a speaker, and the like. The CPU 81 executes various processes in response to commands input from the input unit 86. Then, the CPU 81 outputs the processing result to the output unit 87.

  The recording unit 88 connected to the input / output interface 85 includes, for example, a hard disk, and records programs executed by the CPU 81 and various data.

  The communication unit 89 communicates with an external device via a network such as the Internet or a local area network. A program may be acquired via the communication unit 89 and recorded in the recording unit 88.

  The drive 90 connected to the input / output interface 85 drives a removable medium 91 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, and drives the programs and data recorded therein. Get etc. The acquired program and data are transferred to the recording unit 88 and recorded as necessary.

  A program recording medium that stores a program that is installed in a computer and is ready to be executed by the computer is provided as a package medium by a removable medium 91 as shown in FIG. As the removable medium 91, a magnetic disk (including a flexible disk), an optical disk (including a CD-ROM (Compact Disc-Read Only Memory), a DVD (Digital Versatile Disc), a magneto-optical disk), or a semiconductor memory is applied. be able to. Alternatively, the program recording medium includes a ROM 82 in which the program is stored (recorded) temporarily or permanently, a hard disk that constitutes the recording unit 88, and the like.

  The program can be stored in the program recording medium via a communication unit 89, which is an interface such as a router or a modem, as necessary, such as a local area network (LAN), the Internet, or digital satellite broadcasting. This is performed using a wireless communication medium.

  In addition, although the case where the client apparatus 36 is comprised using the general purpose computer 80 was demonstrated, the submission / examination interface apparatus 32 and the creative optimization server 33 can also be comprised using this general purpose computer 80. FIG.

  Next, the operation of the creative optimization system 30 configured as described above will be described. First, the submission process by the submission / examination interface device 32 will be described with reference to the flowchart of FIG.

  As a premise, when the user attempts to log in to the creative examination system from the client terminal device 36, the following submission process is started.

  First, in step S11, when there is an access from the client terminal device 36 via the computer network 50, the submission / examination discrimination unit 51 of the submission / examination interface device 32 determines whether or not it is a request for submission processing. To do. In the case of the submission process, the login processing unit 52A of the submission processing unit 52 performs the login process of the client terminal device 36 (user). After this process is completed, the process proceeds to step S12.

  FIG. 25 is a diagram illustrating an example of a login screen displayed on the client terminal device 36 during the login process. In this embodiment, the login screen 101 requests the input of the user ID and password. The example of the login screen displayed on the client terminal device 36 during the login process is not limited to this.

  In step S12, the guidance purpose selection unit 52B of the submission / examination interface device 32 causes the authenticated client terminal device 36 to display a guidance purpose selection screen that prompts the user to select a guidance purpose (campaign, site, etc.). After this process is completed, the process proceeds to step S13.

  FIG. 26 is a diagram illustrating an example of a guidance purpose selection screen displayed on the client terminal device 36. On the guidance purpose selection screen 102, for example, the campaign name, deadline date, and URL of the guidance destination are displayed. In this embodiment, since the cursor 102a is on the advertisement A, the advertisement A is selected. The example of the guidance purpose selection screen is not limited to this.

  In step S <b> 13, the guidance purpose selection unit 52 </ b> B determines which guidance purpose (for example, campaign name) is selected from the response content of the client terminal device 36. After this process is completed, the process proceeds to step S14.

  In step S14, the guidance destination / guidance purpose display screen generating unit 52C displays the guidance purpose and posting form display screen 103 as shown in FIG. 27 as the guidance destination / guidance purpose display screen, and further the guidance purpose as shown in FIG. And the posted content confirmation display screen 105 are displayed on the client terminal device 36. After this process is completed, the process proceeds to step S15. The contents of the display screens 103 and 105 shown in FIGS. 27 and 28 will be described later.

  In step S15, when the user completes the creative submission process from the client terminal device 36 through the display screens 103, 105, etc., the submission / examination interface device 32 receives the uploaded creative. After this process is completed, the process proceeds to step S16.

  In step S16, the creative submission unit 52D of the submission / review interface device 32 temporarily stores the received creative in the creative database 33A. After this processing is completed, a series of submission processing is completed.

