JP5595556B1 - System and method for realizing an advertising platform with a new cost model - Google Patents

Abstract

A method for providing an online advertisement is introduced.
In one embodiment, the systems and methods described in this application increase the efficiency and effectiveness of advertisements utilizing the Internet. Increased efficiency can be related to ads being associated with users, reducing advertising space loss or waste, increasing conditions favorable to advertising publishers, displaying ads only for the appropriate duration, and being displayed This is achieved by charging the advertiser according to the degree to which the user is in contact with the advertisement.
[Selection] Figure 12B

Description

  This application relates generally to methods and systems for processing and displaying advertisements. Such processing and display of advertisements includes, for example, determining at least an essential portion of viewable ad space or viewable ad space for a predetermined duration and placing an advertisement in these ad space.

  Advertising in the field of electronic commerce includes various forms and methods of advertising, such as search-based advertising, branding advertising, and the like. One of two important types of advertising mechanisms or e-commerce-based advertising is “direct”, such as cost per click (CPC) where costs are per click or cost per action (CPA) where costs are per action or conversion. -Response advertisement. Another important form of electronic commerce-based advertising is “branding advertising”. In this advertisement, the cost is not based on clicks, behavior or effectiveness, but on the number of “impressions”, usually a set of 1,000 impressions, or cost per mill (CPM). A single online ad impression is a single appearance of an advertisement on a web page. Each time an ad appears on the user's screen, the ad server counts that display as one impression.

  Generally, advertisements are bought and sold based on cost per impression base regardless of the ad slot / slot / frame when it is displayed to the end user. However, not all advertising spaces have the same impact or effect on advertising. Some ad slots do not even exist in the viewable area of a web page. For example, the advertising space may exist at the bottom of the web page outside the initial viewable area, which is visible only when the user scrolls down to the bottom of the web page. Such advertisement impressions may not be effective because they are not even viewed by the user. However, advertisements are still charged for delivered impressions. Thus, not all ad impressions result in the same return on investment (ROI).

  Furthermore, in the Internet and electronic commerce worlds, advertisements are displayed to the user over and over, regardless of whether the user feels it is interesting, relevant, and attractive. That is, the current advertising system distributes advertisements regardless of whether the advertisement is effective or not. Such a technique does not provide a reliable ROI for the advertiser.

  These and other objects, features and characteristics of the present invention will become more apparent to those skilled in the art from a review of the following detailed description and the appended claims and drawings, all forming a part of this specification. The contents of the drawings are as follows.

FIG. 1 is a simple general detail of an exemplary environment in which the present invention can be implemented. FIG. 2 is a block diagram illustrating a typical architecture of a platform server. FIG. 3 is a simple general detail of an exemplary environment in which embodiments of the present invention can be implemented. FIG. 4A is a block diagram illustrating a system for presenting advertising space to an auction. FIG. 4B shows the order of operations performed by the advertising publisher, the web browser, and the SSP to identify the advertising space that will be viewable by the user. FIG. 4C shows an example where an ad slot in a web page is monitored by an area around the ad slot. FIG. 4D shows an example of a website viewed by a user using a web browser. FIG. 5 is a flowchart of a process for presenting an advertising space to an auction. FIG. 6 is a block diagram illustrating measurement of user attention to various content areas displayed on a web page. FIG. 7A is a flowchart of a method that can be utilized by this system to measure the viewable time of advertising content displayed by a given impression on a given web page. FIG. 7B is a flowchart of a method that can be utilized by this system to measure the viewable time of advertising content displayed by a given impression on a given web page. FIG. 8A shows various advertising ecosystem participants in the general environment with which the user interacts when online. FIG. 8B shows the interaction between the various participating entities when the pricing model is implemented by a third party advertising server. FIG. 8C shows the interaction between the various participating entities when the pricing model is implemented by SSP. FIG. 8D shows the interaction between the various participating entities when the pricing model is implemented by the DSP. FIG. 9A shows an overview of a process for implementing a plurality of advertisement placement methods that incorporate the viewable time concept. FIG. 9B shows an overview of a process for implementing a plurality of advertisement placement methods that incorporate the viewable time concept. FIG. 10A shows a process diagram of a use case where one advertisement is delivered on two pages. FIG. 10B illustrates a use in which one ad delivered in a first ad slot (also referred to as an “advertising rectangle”) is later placed in a second ad slot when the user scrolls away from the first ad slot. Another sequence diagram of the case is shown. FIG. 11 shows the probability of getting the number of impressions necessary to acquire viewable time exceeding “T” seconds. FIG. 12A illustrates a method that can be used by an advertising platform to deliver based on an advertising delivery scheme that incorporates advertising targeting and frequency limits. FIG. 12B illustrates a method that can be used by the advertising platform to estimate the relevance score for a given advertisement. FIG. 12C shows the distribution of advertisements for complete viewing distribution and random distribution. FIG. 13A illustrates the interaction between various participants in the advertising ecosystem when the advertising targeting scheme is performed by a third party advertising server. FIG. 13B shows the interaction between the various participants when the advertising targeting scheme is performed by the DSP. FIG. 13C shows between the various participants when the DSP performing the advertising targeting scheme returns bids to the SSP at a bid price lower than the lowest auction price that the advertising publisher is willing to accept to display advertisements from the DSP. Shows interaction. FIG. 13D shows ad serving in response to a client ad tag request by the SSP when the bid amount received from the DSP is lower than the lowest auction price that the advertising publisher is willing to accept to display the advertisement from the DSP. Here are the methods you can use to: FIG. 14A shows a comparison between the GRP for a TV advertising campaign and the eGRP for an online advertising campaign. FIG. 14B shows a general overview of a system that can be used to estimate the eGRP achieved by an advertising campaign. FIG. 14C shows a general overview of a system that can be used to predict the eGRP that can be achieved with a given advertising cost. FIG. 14D is an illustration of a method that can be used to estimate eGRP achieved by an advertising campaign. FIG. 14E is an illustration of a method that can be used to predict possible eGRP that can be achieved by an advertising campaign based on a given bid amount. FIG. 14F illustrates a method that can be used to determine the cost to achieve the desired eGRP for an advertising campaign. FIG. 14G shows an example of eGRP calculated based on the various parameters described above. FIG. 14H is an illustrative example of the interaction between various participants in evaluating eGRP in one embodiment of a system used to evaluate eGRP. FIG. 14I is an illustrative example of a user interface (UI) displayed on a computer system that can be used by an advertiser or advertising agency that interacts with a computer server to estimate or calculate eGRP. FIG. 14J is an illustrative example of a UI displayed on a computer system that can be used by an advertiser or advertising agency that interacts with a computer server to estimate or calculate eGRP. FIG. 14K is an illustrative example of the interaction between various participants in determining the cost of achieving the desired eGRP for an advertising campaign in one embodiment of the system. FIG. 14L is an illustrative example of a UI displayed on a computer system that can be used by an advertiser or advertising agency interacting with a computer server to estimate the cost (or bid amount) to achieve the desired eGRP. FIG. 15 is a high-level block diagram illustrating an example of a computer system architecture.

  The headings used in this application are for convenience only and do not affect the scope or meaning of the claimed invention. In the drawings, the same reference numbers and acronyms represent elements or acts of the same or similar structure or function for easy understanding and convenience. To easily identify a particular element or action statement, the highest digit in a reference number refers to the drawing number in which that element is first introduced (eg, element 114 is first introduced with respect to FIG. 1).・ Discussed).

  FIG. 1 and the following discussion are simple general details of an exemplary environment in which the present invention can be implemented. Although not required, aspects of the invention are described below in the general context of computer-executable instructions, such as routines executed by a general purpose data processing device (eg, a server computer or a personal computer). Those skilled in the art will appreciate that the present invention can be implemented in other communication system configurations, data processing system configurations, or computer system configurations including: wireless devices, Internet devices, handheld devices (personal digital assistants (PDAs)) Including) wearable computers, all types of mobile or mobile phones, multiprocessor systems, microprocessor applications or programmable consumer electronics, set-top boxes, network PCs, minicomputers, mainframe computers, etc. In fact, the terms “computer”, “server” and the like are used interchangeably in this application and refer to any of the devices and systems described above.

  Although aspects of the present invention, such as certain functions, are described as being performed exclusively on a single device, the present invention can also be performed in a distributed environment where functions or modules are shared between disparate processing devices. . The heterogeneous processing devices are coupled via a communication network such as a local area network (LAN), a wide area network (WAN), or the Internet. In a distributed computing environment, program modules can be located in both local and remote memory storage devices.

  Aspects of the invention include magnetic or optically readable computer disks, wiring logic or pre-programmed chips (eg, EEPROM semiconductor chips) nanotechnology memory, biological memory, or other data storage media. It can be stored or distributed on a tangible computer readable medium. Alternatively, computer-implemented instructions, data structures, screen displays, and other data related to the present invention are transmitted over the Internet or other networks (including wireless networks) in a propagation medium (eg, electromagnetic waves, sound waves, etc.) It can also be distributed over a period of time with the above propagating signal. Some implementations provide data over analog or digital networks (packet switched, circuit switched or other schemes).