The guidance purpose and posting form display screen 103 shown in FIG. 27 will be described.
The guidance purpose and posting form display screen 103 displays, for example, a guidance purpose column 103a including a campaign name, deadline date, and guidance destination URL information, a comment field 103b for submission, a posting form 104, and the like. The posting form 104 includes a guidance destination selection column 104a (when there are a plurality of guidance destinations), a title column 104b of a submitted creative, a body text column 104c, a function key group 104d for selecting the type of an image file, a creative display column 104e, It has a photo file storage location display field 104f, an icon-displayed input content confirmation button 104g, and the like. When the user has posted the creative, the user clicks on the input content confirmation button 104g using a pointing device such as a mouse (not shown), and transitions to the posted content confirmation screen display screen 105 (see FIG. 28).

The guidance purpose and posted content confirmation display screen 105 shown in FIG. 28 will be described. In FIG. 28, portions corresponding to those in FIG. 27 are not described in order to avoid duplication.
The guidance purpose and posted content confirmation display screen 105 displays a guidance purpose column 103a, a comment column 103b, a posting form 106, and the like. The posting form 106 includes a guide destination selection field 106a, a title confirmation field 106b, a text confirmation field 106c, a creative confirmation field 106d, a compression confirmation field 106e, a creation button 106f, a return button 106g, and the like. In the creative confirmation column 106d of this example, three creatives 106d1 to 106d3 and a path indicating the position on the file system where each creative is recorded are displayed.

  If there is no problem with the submitted creative, the user clicks the creation button 106f using a pointing device such as a mouse (not shown), and the operation is completed. When the creative is to be reposted, the user clicks the back button 106g, and returns to the guidance screen and post form display screen 103, which is the previous screen, to re-edit.

FIG. 29 is a diagram illustrating an example in which a creative that has already been submitted is displayed under the submission form 104.
When displayed in this way, the user can view creatives posted by other affiliators. In the present embodiment, a posting screen 107 including a creative 107 a and a title 107 b posted from the user 107 c is displayed under the posting form 104. Further, a posting screen 108 including a creative 108a and a title 108b posted from another user 108c in response to the posting screen 107 is displayed. Thus, by displaying the posting screen in a time series with a title, it is possible to grasp the flow of posting in a time series and to post while communicating with other users.

  Next, screening processing by the submission / screening interface device 32 will be described with reference to the flowchart of FIG.

  As a premise, the following examination process is started when the user displays a login screen as shown in FIG. 25 on the client terminal device 36 and attempts to log in to the creative examination system.

  First, in step S21, when there is an access from the client terminal device 36 via the computer network 50, the submission / examination discriminating unit 51 of the submission / examination interface device 32 judges whether or not it is a request for examination processing. . If it is a screening process, the login processing unit 53A of the screening processing unit 53 performs a login process for the client terminal device 36 (user). After this process is completed, the process proceeds to step S22.

  In step S22, the guidance purpose selection unit 53B of the submission / examination interface device 32 displays a guidance purpose selection screen (see, for example, FIG. 26) that prompts the authenticated client terminal device 36 to select a guidance purpose (campaign, site, etc.). Let After this process is completed, the process proceeds to step S23.

  In step S23, the guidance purpose selection unit 53B determines which guidance purpose (for example, campaign name) has been selected from the response content of the client terminal device 36. After this process is completed, the process proceeds to step S24.

  In step S24, the inappropriate creative selection unit 53C generates a screen showing the submitted creative list based on the determination result of the guidance purpose selection unit 53B, and causes the client terminal device 36 to display the creative list screen. After this process ends, the process proceeds to step S25.

FIG. 31 is a diagram illustrating an example of an examination screen.
The screening screen 110 includes, for example, a guidance purpose column 110a including a campaign name, a deadline date, and URL information of a guide destination, a comment column 110b for submission, an instruction column 110c for prompting a check for inappropriate creatives, a guide destination column 110d, a creative It includes a display area 110e and a confirmation end button 110g. In the creative display area 110e, a check box is displayed below each creative, and for example, an inappropriate creative such as a copyright violation or a violation of public order and morals can be checked. Here, the purpose is to collect the number of creatives, so we do not actively check the aesthetic appearance and appeal of the creatives. In this embodiment, the check boxes 110f1 and 110f2 are checked, indicating that these two creatives are to be erased.

  Returning to the flowchart of FIG. In step S25, the inappropriate creative selection unit 53C determines which creative has been selected as inappropriate on the examination screen 110 shown in FIG. After this process is completed, the processes of steps S26 and S27 are executed. The inappropriate creative deleting unit 53D performs the following processing on the creative temporarily stored in the creative database 33A based on the determination result of the inappropriate creative selecting unit 53C.

  In step S26, a creative without a check mark in the check box is stored in the creative database 33A as an appropriate creative.

  Also, in step S27, the creative that is not appropriate because the check box is checked is deleted from the creative database 33A.

  After the processes in steps S26 and S27 are completed, the series of examination processes is terminated.

  Expected values are calculated for a large number of creatives collected as described above, and a creative optimization process is performed so that the creatives with higher expected values are preferentially exposed on the network. The creative optimization processing in the present embodiment will be described with reference to the sequence diagram of FIG.

  First, as a premise, a program 34 for executing access processing to the creative optimization server 33 is distributed to the client terminal device 36 (see FIG. 14).

  Therefore, in step S31, the client terminal device 36 starts the program 34 that calls the creative optimization server 33 based on a predetermined trigger such as starting a browser. After this process ends, the process proceeds to step S32. In the present embodiment, the client terminal device 36 automatically calls the creative optimization server 33 by some trigger, but is not limited to this example. For example, various forms are conceivable, such as calling the creative optimization server 33 in accordance with a user operation.

  In step S <b> 32, the client terminal device 36 transmits a call command to the creative optimization server 33. After this process is completed, the process proceeds to step S33.

In step S33, the creative input / output unit 71 of the creative optimization server 33 refers to the creative table 61 of the creative database 33, and information on the creative and the link destination (hyperlink) with a certain probability among a plurality of candidates. To extract. After this process is completed, the process proceeds to steps S34 and S35.

  In step S <b> 34, the creative transmission unit of the creative optimization server 33 transmits the hyperlinked creative to the client terminal device 36. After this process is completed, the process proceeds to step S37.

  On the other hand, in step S35, after the process of step S33 ends, the event determination unit 72 of the creative optimization server 33 detects that a creative impression has occurred. In addition, the log recording unit 73 leaves the impression of the corresponding creative in the log. Then, the index calculation unit 74 and the reliability / expected value calculation unit 75 operate to calculate the reliability and expected value of the creative at that time and update the information in the creative table 61 (see FIG. 20). . After this process is completed, the process proceeds to step S36.

  In step S36, the read probability calculation unit 76 of the creative optimization server 33 changes the probability of calling the corresponding creative based on the expected value.

  In step S37, after the processing of step S34 is completed, the client terminal device 36 displays the hyperlinked creative sent from the creative terminal device 33 on the display screen. After this process ends, the process proceeds to step S38.

  In step S <b> 38, the client terminal device 36 monitors whether or not a hyperlink embedded in the creative is clicked. If the hyperlink is clicked, the process proceeds to step S39. On the other hand, if the hyperlink is not clicked, the process proceeds to step S43.

  In step S <b> 39, the client terminal device 36 transmits click information to the creative optimization server 33. After this process is completed, the process proceeds to step S40.

  In step S40, the event determination unit 72 of the creative optimization server 33 detects that a hyperlink embedded in the creative is clicked. After this process is completed, the process proceeds to step S41.

  In step S41, the log recording unit 73 of the creative optimization server 33 leaves a click history of the corresponding creative in the log. Then, the index calculation unit 74 and the reliability / expected value calculation unit 75 operate to calculate the reliability and expected value of the creative at that time, and update the information in the creative table 61. After this process is completed, the process proceeds to step S42.

  In step S42, the read probability calculation unit 76 of the creative optimization server 33 changes the probability of calling the corresponding creative based on the expected value.

  In step S43, when the hyperlink embedded in the creative is clicked, the client terminal 36 displays the web content of the guidance destination indicated by the hyperlink on the display screen. After this process ends, the process proceeds to step S44.

  In step S <b> 44, the client terminal device 36 monitors whether any action has been performed on the web content of the guidance destination. If there is an action, the process proceeds to step S45. For example, if the corresponding creative is an advertisement, the action here refers to the user ’s creative that leads to the final outcome (conversion) expected by the advertiser, such as the purchase of a product listed in the advertisement or a catalog request. Refers to the action.

  In step S <b> 45, the client terminal device 36 transmits action information to the creative optimization server 33. After this process is completed, the process proceeds to step S46.

  In step S46, the event determination unit 72 of the creative optimization server 33 detects that the action has been performed on the destination of the creative. After this process is completed, the process proceeds to step S47.

  In step S47, the log recording unit 73 of the creative optimization server 33 leaves the action history of the corresponding creative in the log. Then, the index calculation unit 74 and the reliability / expected value calculation unit 75 operate to calculate the reliability and expected value of the creative at that time, and update the information in the creative table 61. After this process is completed, the process proceeds to step S48.