  As shown in FIG. 1, a user uses a personal computer device (eg, telephone 102, personal computer 104, etc.) to communicate with a network and / or view a display communicated via network 110. The term “telephone” as used in this application refers to a mobile phone having a communication capability to connect to a network, a personal information terminal (PDA), a portable e-mail device (eg, Blackberry®), a portable media player ( For example, iPod Touch (registered trademark)) or other devices. In one example, the telephone 102 may include one or more cellular transceivers or base station antennas 106 (in the cellular device), access points, terminal adapters, routers or modems 108 (in the IP applied telecommunications device) or any of these. Connect using a combination (centralized network embodiment). In some cases, one or more users can use the electronic display device 132 (electronic overhead display device, large outdoor electronic display device, etc.) to view information transmitted via the network. For this description, information communicated includes, for example, advertisements displayed as such, advertisements displayed with web pages or other online media that the user views / experiences. The concept behind the display of such advertisements is explained in more detail in the following sections.

  In some cases, the network 110 is the Internet, which can be a telephone 102 (eg, with WiFi capability), a personal computer 104, or various servers (eg, web server 120) to which the electronic display device 122 is connected via the network. ) Allows access to content provided via. In some cases, the type of network 110, particularly when phone 102 is used to access web content via network 110 (eg, when connecting to network 110 using 3G or LTE services on phone 102). Can be a cellular, IP-based or centralized telecommunications network, including Global System for Mobile Communications (GSM), Time Division Multiple Access (TDMA), Code Division Multiple Access (CDMA), Orthogonal Frequency. Division Multiple Access (OFDM), General Packet Radio Service (GPRS), Enhanced Data GSM Environment (EDGE), Advanced Mobile Phone System (AMPS), Worldwide Interoperability for Microwave Access (WiMAX), Universal Mobile Telecommunications System (UMTS), Evolutio Including, but not limited to, n-Data Optimized (EVDO), Long Term Evolution (LTE), Ultra Mobile Broadband (UMB), Voice over Internet Protocol (VoIP), Unlicensed Mobile Access (UMA), and the like.

  In some cases, the user connects to the platform server 114 via the network 110 using one of the computing devices (eg, telephone 102, personal computer 104, etc.). In one embodiment, platform server 114 includes a server computer 116 that is coupled to a local database 118. “Platform server” as referred to in this application refers to individual or multiple server stations or other computing devices. In one embodiment, the platform server is a web server that can host a website and store content related to the website (eg, various web pages). In some embodiments, the platform server is separate from the web server and communicates with the web server to supply, manage, and / or control content generated by the web server. In general, the platform server 114 is a variety of modules that allow advertisement information to be collected from advertisers who are strategically engaged in advertising activities and wish to coordinate the next advertisement and take over to the end system. (Implemented by software or hardware). In some embodiments, the platform server independently coordinates advertisement processing and final display. In some embodiments, as described in the example of FIG. 2, the platform server may include one or more specific advertising activities (eg, providing bidding functionality, providing a campaign conversion module, etc. Provide an interface (e.g., API) with an existing ad network platform. This will be described in more detail below. Also, as described in further detail in this application, the management server 114 includes one or more functional building blocks to achieve each of the functions described above.

  As shown in FIG. 1, in some embodiments, the personal computer device and management server 114 are connected to one or more web servers (eg, web server 120) via a network 110. Each web server corresponds to a computer station that allows website providers to provide web content (eg, web pages) that can be accessed via network 110 by, for example, a personal computing device.

  A platform server as defined in this application may be a separate server that provides the services described in this application to, for example, one or more web providers. In other examples, the management server can itself be a website provider that also performs services that implement the techniques described in this application. Other examples of implementing a management server are of course equally suitable for performing the techniques described in this application, as will be appreciated by those skilled in the art.

  With respect to the system described in this application, in one embodiment, the platform server is implemented as a search system that enables an advertisement display method, and in advertisement data (eg, from the advertisement client 132) obtained via a network. Based on, it is possible to show one or more advertisements simultaneously or at various different times. The platform server 114 then communicates advertisements to an advertisement display system (eg, the user's personal computer device) where individual advertisements are displayed for a predetermined length of time or according to variables set by the advertisement client. .

  Separate advertisements x1, x2, x3,. . . Consider a typical scenario where xp is shown to the user as being processed and output by the platform server 114. These advertisements are t1, t2, t3,. . . It is predetermined as shown over the length of tp. However, this does not mean that the displayed advertisement is also predetermined. For example, when a user walks around the Internet using a PC, various advertisements are presented in various situations, and the techniques described in this application are based on the predetermined time that these advertisements are posted. Including the case where it lasts for a long time.

  The advertisement described in the present application includes an unlimited number of moving images, still images, banners, animation images, and the like. When processed by the platform server, such advertisements are presented over a period of time, such periods being predetermined by the advertiser, for example. When the advertisement is a moving image, the length of the created moving image or the playback time specified by the advertiser is the display time of the advertisement. When the advertisement is a still image, the display time is a display time specified by the advertiser.

  “Display” of an advertisement refers to the display of the advertisement that is substantial or meaningful. For example, when the user scrolls down the screen on the web screen, it is preferable that the advertisement is also scrolled in parallel to fit on the screen on which it is displayed. However, if the above method is not possible and the user scrolls so that the advertisement is no longer visible on the display screen, the advertisement should be stopped and the time that the advertisement was played is recorded (as described below). At least for the purpose of calculating the cost per second of ad display). When the advertisement returns to the on-screen display, the advertisement playback should resume and the total playback time is recorded at the end of the advertisement or at the next stop.

  The determination of whether or not an advertisement is displayed is, for example, when a certain ratio of advertisement is not displayed on the screen, it can be considered that the advertisement is not displayed on the screen be able to. Here, the “certain ratio” can refer to a ratio at which it is difficult to consider that the advertisement is substantially viewed, for example, a ratio of 50% or more. However, over 50% is merely an example, and this ratio need not be limited to 50% or more. For example, you can divide the part where the advertising display is important (for example, the part showing the name of the advertised product or service) and the part that is less important. It can also be determined that the advertisement is displayed on the screen.

  The techniques discussed in this application include a bidding system that allows advertisers to bid for a certain ad slot and its duration. As shown with respect to FIG. 2, platform server 114 includes a bid platform module 202 that, in some embodiments, enables bid operations. Illustratively, the bidding platform module presents the appropriate GUI to the advertising client 132 that allows the advertiser to make appropriate selections and inputs. These are then captured by the bid platform module 202 for further processing and bid evaluation.

  If the advertiser knows the display length in advance, in some embodiments, the advertiser can sell the ad display time (from the advertiser ’s point of view) to determine the priority for displaying the ad (s). Bid as a method of “purchasing”). In other words, the amount of advertisements that can be displayed in the advertising space is generally finite. In the case of a web screen, when there are a plurality of advertisements that can be displayed on the same advertisement space, the order in which these advertisements are displayed becomes important. In particular, when displaying an ad in a specific ad space or displaying an ad for a specific keyword, the input is made at the maximum cost / price that the advertiser can afford for that specific combination of duration and order (e.g. , In a bid format). Clearly, the higher this cost / price, the higher the order or priority.

  In some embodiments, the platform server 114 includes logic for the purpose of two types of cost determination decisions and logic that reveals the target and the correlation between the two types. In some embodiments, as shown in FIG. 2, the platform server includes one or more of the following modules, each implemented by hardware, software, or firmware, or combinations thereof: An advertisement proposal module 222 that makes decisions about the type, content, duration, etc. of advertisements displayed on the site of the advertising media company and makes a harmonized proposal; The logic contained in this module includes, for example, an algorithm that identifies the importance, meaning, background, relevance, etc. of a particular website and reveals related advertisements accordingly. Further, the platform server 114 includes an advertisement accepting means 204 that accepts advertisements uploaded by the advertisement client 132. In some embodiments, the platform server 114 also includes an advertisement memory 208 that stores advertisements received from advertisers and an advertisement information memory 210 that stores information related to advertisements (eg, related information, order or priority information, etc.). . In some cases, these modules further include an advertisement selection module 216 and an advertisement distribution module 218. These are each configured to select an appropriate advertisement and send the advertisement to a predetermined web screen based on a decision made by the platform server.

  In some embodiments, these include means accessible online from the advertiser. Each component / module shown above can be implemented as a separate software or hardware unit or a combination thereof. In some embodiments, for example, an advertisement space suggestion module and advertisement bidding means for proposing advertisements to be placed on the advertisement space can be combined with or combined with the web server 120. In some embodiments, the platform structure may include, for example (when displaying an advertisement on a search results page), the keywords used for the search, and various attributes (gender, age) of the user for whom the advertisement is desired to be displayed. , Region, occupation, educational background, hobbies, etc.), and a GUI that suggests a page that allows the user to input a preferred time zone of the day for displaying the advertisement. According to these input data, a price per unit time for purchasing the advertising space and an input field for purchasing the advertising space (or input page) are displayed next. For the proposed advertising space, the advertiser enters the desired unit time price for purchasing the advertising space and the number of advertising spaces to purchase (eg, via a bidding platform). However, in some embodiments, the purchase of advertising space may be targeted for the total length of time that the advertisement is displayed.

  In some embodiments, the advertisement information memory 210 and the advertisement data memory 212 storing the advertisement itself include, for example, an advertisement information database means for storing information about the advertisement and an advertisement data database for storing the advertisement itself, respectively. Can be included.

  Returning to the drawing of FIG. The process of processing / displaying advertisements using the platform server will be described below with two scenarios: (1) When an advertisement is returned to a user viewing an advertisement on a web screen, (2) an electronic display device ( For example, when an advertisement is displayed to a plurality of users by an outdoor large electronic display device).

  When the advertiser accesses the bidding platform module of the platform server 114 as shown in FIG. 1, the system, for example, presents an input field for desired conditions for advertisement display. Advertiser 132 inputs the desired conditions accordingly. In response, the platform server 114 requests input of advertisements. The received advertisement and advertisement information are then stored in the advertisement video database and advertisement information database by the advertisement accepting device. In some embodiments, the information stored in the advertisement video database and the information stored in the advertisement information database are associated and characterized by a unique advertisement ID for each advertisement. In some embodiments, when advertising information is sent to the display device, the related information is also characterized by the advertising ID.

  In the first scenario, the user typically has an advertising display device that is loaded into a web browser (eg, a widget in a web page, etc.). At this time, in order to display an advertisement that matches the user's interest, information about the indicated page and a user ID are sent to the advertisement selection device of the platform server. The advertisement selection module 216 selects an advertisement to be displayed based on the received information and advertisement data stored in the advertisement information database. When the advertisement selection module 216 selects an advertisement to be displayed, the advertisement ID of the advertisement to be displayed is then sent to an advertisement screening device (eg, a user's computer).

  After receiving one or more advertisement IDs from the advertisement selection module 216, the advertisement transmission or distribution module 218 sequentially sends the one or more advertisements to the advertisement display device. Upon receipt, the advertisement screening device displays the advertisement to the user. In the case of video, the display time is generally determined by the length that the video advertisement is played. In the case of a still image, the display time is determined by the time specified by the advertiser.

  FIG. 3 shows a second scenario. In this case, the advertisement display device on which the user views the advertisement is not a device on the browser but an Internet connection device that performs LCD display such as a street advertisement (for example, device 122). In this scenario, since the advertisement display device is not provided on the web browser, there is no information on the basis for selecting the advertisement to be displayed. In such a case, the advertisement display device does not send information for selecting an advertisement, but instead only displays the advertisement continuously in a predetermined order. However, for example, when digital signage devices are provided in various stores and places, conditions for selecting advertisements, for example, display from 5 pm to 11 pm on a ramen shop device in a shopping street, etc. You can specify and display advertisements accordingly. In such a case, an advertisement that best matches such conditions is selected. In the case of video, the display time is generally determined by the length that the video advertisement is played. In the case of a still image, the display time is determined by the time specified by the advertiser.

  A third scenario for processing and displaying advertisements according to the techniques described in this application will be described with reference to FIG. In this example, the advertisement is displayed and displayed on the advertisement viewing device is not directly performed between the advertiser and the user, but uses an interface to the Demand Side Platform (DSP) 330 and the Supply Side Platform (SSP) 340. Done. The configuration of this exemplary embodiment constitutes an advertisement exchange that can include conventional advertisement exchanges in addition to the present teachings.

  In some embodiments, a DSP, SSP or both are included. The configuration can be a connection with a DSP, a configuration with a direct connection to an advertiser, or a combination thereof. Similarly, the configuration can be a configuration through a connection with an SSP, a direct connection to a user, or a combination. As contemplated by those skilled in the art, other similar combinations of one or more DSPs and SSPs can also be used as alternatives or variations of the above-described configurations.

  In this example, when an advertisement is sent by an advertiser, it is stored in a DSP that acts as an arbitrator on the advertiser side. The DSP then selects one advertisement exchange (not shown) from the advertisement exchange group. And this advertisement is released. In order for the device devised according to the invention to receive the advertisement, a bid is received from the advertiser via the DSP to determine the price of the advertisement.

  On the other hand, on the user side, the advertisement is received from the SSP, not directly from the device devised by the present invention. And this advertisement is shown. After receiving the advertisement display request from the user, the SSP selects one or more advertisement exchanges from which to receive advertisements and requests for advertisements. At this time, the system devised by the present invention (advertising exchange) has already received the advertisement request, but also receives information necessary for selecting the advertisement most suitable for the user and can display it according to this information. The most suitable advertisement is selected from the advertisements, and the advertisement is sent to the SSP. After receiving the advertisement, the SSP sends the advertisement to the user. The user sees the advertisement.

  In some embodiments, such an arrangement allows an advertiser to increase the effectiveness of their advertisements by expanding the capacity of the media / advertisement media company that displays their advertisements. Quantifiable profit outcomes on both sides of the source--the advertising media / advertising company will be able to select advertising that best fits the user's interest from a large number of advertisements. Revenue increases. From the user's perspective, for the same reason, advertisements are selected from a wider variety, and the user can see advertisements that fit their interests.

  FIG. 4A is a block diagram illustrating a system 400 for presenting advertising space to an auction in accordance with an embodiment of the disclosed technique. The system 400 includes an SSP 410 (similar to the SSP 340 of FIG. 3) that identifies the advertising space on the web page 405 and presents it to the auction. A DSP 415 such as an advertiser (similar to the DSP 330 in FIG. 3) participates in the auction by bidding on the advertising space. The web page 405 can include any online content that can be accessed by a user via, for example, a web browser or other proprietary application. Web page 405 includes one or more ad slots, each of which can represent one or more advertisements handled by SSP 410. A user who accesses the web page 405 can view advertisements represented in these advertisement spaces and perform operations related to the advertisements.

  The SSP 410 is configured to identify advertisement slots that can be viewed by the user for a predetermined duration and to present only these advertisement slots in the auction. In one embodiment, the SSP 410 identifies an advertising space in which a predetermined portion of the advertising space, typically a significant portion, can be viewed by the user. In one embodiment, the critical portion can be at least 50% of the advertising space and the predetermined duration can be at least 100 milliseconds. After the ad slots are determined, the SSP 410 presents these ad slots to the auction and requests the advertiser to bid on these ad slots. This request also includes a response time indicating the deadline for which the advertiser must submit his bid.

  In one embodiment, the SSP 410 identifies advertising space that can be viewed by the user using a script pre-installed on the web page 405. Web page 405 includes various preinstall scripts that monitor web page 405 as it is viewed by a user via a web browser. These scripts monitor whether one or more pixels around a given ad slot (also referred to as an “ad slot”) on web page 405 have become visible to the user. This script uses a web browser to indicate that these ad slots may be viewed by the user (which can be used to identify and request ad content, as described below) To SSP 410 (or advertisement server) upon request.

  In response to a bid request from the SSP 410, the advertiser / DSP 415 submits a bid to the SSP 410 along with the advertisement to be handled in the advertising space. After receiving the bid from the advertiser, the SSP 410 will place a successful bid based on predetermined criteria such as the second price auction procedure or other convenient method described in US patent application Ser. No. 13/609146. decide. SSP 410 notifies the advertiser about the winning bid for the impression. Next, the successful bid advertiser's advertisement is placed in the advertising space. Therefore, by placing the advertisement in the viewable advertisement space, the probability that the advertisement is not viewed by the user decreases.

  FIG. 4B shows a sequence of operations performed by the advertising publisher, the web server, and the SSP 410 to identify advertising slots that are likely to be viewed by the user. Here, one or more scripts are installed on the advertising publisher's web page that can be used to monitor the area around the advertising slot of the web page (eg, an area defined by pixels). When the user's web client requests a web page, such a script is included in the web page provided to the web client, allowing for monitoring of user operations related to this web page. When the script detects that a portion of the monitored area (eg, one or more pixels of the monitored area) is available for viewing by the user, the web client is used to request an advertisement tag from the SSP 410 To do. Here, an advertisement that can be shown in the advertisement slot in the monitored area is downloaded using the advertisement tag.

  FIG. 4C shows an example of a web page 435 having an advertisement slot 455 monitored by an area 450 around the advertisement slot 455. The script monitors an area 450 around the advertisement slot 455 defined by x and y wide pixels around the slot 455. Thus, when the region 450 around the ad slot 455 enters view on the web browser 440, the monitoring script can inform the SSP 410 that the ad slot 455 has become visible to the user. By monitoring such areas around the ad slot, the true degree of viewability of the ad slot and thus its true price can be assessed. For example, if an ad slot is at the bottom of the page, it will not be very viewable and is usually low in price. However, since the content of a page usually ends at the bottom, user involvement tends to shift from content to ad slots.

  In this situation, the true value of this advertising slot is recognized and generates higher revenue for the advertising media company. In some embodiments, the bottom slot of a page is auctioned only if the user's involvement is directed to the bottom content of the page. By doing so, the advertising publisher of the web page can require the user to pay a higher price for the increased viewability of the ad slot, and the user's involvement is directed toward the bottom content of the page. The ad slot will only be auctioned (and this low-priced ad slot will not be auctioned when there is little or no ad slot viewability) be able to.

  In some embodiments, elements (such as video content) can be dynamically added to the end of a web page from time to time or as time passes, depending on user operations. For example, when a user is scrolling over a video hyperlink, the video can be dynamically appended to the bottom of the web page currently loaded on the user's web browser. Here, the user is either automatically guided to the added video or requested to be guided based on some user operation. Therefore, such dynamic tail addition can be performed outside the area that can be viewed by the user. By utilizing the monitoring area in the appended elements, the effectiveness of the advertisement slots included in the appended elements on these pages can be increased.

  In some embodiments, a monitoring area is used to monitor a portion of a web page regardless of the content of the monitored portion of the web page, and a portion of the monitored portion of the web page can be viewed by the user Sometimes it performs one of many possible actions. For example, when a monitoring area is added to the bottom of a web page and pixels belonging to the monitoring area become available to the user, an ad slot can be dynamically added to the bottom of the web page. . In this case, the advertisement slot can completely occupy the newly appended portion of the web page. As a result, the user is attracted to the advertisement displayed in the newly added advertisement slot without interruption.