  In step S48, the read probability calculation unit 76 of the creative optimization server 33 changes the probability of calling the corresponding creative based on the expected value.

  Next, the result display and point distribution of the submitted creative will be described with specific examples.

FIG. 33 is a diagram illustrating an example of a result display screen of a submitted creative.
For example, the display of the result starts when the advertiser closes the application for creative submission. In the present embodiment, when the comment screen 121 in which the advertiser has closed the creative application and declared the start of posting is displayed, the creative ranking screen 122 showing the result is displayed almost simultaneously. On the creative ranking screen 122, items such as a submitted creative 122a, an affiliator 122b, an impression number (Imp number) 122c, a click number 122d, and a click rate (CTR) 122e are displayed.

FIG. 34 is a diagram illustrating an example in which points are distributed according to an expected value for each creative illustrated in FIG. 33.
On the creative ranking screen 122A, a comment screen 121A in which the advertiser makes a comment on the result is displayed. In addition to the items shown in FIG. 33, an expected CTR 122f and acquired points 122g are displayed on the creative ranking screen 122A. In this embodiment, an acquisition point 122g is determined based on the number of click rates 122, and bonuses corresponding to the ranks are given to the top three affiliates (A, B, C). Yes. Acquired points are not limited to the number of clicks, but may be a click rate, an expected CTR, or may be determined based on other numerical values. Furthermore, you may make it calculate an acquisition point combining them.

FIG. 35 is a diagram showing an example of improving the relationship between the stakeholders after applying the present invention.
By using the new advertising system (creative optimization system) of the present invention, the advertiser 1 is provided with the actual effect of increasing the click rate of the creative, that is, the actual effect of optimization from many creatives. Can do. For the advertiser 1, matching with the affiliator can be obtained by the convincing feeling and the effect of automatically selecting a good one out of a large number of creatives by highly motivated creators. In addition, it is good also as a mechanism in which a good creative is bought by the advertiser 1.

  Further, the agency 2 can be provided with a profit that reduces the sales margin as a medium cost by operating a new advertising system (creative optimization system). In other words, since a general user can become an affiliate and collect a large number of high-quality creatives, production costs can be reduced.

  In addition, the affiliator 3 can be provided with a satisfactory distribution (reward) and motivation by the community. The Affiliator 3 is seeking not only money but also evaluation from the world, so that the position of the Affiliator 3 can be improved.

According to the present embodiment described above, the future value can be predicted by functionalizing the reliability of the click rate and the conversion rate, and a better creative can be extracted efficiently and quickly.
Also, instead of selecting the creative itself, it is possible to pick up a good creative by tilting the exposure probability. In addition, it is based on the reality by adopting the concept of reliability, and in this respect, it can be said that the effectiveness is extremely high as compared with that described in Patent Document 1.
Unlike the invention described in Patent Document 1, the present invention is different not only in advertising but also in optimizing hyperlinks. Further, although the click rate and the conversion rate are used, the click rate and the conversion rate itself are not evaluated.

From a business model perspective, it is possible to collect more creatives by gathering openly, including externally, where many creative creations themselves do not match economic rationality. .
In addition, it is possible to increase the possibility of extracting good creatives by not exposing these creatives but by exposing them all for the time being.
In addition, the optimal one can be extracted from the creative with a small number of exposures and a short period of time.
Also, by introducing the effect to advertisements with high value gaps, search result pages, and the like, the affiliator can obtain a better reward.

  Furthermore, the fastest and the highest click rate or the high conversion rate can be extracted. Rather than simply comparing the click rate and conversion rate at the present time, it is possible to calculate the creative value from a small number of samples (number of exposures) by calculating the expected value of the creative by combining statistical methods Become.

  The series of processes described above can be executed by hardware, or can be executed by software. Needless to say, the function for executing these processes can also be realized by a combination of hardware and software. When a series of processing is executed by software, it is possible to execute various functions by installing a computer in which a program constituting the software is incorporated in dedicated hardware, or by installing various programs. For example, it is installed from a program recording medium in a general-purpose computer or the like.

  Further, in the present specification, the processing steps describing the program stored in the program recording medium are not limited to the processing performed in time series in the described order, but are not necessarily performed in time series. This includes processing that is executed manually or individually (for example, parallel processing or object processing).

  Further, the program may be processed by a single computer, or may be processed in a distributed manner by a plurality of computers. Furthermore, the program may be transferred to a remote computer and executed.

  Furthermore, the embodiments of the present invention are not limited to the above-described embodiments, and various modifications can be made without departing from the gist of the present invention.