  In another example, the monitored area can be embedded in the center of the web page. When the user scrolls through the web page and the portion of the monitored area in the center of the web page becomes available for viewing, the advertising slot is replaced with a designated portion of the web page (e.g., by replacing the content in the designated portion of the web page (e.g. In the monitored area). Here, the ad slot is not appended to the web page. The size of the web page remains the same. Instead, a portion of the web page content can be dynamically changed to accommodate the new ad slot. This advertisement slot can be sold at an auction in the same manner as the advertisement slot in the monitored area. In some embodiments, the content to be changed is not changed (eg, moved through an invisible part of the content to accommodate a new ad slot) and the original is generated in response to a user operation where the user is newly created and closes the ad slot. It can be returned to the form.

  FIG. 4D shows an example of a web page 460 (also referred to as a “website”) viewed by a user via the web browser 465. This web page includes a monitoring area (not shown in FIG. 4D) that monitors all pixels at the bottom of the web page 460. When the user scrolls the web page 460 and the bottom pixel in the monitored area becomes viewable to the user by the web browser 465, 475, the web browser 475 loads the ad slot 480 and displays it. A request is made to append to the monitored area, that is, at the bottom of the web pages 460 and 470. The advertisement slot can then be auctioned to display advertisements using the various viewable advertisement slot identification methods described above.

  FIG. 5 is a flowchart of a process 500 for presenting advertising space to an auction in accordance with one embodiment of the disclosed technique. Process 500 may be performed in a system such as system 400 of FIG. 4A. In step 505, the SSP is notified of the user's web visit. In decision step 510, the SSP determines whether or not a predetermined portion (important portion) of the advertising space was viewable. In response to determining that a significant portion of the advertising space was viewable, the SSP determines in step 515 whether the significant portion of the advertisement was viewable for a predetermined duration.

  In response to determining that the inventory was viewable for a predetermined duration, the SSP presents the inventory to the advertiser / DSP server in an auction at step 520 and prompts the advertiser for the inventory. Invite to submit a bid. In step 525, the SSP receives a bid from the advertiser and determines a successful bid based on predetermined criteria. In step 530, the SSP notifies the advertiser of the successful bid and the successful bid price. In step 535, the successful bidder's advertisement is placed in the advertising space.

  In one embodiment, process 500 is performed for all ad slots on a web page. Reference is again made to 505-515. Tags such as HTML tags are embedded in the web page in order to obtain data related to the user's web page visit, viewability of advertisement space, duration of viewability, etc., and these data are sent to the SSP.

Refer again to step 520. When sending a request for the submission of bids to the selected inventory to SSP advertiser, SSP also required to perform its bid predetermined time limit, in t _b. The SSP designates this deadline to the advertiser by including such a deadline in the request or by any other suitable method.

  Refer again to step 525. In addition to the bid price, the advertiser responds to the bid request by submitting additional information such as an advertisement to be placed in the advertising space when the advertiser has made a successful bid. Refer again to step 530. The winning bidder and price are determined using a second price auction method or other suitable method.

<< Measurement of viewing time of ad creative >>
As described above, in some cases, a user's level of engagement with advertisement creation can be measured by the user's response to advertisement creation (also referred to as “advertisement” or “advertisement content”). In addition, the level of user engagement is also used to measure the effectiveness of advertisements to the user. In some embodiments, the effectiveness of an advertisement for a user is utilized when determining the cost of displaying the advertisement to the user. In one example, the effectiveness of an advertisement for a user is measured as a function of the duration that the user may be alerted to the displayed advertisement (the “measured viewable time” of the displayed advertisement). Also called).

In some embodiments, the user's attention is a ratio of display advertisements that are viewable (ie, visible) to the user via a web client (eg, a web browser) (or a ratio of that content that is related to the content viewing user's attention measurement). Measured as a function. If the rate at which an ad is displayed is below the minimum threshold that might be needed to attract the user's attention to the ad, it is unlikely that the user has paid attention to the portion of the ad that is displayed ( That is, the measured viewable time is zero). In one embodiment, information about the location of the advertising content with respect to the region of advertising content displayed by the web client is collected every _ts seconds. Let h and w be the height and width of the content, respectively. Using this position information, the ratio to the overall size of the advertising content viewable portion of advertising content, r _v is computed as _{r v = (S V) /} (h × w). Here, _SV is the area of the portion of the advertising content shown on the web client. In one example, the minimum threshold (θ _v ) is set to 0.5. In this case, it is considered that only the advertising content with the ratio r _v > = θ _v gets the user's attention.

  In some embodiments, user attention is further measured as a function of the visibility (also referred to as “focus”) of the web client displaying the advertisement to the user. In one example, when the web client is not visible to the user, the user can be considered to pay little attention to the advertisement being displayed by the web client (i.e., the measured viewable time is Zero). For example, a window of another application running on the user's computing device blocks the web browser displaying the advertisement from the user's view, and the user's attention is mostly directed to the displayed advertisement. It may disappear. In another example, if a web browser with multiple tabs is displaying multiple web pages at the same time, only the content that is displayed in one of the tabs that are actually viewed by the user will be focused. (Ie visible to the user). If the displayed advertisement does not exist on the tab that is actually viewed, the user pays little attention to the displayed advertisement.

In some embodiments, user attention can also be measured as a function of one or more operations related to the web client. In one embodiment, since a user's last user operation (with respect to the web client), at least one operation per _tu seconds (a possible user indicating user interaction with the web client) When doing (from a subset of operations), the user is paying attention to the web client and possibly the portion of the advertising content being displayed by the web client. In one example, t _u is a predetermined threshold that is empirically determined so that a user action performed at least once every t _u seconds represents the user's continued attention.

  FIG. 6 is a block diagram 600 illustrating measurement of user attention for various content regions 606-612 (eg, ad slots) displayed on web page 602. Here, some of the content areas 608 and 610 are displayed to the user by the web browser 604, but the other 606 and 612 are not displayed. As noted above, user attention can be measured by a function of the percentage of a given content area being displayed to the user, the focus of the web browser 604 on the user, and the user operation with respect to the web browser 604. The content of the web page 604 in the web browser 604 is displayed to the user, but the content of the web page 604 outside the web browser 604 is not displayed.

Here, content # 1 606 and content # 4 612 are outside of the web browser 604, and thus no part of the content 606, 612 is displayed to the user, and therefore each gets little user attention (or zero). Measurement viewing time). On the other hand, content # 2 608 and content # 3 610 are partially viewable for the user. Content # 3 610, the ratio of content displayed to the user by the web browser 604 (more than perhaps theta _v threshold) high, if the other conditions related to attention measurement of the user (if any) is met As a result, the viewable time can be acquired. On the other hand, the content # 2 608 is in part being displayed, the ratio of the content displayed to the user (below perhaps theta _v threshold) lower by the web browser 604, other related attention measure user Even if the condition (if any) is satisfied, the viewable time cannot be acquired.

  FIG. 7, including FIGS. 7A and 7B, is a method that can be utilized by system 415 to measure the viewable time of advertising content displayed by a given impression on a given web page (also referred to as a “website”). 700 is shown. The system 415 begins delivering advertising content that is displayed to the user with a given impression at step 702 of the method 700. The system 415 determines at step 704 of the method 700 whether the user is still left on a given web page with a given impression. Step 704 may be by any known method such as determining whether at least one of the web pages open in the web page has the same web address as the web address for the given web page. Can be executed. If it is determined in step 704 that the user is still on the web page, the system 415 determines in step 706 of the method 700 that the viewable time of the displayed advertising content (ie, the user's attention duration). ). However, if it is determined in step 704 that the user has left the web page for this impression, the system 415 stops delivering advertising content displayed to the user in step 708.

In one embodiment, system 415 uses method 710 to measure the viewable time of the displayed advertising content. System 415 begins measuring the viewable time of the advertising content at step 712 of method 710. System 415, in step 714, determines whether the next three conditions are met: (1) whether the ratio of the area of a given content to be displayed to the user is greater than theta _v, (2) the web・ Focus on browser users, (3) User operations related to web browsers. If these conditions are met, the system 415 proceeds to step 716 and continues to measure the duration as long as the advertisement is displayed at a given impression (or if it begins the viewable time measurement process). ,Start). After a predetermined period of time, the system 415 evaluates step 714 again to determine whether the user's attention is probably still concentrated on the advertisement before continuing to measure the viewable time of the advertisement. To do. If the condition is met, the system 415 continues to measure the viewable time of the displayed advertisement. However, if the condition is not met at step 714, the system 415 proceeds to step 718 and stops measuring the measurement time. The system 415 determines in step 704 whether the user is still on the web page before restarting the viewable time measurement in step 706.

<< CPV pricing model >>
A CPV model is a pricing model that can be used to charge an advertiser only if the advertiser's advertisement has acquired a viewable time longer than a predetermined duration (eg, given by the advertiser). Thus, if the advertisement is not delivered for a predetermined duration, there is no cost to the advertiser for the advertisement already displayed to the user. The following discussion shows the methods that can be used to implement the CPV pricing model.

In the case of this method, a predetermined value “T” equal to the length of the guaranteed viewable time for a given advertisement is set by any one of the parties such as the advertiser, the DSP and the like. If an ad delivered (or a series of deliveries using the “ad follow” method described below) fails to acquire a viewable time longer than “T” seconds, the advertiser is charged for any portion of this ad delivery Not. Therefore, this method measures the viewing time T _v resulting from ads given advertisement (or a series of published by "Ad follow" method) for the user.

If T _v > T, the method charges the advertiser for advertising delivery. The billing amount can be determined by several methods including: (1) the amount determined by consultation with the advertiser, (2) by auction of impressions (used to display the advertisement to the user) The amount to be reached. The proposed CPV pricing model may be implemented by an SSP, DSP, or any third party advertising server that is given the possibility to collect or receive viewable time for a given advertisement. Further details of how different parties implement different pricing models are described in detail below.

<< CPV⇔CPM bid conversion >>
The following discussion shows one way to perform the conversion from CPV to CPM (and vice versa). Let “T” represent the length of time that a given advertisement can be viewed in the CPV model. For impressions that are auctioned (used to display the advertisement), collect data related to the impressions. For example, the web page www. nytimes. com, the impression is related to the top banner ad slot at the com. nytimes. com represents a specific opportunity to display in the top banner ad slot. The data collected for this impression is the number of impressions generated for that ad slot during a given period (eg, 1 day, 1 week, or 1 month) for the same given period The number of impressions that achieved a viewable time longer than T (“N _T ”).

The CPV can then be converted to an equivalent CPM value using equation (1) below, and vice versa:

  If the advertising platform (or advertising publisher) uses CPM as a cost metric to evaluate bids and / or charges for advertising delivery, as described above, the advertising platform (or advertising publisher) Thus, bidding (or pricing) based on CPS, CPX, and CPV can be converted to CPM bidding (or pricing). As can be clearly seen from the above discussion, the advertising platform (or advertising publisher) can use the above method to convert CPM bids to any other pricing, and use CPM as an intermediate cost criterion. Conversions can be made between pricing decisions (eg CPX → CPM → CPV).

<< Overview of various systems that realize pricing models >>
Viewable time, which is the basis for various pricing models and methods for evaluating bids, distribution fees, revenue, etc., is measured by various participating entities including DSPs, SSPs, third-party advertising servers, advertising publishers, etc. And can be realized.

  FIG. 8, which includes FIGS. 8A, 8B, 8C, 8D, shows illustrative examples of various methods that can be utilized by various participating entities that implement the pricing model.

  FIG. 8A illustrates various participating entities in a general environment 800 with which a user interacts when the user engages in an activity via the Internet. In FIG. 8A, the user is engaged online using a portable device 812A running a web browser 812B (ie, a web client). When a user visits a website hosted by advertising publisher 810 (advertisement publisher's server), advertising server 806 or SSP 804 sends an advertising tag to the user's web browser 812B. The tag then instructs the web browser 812B to download the advertisement from the Content Delivery Network (CDN). Furthermore, the DSP 802 interacts indirectly with the web browser via the SSP 804. The logging server 808 may also be used to receive information about viewable time acquired for each given advertisement that was posted at a given impression during the measurement of user attention. This report may degrade the quality of the measurement (since it takes time to report to the server), but the measurement performed before this report is made will have a viewable time of “T” seconds. It has been confirmed that it has been acquired, and the data that it has acquired “T” seconds remains valid.

  FIG. 8B illustrates the interaction 820 between the various participating entities when the pricing model is implemented by the third party advertising server 806. Here, the web client (ie, web browser) of the client 812B initiates a page request to the advertising media company 810. Advertising publisher 810 provides the requested page to client 812B. Client 812B then sends an ad tag request to ad server 806 for one or more ad slots (in the page received from ad publisher 810). The ad server 806 then returns an ad tag (for the ad displayed in the ad slot) with scripts that perform various functions such as monitoring user operations, measuring user attention, and the like. The client 812B sends the received advertisement tag to the CDN and requests the advertisement content related to the advertisement tag. The CDN responds to this request by sending advertising content related to the advertising tag to the client 812B. The client 812B displays the received advertisement content to the user. Furthermore, the client 812B further measures the viewable time of the displayed advertisement (when various conditions regarding user attention are satisfied as described above). The measured viewable time is transferred to the logging server when the advertising content has been displayed or when the user ends the session. It can be used to implement pricing models with other participating entities (including third party advertising servers) and to determine the cost of advertising delivery.

  FIG. 8C shows the interaction 830 between the various participating entities when the pricing model is implemented by SSP 804. Here, various scripts (described above) are installed in the website of the advertising media company 810 to collect the necessary data when the user visits the website. Here, the SSP 804 also receives a request for an ad tag (instead of the ad server 806 described above). In response to the ad tag request, the SSP 804 then sends a bid request to the DSP 802 regarding the advertisement displayed in the ad slot for the ad tag request. The DSP 802 sends an advertisement that is displayed in the advertisement slot when the bid price and DSP 802 bid is successful. The SSP 804 evaluates the various bidding bids and forwards the advertising tags related to the advertisements in the successful bid to the client 812B. The remainder of this function, including the measurement and storage of viewable time for each displayed advertisement, is performed in a manner similar to that performed by the advertisement server 806 as a participating entity (in FIG. 8B).

  FIG. 8D shows the interaction 840 between the various participating entities when the pricing model is implemented by DSP 802. Here, various scripts (described above) are provided in the ad tag that is forwarded to the client. In response to the bid request, the DSP 802 then sends bid requests to the various advertisers. The advertiser evaluates the bid request and sends back an advertisement tag for the advertisement that is displayed in the advertisement slot (for the bid request) when it bids. The DSP 802 evaluates various received bids and forwards one of the bids (eg, the highest bid amount bid) along with the script to the SSP 804. The SSP 804 then evaluates the bids from the various DSPs 802 and sends a bid bid ad tag to the client 812 B. The remainder of this function, including the measurement and storage of viewable time for each displayed advertisement, is performed similarly to the function performed by the advertising server 806 as a participating entity (FIG. 8B).

<< Multiple advertisement placement and viewing time >>
The viewable time concept in the present invention can be extended to other aspects of online advertising, such as posting multiple advertisements to a given user with a given ad slot on a web page. Several methods are known for displaying several advertisements in a given advertisement slot. For example, one typical method is to change a banner ad every few seconds, even if the ad slot is not visible. However, in known methods, advertisement delivery does not consider whether the advertisement was actually located in the viewing portion of the web browser when displayed, and even less care is taken by the user.

The following discussion provides a way to utilize viewable time (as supporting evidence of user involvement) when displaying multiple advertisements to the user. As described above with reference to FIG. 8, a script for advertising placement, monitoring and measurement is downloaded to the user's web client (eg, web browser) along with the return advertisement tag. This tag starts measurement of the viewable time of the displayed advertisement from the time when the advertisement display is disclosed. Here, “a” represents an advertisement currently displayed to the user. The total length of viewing time t _a which is acquired by the advertisement "a" in the distribution, advertising "a" is compared with a predetermined minimum viewing time value T _a necessary in order to be considered successful delivery. If t _a > T _a , then the script requests another tag from the ad server 806 (or SSP 804 or other participating entity in the ad ecosystem that performs the impression auction) and sends the next ad a ′ Obtain. When the delivery of the advertisement “a” is finished, the next advertisement a ′ is provided to the user. However, if t _a <T _a , the method continues to display the advertisement to the user and measure the viewable time. The method repeats the above steps of measuring the achieved viewable time and requesting an advertisement replacement when the condition is met until the user's session of this web page ends.

  In some embodiments, when requesting replacement advertisements, the method requests and downloads multiple advertisements and cycles them in the user's web client. In one embodiment of this “download once”, when an advertisement is to be displayed, the display attribute in the advertisement HTML is dynamically changed from “none” to “inline”, and the advertisement display is ended (or If you want to remove the advertisement from the display), change it in reverse (ie, from “inline” to “none”).

  FIG. 9, including FIGS. 9A and 9B, shows a process overview 900, 910 that implements a method for placing a plurality of advertisements that incorporate a viewable time concept. In FIG. 9A, similar to the process described in connection with FIG. 8 and the discussion above, the web client requests multiple advertisements as replacement advertisements when conditions are met, and the displayed replacement advertisement The viewable time measured for each is sent to the logging server 808.

  In FIG. 9B, in addition to the initial advertisement request, the web client requests a replacement advertisement when the measured viewable time of the first advertisement meets a predetermined viewable time requirement. In some embodiments, the predetermined viewable time requirement is set to a value that is less than the duration of the first advertisement so that a replacement advertisement request is sent before the first advertisement finishes displaying. In response to the replacement advertisement request, SSP 804 and DSP 802 obtain the replacement advertisement displayed to the user through a second auction (the first advertisement displayed to the user in the current user session (or the first advertisement). In accordance with the previous auction)). The web client then receives a replacement advertisement ad tag that is utilized in a plurality of advertisement placement processes. The web client further sends the measured viewable time for each of the displayed replacement advertisements (and the first advertisement) to the logging server 808.

<< Ad following >>>>
As described above, the various pricing models discussed so far utilize viewable time as one of the important parameters. The viewable time depends on the user's attention (or involvement) to one or more displayed advertisements. Various participating entities in the advertising ecosystem can increase revenue under viewable time by raising the user's attention to the displayed advertisement. In one method, the user's attention to the displayed advertisement can be enhanced by ensuring that the user watches the displayed advertisement from start to finish. Furthermore, the effect on the user of viewing the entire advertisement is much more valuable to the advertiser than the effect achieved for a user who sees only a portion of the displayed advertisement.