It is a figure which shows the needs of each stakeholder. It is a graph which shows the example of the click rate at the time of 1 million impression. It is a histogram which shows the example of the number of creatives, and a click rate. It is a graph which shows the normal distribution of the number of creatives with respect to a click rate. It is a graph which shows the example of change of normal distribution based on the number of creatives. It is a figure which shows the example of the creative published on the web page. It is a graph which shows the example of a transition of the click rate of a 0.063% banner advertisement. It is a graph which shows the example of a transition of the click rate of a 0.063% banner advertisement. It is a graph which shows the example of transition of the click rate by a plurality of 0.063% banner advertisements. It is a figure which shows the example of the normal distribution (standard deviation large) of the number of creatives with respect to a click rate. It is a graph which shows the example of transition of the click rate by a plurality of 0.063% banner advertisements. It is a figure which shows the example of the normal distribution (standard deviation small) of the number of creatives with respect to a click rate. It is a graph which shows the relationship between normal distribution and a standard deviation. It is a figure which shows the outline | summary of the system which concerns on one embodiment of this invention. It is a flowchart which shows the process example of the whole system which concerns on one embodiment of this invention. It is a figure which shows the device of the creative submission which concerns on one embodiment of this invention. It is a figure which shows the outline | summary of the examination which concerns on one embodiment of this invention. It is a block diagram which shows the structural example of the system which concerns on one embodiment of this invention. It is a figure which shows the example of an internal structure of the submission and examination interface apparatus which concerns on one embodiment of this invention. It is explanatory drawing of creative DB which concerns on one embodiment of this invention. It is explanatory drawing of case DB which concerns on one embodiment of this invention. It is a figure which shows the example of an internal structure of the creative optimization server which concerns on one embodiment of this invention. And FIG. 11 is a block diagram illustrating a configuration example of a general-purpose computer. It is a block diagram which shows the example of a submission process which concerns on one embodiment of this invention. It is a figure which shows the example of the login screen which concerns on one embodiment of this invention. It is a figure which shows the example of the guidance purpose selection screen which concerns on one embodiment of this invention. It is a figure which shows the example of the display screen of the induction | guidance | derivation objective and posting form which concerns on one embodiment of this invention. It is a figure which shows the example of the display screen of the guidance objective and posting content confirmation which concerns on one embodiment of this invention. It is a figure which shows the example of a display of the submission creative which concerns on one embodiment of this invention. It is a flowchart which shows the example of a screening process of the submission / screening interface apparatus which concerns on one embodiment of this invention. It is a figure which shows the example of the examination screen which concerns on one embodiment of this invention. It is a sequence diagram which shows the example of a creative optimization process which concerns on one embodiment of this invention. It is a figure which shows the example of a result display which concerns on one embodiment of this invention It is a block diagram which shows the example of point distribution which concerns on one embodiment of this invention. It is a figure which shows the example of a relationship improvement of each stakeholder which concerns on one embodiment of this invention.

Explanation of symbols

  32 ... Submission / examination interface device, 33 ... Creative optimization server, 33A ... Creative database, 51 ... Submission examination determination unit, 52 ... Submission processing unit, 52A ... Login processing unit, 52B ... Guidance purpose selection unit, 52C: Guidance destination / guidance purpose display screen generation unit, 52D ... Creative submission unit, 53 ... Examination processing unit, 53A ... Login processing unit, 53B ... Guidance purpose selection unit, 53C ... Inappropriate creative selection unit, 53D ... Inappropriate Creative deletion unit 61 ... Creative table 62 62 Case table 71 71 Creative input / output unit 72 Event determination unit 73 Log recording unit 74 Index calculation unit 75 Reliability / expected value calculation unit 76 ... Reading probability calculation unit, 77 ... Price calculation unit

Claims (10)