The following discussion provides a method that can be used by participating entities in the advertising ecosystem to implement an advertising follow-up system (when possible) that displays the advertisement to the user on the web page until the advertisement is fully displayed. . In this way, when a new impression i ₀ in the middle of the ads advertising a ₀ for the "u" is created for the user "u" (and the current impression is no longer visible), new impressions i ₀ is possible If evaluated for acquisition. A new impression i ₀ can be generated for user “u” in a number of ways, including when the user moves to another web page or another part of the current web page with a new ad slot, and so on. If i ₀ is provided by an advertising publisher (or obtained through an auction), the web page generates an impression, and the provided impression i ₀ is a condition regarding the display of the advertisement “a ₀ ” (ie, , Targeting condition, advertiser request condition, etc.), the advertisement a ₀ is displayed to the user “u” from the point where the distribution of a ₀ was stopped in the previous advertisement frame where it was displayed.

Various conditions for displaying the advertisement “a ₀ ” with a new impression i ₀ include, but are not limited to, one or more of the following: (1) Appropriate in the quality, category, etc. of a ₀ for i ₀ (2) the predetermined viewable time to be achieved for the advertisement is not yet met (between the last time the viewable time was measured for “u” and the time i ₀ was created) Determined based on the difference), (3) whether the page that created i ₀ is different from the page that currently displays a ₀ in “u”, or (4) an advertisement that currently posts a ₀ Whether the slot is visible.

In addition, if you want to use the ad follow-up method, when the ad "α _0" to the required acquisition of the requested another ad slot to display conditions are satisfied, the ad "α _0" is, it is participating entity revenue Is selected to maximize. Here, α ₀ , α ₁ ,. . . , Α _n represent a series of advertisements suitable for i ₀ . Furthermore, v _j (j = 0,..., N) represents the bid price of the corresponding advertisement “a _j ” (for example by CPV), and α _j was delivered for i ₀ by p _j Sometimes it represents the probability that the full bid price will be acquired (ie, CPV is fully realized by meeting the required minimum viewable time). Furthermore, let α1 (α _j ) be the probability that the user “u” will create another impression i _I (a _j can be posted within time I from the advertisement distribution for i _{0 in} this impression).

In this method, the score S (α _j ) of each α _j is calculated by the following equation (2):
S (α _j ) = v _j · (P _j + (1−P _j ) · α ₁ (α _j )) (2)

The highest-scoring advertisement a _j represents the highest possibility of realizing the full value of the bid amount by satisfying all the display conditions required for the advertisement a _j . Therefore, the advertisement “α” to be delivered to the user is α = argmaxS (α _j ). These probabilities can be calculated using any suitable method used with the pricing model described above. For example, in the case of CPV, this probability is determined as a function of the predetermined viewable times to be achieved for advertisement a _j and the history of impressions that achieve those viewable times. Of course, this method of calculating scores is an example, and when evaluating the best advertisements to serve to the user, participating entities can use other machine learning methods to calculate the score for each advertisement.

  As described above, the advertisement follow-up method facilitates completion of advertisement placement even when the user moves to another page during distribution. FIG. 10A shows a process diagram 1000 for a use case where one advertisement is delivered to two pages. FIG. 10B shows a use case where one advertisement delivered in the first advertisement slot (also referred to as “advertisement rectangle”) is later handled in the second advertisement slot when the user scrolls away from the first advertisement slot. Another sequence diagram 1010 is shown. Here, the advertisement server determines which advertisement to display to the user when it receives an advertisement request for the user (in the form of an advertisement tag request). The ad server evaluates the previously placed ad and decides whether to provide the user with a new ad or continue to display the current ad in the second ad slot (for ad tag request) (monitor previous ad display). Use playback time information)

  Also, in process 1010, if the advertisement displayed in the first advertisement slot is continued in the second advertisement slot, the measured viewable time in the first advertisement slot and the second advertisement slot is achieved for the displayed advertisement. Represents the total viewable time. The rest of the process for monitoring and recording the viewable time is the same as the process described in FIG. In the method and process described above, the identity of the user is managed by a logging server that stores the user ID stored in the browser cookie, or it is provided by the function sessionStorage provided by JavaScript. Managed. However, we can also use some other method to verify the user's identity, such as the unique user_id number given by the RTB in each ad bid request.

<< Example of CPV model >>
The following discussion provides a simple analysis of CPV pricing. In this example, the following conditions are assumed for consideration: (1) When a viewable time of less than 1 second is achieved in one advertisement placement, a viewable time of zero seconds is recorded. p ₀ represents the probability that an impression with a viewable time of less than 1 second is created.

When a viewable time of k seconds (for example, k−0.5 to k + 0.5 seconds) is obtained, the viewable time of k seconds is achieved. p _k represents the probability that the viewing time of the impression of about k seconds is created. Also assume that each page has only one ad slot and the user moves from page to page. Furthermore, it is assumed that an advertisement follow can be performed each time the user moves to another page.

Under these conditions, the actual value of p _k is determined using Equation (3) shown below:

  FIG. 11 shows the probability of obtaining the number of impressions 1100 necessary to obtain viewable time exceeding “T” seconds. The distribution was plotted for T = 15 (1102) and 20 (1104). If T = 15, the average of the distribution is 2.93, and if T = 30, it is 4.88. In practice, there may be cases where advertising follow-up cannot be performed. Based on distribution 1100, for T = 15, CPV is more than 3 times higher than simple CPM and CPV revenue is more than 5 times higher.

<< Ad targeting >>
The various pricing models described so far use advertisement viewing time as one of the main parameters. The advertisement viewing time depends on the user's attention (or engagement) with respect to the one or more displayed advertisements. Various participants in the advertisement ecosystem can increase the realized profit under advertisement viewing time by increasing the user's attention to the displayed advertisement. In one example, posting an advertisement for a given user can increase the user's involvement. That is, by distributing an advertisement that is likely to be of interest to the user, the user's attention to the distributed advertisement is increased.

  The following discussion provides an explanation of ad targeting methods that can be performed by one or more participants in the advertising ecosystem to increase user engagement with displayed advertisements. In this embodiment of the advertising targeting method, the priority (or order) of advertisements delivered to the user can be determined based on information about the user such as user interest, age, gender, and the like. In this case, the prioritization of the advertisement for increasing the user's attention is referred to as targeting.

  Further, in an embodiment, the advertising targeting method may limit the number of times a particular advertisement is shown to a user (also referred to as the “frequency limit” for a given advertisement) and, instead, By increasing the user's involvement in the display advertisement by showing an advertisement that has not been shown until or shown only a small number of times (ie, less than a predetermined number). Also, by limiting the number of times a particular ad is shown to the same user, a limited budget ad campaign will not show the ad's ad for that small subset of users repeatedly, thereby increasing the number of viewers. Can be reached. In an embodiment, the number of times a particular advertisement can be shown to a given user (ie, the limit) is that the given user's viewing of a given advertisement that exceeds the limit in achieving the purpose of the given advertisement is It can be determined empirically as a limit that only has a significant impact. In an embodiment, a given user may be identified using known online user tracking techniques such as the use of cookies, IP address associations, UIDs (user IDs) included in bid requests from RTB platforms, etc. it can.

  In this embodiment of the advertising targeting method, a single occurrence of the presentation of a given advertisement at a given time for a given user is the presentation of the given advertisement at that given time for that given user. Determined as a function of advertisement viewing time achieved. In an embodiment, this ad targeting method assumes that a single occurrence of a presentation at a given time for a given user at a given time is achieved. This is a case where the advertisement viewing time measured during presentation at a given time for a given user exceeds a certain predetermined threshold. In an embodiment, this constant predetermined threshold may be defined as a percentage of the total display duration of the given advertisement at the given time. In an embodiment, the total display duration of the given advertisement at the given time is based on the length of the given advertisement (eg, a rich media advertisement such as a video advertisement).

  In an embodiment, the total display duration of the given ad at the given time is available for a given ad slot (eg, a static image banner ad) in which the given ad is displayed. Based on length of display time. For example, a video advertisement “A” with a total length of 20 seconds is displayed to a user “u” at a given time, and a certain predetermined threshold for evaluating the display of the advertisement “A” for user “u” is Assume 40%. Therefore, when the advertisement viewing time (described above) measured from the display of “A” for the user “u” at the given time exceeds at least 8 seconds, the display of the advertisement “A” for the user “u” It is considered that the display of “A” for the user “u” is a single appearance (also called “complete viewing”).

  In this embodiment of the advertising targeting method, a single occurrence of a presentation for a given user for a given advertisement is achieved in the presentation to the given user for the given advertisement over one or more display opportunities. Determined as a function of advertisement viewing time. In the above example, the user “u” has seen the advertisement “A” for a certain number of times (ie, the measured advertisement viewing time), but each time it is considered a complete viewing of the advertisement “A” by the user “u”. If it is shorter than 8 seconds, each viewing of the advertisement “A” over this certain number of times, each less than the threshold of 8 seconds, can be considered as a single appearance of the complete viewing by the user “u” of the advertisement “A”. In the embodiment, this certain number of times considered to be equivalent to “complete viewing” is such that the certain number of advertisements “A” viewing has the same effect as the case of “complete viewing” of advertisement “A”. Can be determined.

This targeting method will be further described using the following example. When the user “u” accesses the web page “m”, an impression is generated in the advertisement slot F _mj existing in the web page “m”. The impression represents the opportunity for the advertiser to display the advertisement to the user in slot F _{mj when the} user is viewing the web page “m”.

When bidding for an impression, various participants in this advertising ecosystem, such as third-party advertising server 806, DSP 802, etc., can take one or more of the available advertisements that can be displayed to user “u”. Targeting is performed by determining a relevance score for a particular advertisement and determining a display priority for one or more possible advertisements based on the determined ad relevance score. Elements relating to “a”, “u”, F _mj etc. used in targeting include, but are not limited to:

The various factors related to a given ad “a” that are analyzed during ad relevance score determination include: (1) degree of achievement of the campaign to which ad “a” belongs, (2) includes “a” The remaining budget of the campaign, (3) the desired bid amount of the campaign including “a”, (4) the creative quality of the advertisement “a”, and (5) the advertisement “a” is a candidate for continuous distribution (detailed later). (6) Whether the advertisement “a” is delivered to another advertisement slot F _ml somewhere on the web page “m”.

  Various factors relating to user “u” analyzed during the ad relevance score determination include: (1) Geographic location of user “u”, (2) IP address for user “u” Internet connection, (3) Cookie information from the user's computer device, (4) User browser or OS type, (5) User language, (6) Preceding URL (URL of website visited just before “m”).

The various factors related to ad slot “F _mj ” and web page “m” that are analyzed during ad relevance score determination include: (1) Estimated price for the ad slot (CPM, CPS, etc.), ( 2) Average performance of the ad slot (click-through rate, conversion rate, etc.), (3) Average ad viewing time, (4) Impressions per unit time, (5) Average user dwell time, (6) Ad Slot dimensions (height and width), (7) URL of web page “m”, (8) Type of “m”, (9) Whether “m” can acquire other advertising slots F _ml (Enables ad follow and other strategies described below).

In this advertisement targeting method, the advertisement relevance score for displaying each considered advertisement in the slot F _mj is represented by “S”. Here, “S” of each considered advertisement is B (a, t) (the bid price of “a” at time t when no advertising targeting is applied), T (a, u, F _mj ) (targeting effect) Bid price premium), c (u, a) (number of times user “u” has completely viewed advertisement “a”), C _a (number of times that advertisement “a” can be shown to a given user) It is determined as a function of (limit) (formula (4) shown below).

Where 0 <λ <1 is a constant that determines the weight of the parameter in equation (4), and 1 _{c (u, a) ≦ Ca} is c (u, a)> Ca (ie, achieved) The function is set to 0 when the number of complete viewing is less than the frequency limit, and set to 1 in other cases.

により表すこととする。広告「ａ」の達成された目標および期待される目標を使用して、時刻「ｔ」における広告「ａ」の入札金額Ｂ（ａ，ｔ）を計算する。時刻「ｔ」において、広告「ａ」の実際の達成ｒ_ａ（ｔ）と期待達成

を比較し、以下に示す式（５）を使用してＢ（ａ，ｔ）を計算する。

In one embodiment, B (a, t) is estimated as a function of the bid amount that the advertiser on “a” is willing to pay and the current campaign goal (of the advertiser on that ad “a”) performance progress. The In general, an advertising campaign goal is generally defined by certain metrics such as the total number of complete views of various advertisements belonging to the advertising campaign, the total advertising viewing time achieved by all advertisements in the campaign, and the like. The achievement of the campaign corresponding to the advertisement “a” at time “t” is represented by 0 <r _a (t) <1, and the achievement of the campaign at time “t” is achieved.

It shall be represented by The bid amount B (a, t) of advertisement “a” at time “t” is calculated using the achieved and expected goals of advertisement “a”. At time “t”, the actual achievement r _a (t) and the expected achievement of the advertisement “a”

And B (a, t) is calculated using the following equation (5).

  Where γ is a constant, F: (0, ∞) → (0, 1) is any suitable monotonically increasing function, and B (a, t−1) is the time “t−1”. Is the previous bid amount (ie, the last calculated bid amount) calculated for the advertisement “a” using equation (5). As a result of this formula, if the advertising campaign (related to the advertisement “a”) exceeds its target at time “t”, the bid amount B (a, t) decreases and the advertising campaign is time “t”. Increases if the target is not achieved.

  In one embodiment, “T” is estimated as the probability that the delivery of advertisement “a” to user “u” will achieve full viewing. Here, a higher probability indicates that the user is more involved in the advertisement “a”. By including “T” as an element in the evaluation of the advertisement relevance score “S”, the user's higher involvement in “a” is associated with the relevance score “S” for the advertisement “a” compared to the remaining considered advertisements. To increase. As a result, when the relevance score “S” of the advertisement “a” is higher than that of the other consideration-targeted advertisements, even if the bid amount B (a, t) is lower than that of the other consideration-targeted advertisements, the advertisement “a "Is displayed to the user" u ".

に設定する。ただし、

は、以下に示す式（６）を使用して決定することができる。

In the embodiment, let p (a, u) represent the probability that the distribution of the advertisement “a” to the user “u” will achieve complete viewing. That is, p (a, u) is a probability that the duration of the measured advertisement viewing time of the advertisement “a” is equal to or greater than a predetermined threshold value regarding the advertisement “a”. Further, let X _umj represent the feature (ie, attribute) vector of user “u” and advertisement slot F _mj , and β _a represent the feature vector of advertisement “a”. Where T (a, u, F _mj ) is

Set to. However,

Can be determined using equation (6) below.

However, _C0 is a certain constant. In equation (6), the overlap (ie, match) between the features of X _umj (ie, the features of user “u” and ad slot F _mj ) and the features of β _a (ie, the feature vector of “a”) is A higher probability value p (a,) indicates that the larger the greater the match for the user “u” of the advertisement “a” displayed in the ad slot F _mj, and the more likely that the user ’s participation in the advertisement “a” will be achieved. u). Additional details of methods of ad targeting that can be attribute matched between users and available ads (which can be used with other ad targeting methods disclosed in this application) are described, for example, in commonly assigned US Patent Application No. 13 / 605,915.

In an embodiment, an ad relevance score “S” indicating whether a given ad, for example ad “a”, has exceeded the allowed number of full viewings of ad “a” (ie, the frequency limit of ad “a”). The element 1 _{c (u, a) ≦ Ca} in can be determined as a function of max (0,1c (u, a) / C _a ). The ad relevance score “S” drops to zero when the ad “a” has reached its relevance frequency limit, so that the ad relevance score has not yet reached the frequency limit and is greater than zero. Allows other advertisements with score “S” to be ranked higher than advertisement “a” to be displayed to user “u”. On the other hand, if the advertisement “a” has not achieved its associated frequency limit, the score “S” increases in proportion to the remaining number of complete viewings of the advertisement “a” (if the other elements are the same) B) allowing ads that have achieved fewer full views to have a relevance score “S” greater than ad “a” and a higher display priority.

人のユーザのいずれかについて所与の広告の完全視聴を達成する確率（ｐ）をｐ＝０．０５に設定する。また、所与の広告のＣ_ａ（すなわち、最大完全視聴回数）をＣ_ａ＝１に設定する。これは、頻度限界に到達するために当該広告のただ１回の完全視聴を必要とする。 The following example illustrates changes in the maximum number of complete views associated with an advertisement and the resulting ad campaign arrival (eg, the number of unique viewers reached by an ad campaign for a given budget). “A” and “B” represent two advertisements that can be posted to the user visiting the website “m”. N _u represents the unique number of users who visit the website “m” twice. Opportunity to display ads on each of the N _u of users is that there is at least 2 times. Assume that “A” and “B” have the same cost criterion C _B. In embodiments, this cost criterion C _B can be determined as a function of the estimated bid amount for each advertisement. In embodiments, this cost criterion C _B can be determined as a function of the bid price premium based on the effect of the estimated bid amount and targeting of the advertisements. Furthermore, on the website “m”

The probability (p) of achieving full viewing of a given advertisement for any of the users is set to p = 0.05. Also, set C _a (ie, the maximum number of complete viewing times) of _a given advertisement to C _a = 1. This requires only one complete viewing of the advertisement in order to reach the frequency limit.

Based on the above values, advertising platform server, such as third-party ad server 806, DSP 802, which performs ad operations in website "m" using the ad targeting methods are provided for each N _u users Estimate each relevance score “S” for each potential advertisement and post the highest scoring advertisement at a given time. As described above, the relevance score “S” of a given advertisement can be estimated using the following equation (7).
S (u, a) = max (0, (1-c (u, a) / C _a )). C _B (7)

  When the advertisement relevance score “S” becomes the same score, each advertisement having the same score may be displayed with an equal probability. To estimate the effect of implementing a frequency limit, assume that the advertising platform implements a random advertisement posting scenario. In this case, no frequency limit is set and any given advertisement can be displayed with a probability equal to a given user at a given time. In another scenario, it is assumed that the advertisement platform implements a distribution method based on an advertisement relevance score including a frequency limit.

  Table # 2 1205 shown in FIG. 12C shows the distribution of the number of times that “A” and “B” were each viewed completely by a given user who visited the website “m”. As shown in Table # 2 1205, when neither advertisement “A” nor “B” achieves complete viewing (ie, (0, 0)), both advertisement distribution schemes are advertisements “A” and “B”. Have the same distribution. That is, when p = 0.05 (probability of achieving full viewing for a given advertisement (p)), the probability of not achieving full viewing for a given advertisement is 1-p = 0.95. Therefore, the probability that neither advertisement “A” nor “B” achieves complete viewing is 0.95 * 0.95 = 0.9025. Similarly, if either advertisement “A” or “B” achieves full viewing but the other does not (ie, (0,1) and (1,0)), “A” and “B” have the same distribution. This is understandable because the advertising platform delivers random advertisements in any scenario. That is, for delivery based on “S”, the ad relevance scores for both ads “A” and “B” are either (0, 1) or if one of these ads is randomly picked and delivered. The score is the same until (1,0) is achieved.