For a plurality of creatives in which the relationship between the evaluation index, which is information for evaluating the effectiveness of the creative, and the number of creatives is normally distributed, an index calculation unit that calculates each evaluation index at the present time ,
Reliability for calculating the reliability of the evaluation index based on the number of times the creative is exposed to the web page at the time of calculating the evaluation index and the past statistical data corresponding to the number of exposures stored in the database An arithmetic unit;
An expected value calculator that calculates the expected value of the evaluation index in the future by multiplying the value of the evaluation index by the reliability ;
Based on the expected value of the evaluation index calculated by the expected value calculation unit, a read probability calculation unit that changes the exposure frequency that is the number of exposures to the web page within an arbitrary time of the corresponding creative;
Wherein the plurality of high frequency of exposure from the creative creative extracted, and a creative output unit that outputs the U Ebupeji comes a hyperlink to the creative,
The reliability calculation unit updates past statistical data associated with the number of exposures of the creative stored in the database based on the performance of the current evaluation index.
A creative optimization server characterized by that . The creative optimization according to claim 1, wherein the reliability of the evaluation index calculated by the expected value calculation unit is multiplied by a constant set in advance according to the number of exposures of the corresponding creative at the time when the evaluation index is calculated. server. The creative optimization server according to claim 2, wherein the evaluation index is a click rate. The creative optimization server according to claim 2, wherein the evaluation index is a conversion rate. A creative optimization system comprising a database and a creative optimization server for optimizing creatives,
The database is
Accumulate multiple creatives and historical statistics associated with the number of creative web page exposures ,
The creative optimization server
For a plurality of creatives in which the relationship between the evaluation index, which is information for evaluating the effectiveness of the creative, and the number of creatives is normally distributed, an index calculation unit that calculates each evaluation index at the present time ,
Based on the exposure times to the appropriate creative web pages, and past statistical data corresponding to the number of times out said exposure stored in the database at the time of calculation of the evaluation index, calculates the reliability of the evaluation index trust Sex calculation unit,
An expected value calculator that calculates the expected value of the evaluation index in the future by multiplying the value of the evaluation index by the reliability ;
Based on the expected value of the evaluation index calculated by the expected value calculation unit, a read probability calculation unit that changes the exposure frequency that is the number of exposures to the web page within an arbitrary time of the corresponding creative;
Wherein the plurality of high frequency of exposure from the creative creative extracted, and a creative output unit that outputs the U Ebupeji comes a hyperlink to the creative,
The reliability calculation unit of the creative optimization server updates past statistical data associated with the number of exposures of the creative stored in the database based on the performance of the current evaluation index.
A creative optimization system characterized by that . Creative optimization system of claim 5, the user is received via the electric communication line creative input with trafficking form from the client terminal, before the interface device to accumulate Kide database, further comprising a. The interface device, the examination screen data for a particular user selects a creative from multiple creatives received, a client terminal of the particular user via the telecommunication line in accordance with the login process from a specific user The creative optimization system according to claim 6. The interface device transmits the screen data including the evaluation index of the received plurality of creatives, the rank of the expected value , and information on the grant points based on the rank through the telecommunication line. The creative optimization system according to claim 7. The index calculation unit provided in the creative optimization server calculates each evaluation index at the present time for a plurality of creatives in which the relationship between the evaluation index, which is information for evaluating the effectiveness of the creative, and the number of creatives has a normal distribution. Steps,
The number of exposures to the web page of the corresponding creative at the time when the evaluation index is calculated by the reliability calculation unit provided in the creative optimization server, and past statistical data corresponding to the number of exposures stored in the database Based on the reliability of the evaluation index,
A step of calculating an expected value in the future of the evaluation index by multiplying the reliability of the value of the evaluation index by an expected value calculation unit included in the creative optimization server ;
Based on the expected value of the evaluation index calculated by the expected value calculation unit, the exposure frequency that is the number of times the creative is exposed to the web page within an arbitrary time is changed by the read probability calculation unit included in the creative optimization server And steps to
Wherein the creative output unit creative optimization server comprises extracts creative high exposure frequency from the plurality of creative and step of outputting the U Ebupeji comes a hyperlink to the creative has,
In the step of calculating the reliability of the evaluation index, the past statistical data associated with the number of exposures of the creative stored in the database is updated based on the actual performance of the current evaluation index.
Creative optimization method characterized by that . Computer
For multiple creatives relationship information in which the evaluation index and creatives number to evaluate the effectiveness of the creative is a normal distribution, means for calculating the respective evaluation index at the present time,
Means for calculating the reliability of the evaluation index based on the number of exposures to the web page of the corresponding creative at the time of calculating the evaluation index and the past statistical data corresponding to the number of exposures stored in the database ; ,
Means for multiplying the value of the evaluation index by the reliability and calculating an expected value of the evaluation index in the future ;
Based on the expected value of the evaluation index out calculation, and means for changing the exposure frequency is the number of exposures to a web page in the appropriate creative any time,
Extract the creative high exposure frequency from the plurality of creative, and means for causing output c Ebupeji comes a hyperlink to the creative,
As means for updating past statistical data associated with the number of exposures of the creative stored in the database based on the performance of the current evaluation index
A program to make it work .

Images