Furthermore, it is assumed that the frequency limit C _a is set to 1 under the method based on the advertisement relevance score. Since neither “A” nor “B” will be posted more than once, either of the advertisements “A” or “B” will be viewed twice (ie (0,2) or ( 2,0) is zero, and finally, under a method based on the ad relevance score, suppose frequency limit C _a is set to 1. Advertising within the allowed frequency limit The probability (0.05 * (1 * 0.05) = 0.0025) that both “A” and “B” achieve complete viewing (ie, (1,1)) Higher than the probability of achieving (0.05 * (0.5 * 0.05) = 0.00125), which is one of the given ads, eg “A”, under a scheme based on the ad relevance score Will be posted to the user on the next visit after the frequency limit is reached This is because the advertisements that are played are the remaining advertisements (ie, advertisement “B”), and the probability that both advertisements “A” and “B” achieve full viewing is simply a function of p. By implementing an ad delivery system that includes frequency limits, the advertising platform can expand the reach of a given campaign for a given advertising budget.

  FIG. 12A shows a method 1200 that can be utilized to deliver advertisements based on a delivery scheme that incorporates advertising targeting and frequency limits by an advertising platform (eg, third party ad server 806, DSP 802, etc.). At step 1210 of method 1200, an advertisement placement request is received by an advertisement platform. Here, this request can be in the form of an RTB bid request without limitation. At step 1220, a relevance score for each of one or more available advertisements in the advertisement inventory is determined. In an embodiment, an ad relevance score is determined for a subset of ads selected from among the available ads in the ad inventory. Here, the subset of advertisements can be selected based on various criteria such as whether or not the bid amount for the available advertisement is within a predetermined range of the suggested bid price in the received bid request. At step 1230, the advertisement associated with the highest advertisement relevance score is selected. At step 1240, an ad tag for the advertisement (which can be used to retrieve an ad creative from the repository) and a bid amount for displaying the advertisement for the ad tag are returned as part of the bid response.

  FIG. 12B illustrates a method 1250 that can be utilized by an advertising platform (eg, third party ad server 806, DSP 802, etc.) to estimate a relevance score for a given advertisement. This relevance score can then be used to prioritize available advertisements that can be displayed to the user or given in a bid response. In step 1260 of method 1250, a bid amount that the advertiser is willing to pay for display of the given advertisement is determined. In the embodiment, the bid price is increased or decreased based on the performance of the campaign to which the given advertisement belongs. In step 1270, a bid price premium based on the effect of targeting is determined. In an embodiment, the bid price premium is determined as a function of the probability of achieving full viewing of the given advertisement when the given advertisement is displayed to the user. This probability increases when the user's interest (ie, attribute) shows a high degree of match to the content of the advertisement (ie, the attribute of the advertised content), and the user's interest is less Decrease when showing a match.

  In step 1280, a display limit score is determined that determines the degree of complete viewing achieved by the given advertisement. In an embodiment, the display limit score is determined as a function of the percentage of complete viewing achieved for the given advertisement associated with the user and the frequency limit for the given advertisement. In an embodiment, the display limit score is set to zero if the full viewing achieved for the given advertisement involving the user is equal to the frequency limit for the given advertisement. In step 1290, an ad relevance score is determined as a function of the determined bid amount (step 1260), the determined bid price premium (1270) and the display limit score (step 1280). In the embodiment, the advertisement relevance score based on the above elements can be determined using Equation (4).

  FIG. 13, including FIGS. 13A, 13B, 13C, and 13D, is an illustrative example of various methods that can be used to implement an advertising targeting scheme by various participants in the advertising ecosystem.

  FIG. 13A shows an interaction 1300 between various participants in the advertising ecosystem when the advertising targeting scheme is implemented by a third party advertising server 806. Here, the web client (ie, web browser) of client 812B initiates a page request to advertising media company 810. Advertising publisher 810 provides the requested page to client 812B. Client 812B then sends a request to advertisement server 806 for one or more advertisement slot advertisement tags (present on the page received from advertisement publisher 810). The ad server 806 that executes the ad targeting scheme is configured to obtain the highest ad relevance score from the available ad inventory that is managed (directly or indirectly) by the ad server 806 for each of the ad slots available for ad display. Determine ad tags. In an embodiment, the advertisement server 806 performs the targeting method 1300 (described in connection with FIG. 13A) to perform the advertisement targeting scheme.

  The ad server 806 then returns the determined ad tags (which are related to the ad displayed in the ad slot) along with scripts that perform various functions such as monitoring user behavior, measuring user attention, etc. To do. Client 812B sends the received advertising tags to the CDN to request advertising content associated with those advertising tags. The CDN responds to this request by sending advertising content for those advertising tags to the client 812B. The client 812B displays the received advertisement content to the user. Further, the client 812B measures the advertisement viewing time of the displayed advertisement (when various conditions regarding user attention are satisfied as described above). When the display of the advertising content ends, or when the user ends the session, the measured advertisement viewing time is transferred to the logging server. It can be used by other participants (including third party ad servers) to execute pricing models and determine the cost of ad delivery.

  FIG. 13B shows the interaction 1310 between the various participants when the advertising targeting scheme is performed by the DSP 802. Here, various scripts (described above) are provided in the ad tag that is forwarded to the client. In response to a bid request that seeks to provide an advertisement to a given user, the DSP 802 forwards the bid request to various advertisers. In an embodiment, the DSP 802 evaluates the advertisements of the various advertising campaigns and determines whether advertisements from the various advertising campaigns have reached their frequency limit for total complete viewing, and then the bid. Forward the request to the advertiser for the advertising campaign. If the advertisement from a given advertising campaign has reached the total full viewing frequency limit for a given user, the DSP 802 will not forward the bid request to the advertiser associated with the advertising campaign.

  The advertiser who receives the bid request from the DSP 802 evaluates the bid request and sends back the advertisement tag of the advertisement displayed in the advertisement slot (related to the bid request) when making a successful bid in addition to the bid. The DSP 802 evaluates the various bids received and forwards one of the bids along with the script to the SSP 804. In an embodiment, the DSP 802 that executes the ad targeting scheme determines an ad relevance score for each of the ads related to the various receiving bid ad tags and selects the ad tag with the highest ad relevance score. To do.

  In an embodiment, the DSP 802 performs the method 1300 (described in connection with FIG. 13A) to implement the ad targeting scheme. In the method 1300, the frequency limit for the total complete view of a given ad for a given user is incorporated into the ad relevance score estimate and the DSP 802 enforces the frequency limit after a bid is received from the advertiser. to enable. SSP 804 then evaluates bids from various DSPs (such as DSP 802) and sends the winning bid ad tag to client 812B. The rest of the functions including measurement and storage of advertisement viewing time for each displayed advertisement is performed in the same manner as the function performed by the advertisement server 806 as a participant (in FIG. 13A).

  FIG. 13C illustrates various participations when the DSP 802 executing the advertising targeting scheme returns to the SSP 804 a bid with a bid amount lower than the lowest auction price that the advertising publisher 810 is willing to accept display of the advertisement from the DSP. Interaction 1330 between is shown. Here, various scripts (described above) are installed in the advertising company's 810 website to collect the necessary data when the user visits the website. Also here, the SSP 804 receives a request for an ad tag (instead of the ad server 806 described above). In response to the ad tag request, SSP 804 sends a bid request to DSP 802 to cause the advertisement to be displayed in the ad slot for that ad tag request. The DSP 802 performs advertising targeting and sends the bid price and the advertisement (based on its advertisement relevance score) to be displayed in the advertisement slot if the DSP 802 bid is successful. SSP 804 evaluates the various return bids and advertisements in winning bids (which have the highest bid amount among the bids received and have a higher bid amount than the advertising agency's 810 minimum auction price). The advertisement tag related to is transferred to the client 812B. The rest of the functionality, including searching for advertisements associated with the ad tag, displaying the searched advertisements, measuring and storing the advertisement viewing time for each displayed advertisement, is performed by the ad server 806 as a participant (in FIG. 13A). ) Performed in the same way as the function.

  If any of the bids received do not exceed the minimum auction price of advertising publisher 810, SSP 804 sends the tag of the advertising network or advertising exchange (eg, advertising server 806) to client 812B. Client 812B then distinguishes the tag of the ad network (or ad exchange) from the ad tag associated with the winning bid and sends an ad tag request to the ad network associated with the tag received from SSP 804. The advertisement network that has received the advertisement tag request returns an advertisement tag that satisfies the minimum auction price of the advertisement publisher 810 (where such information is included in the advertisement request, if possible). The rest of the functionality, including searching for advertisements associated with the ad tag, displaying the searched advertisements, measuring and storing the advertisement viewing time for each displayed advertisement, is performed by the ad server 806 as a participant (in FIG. 13A). ) Performed in the same way as the function.

  FIG. 13D shows that the SSP 804 receives a bid from the client when the bid amount received from the DSP (such as DSP 802) is lower than the minimum auction price that the advertising publisher 810 is willing to accept to display the advertisement from the DSP. A method 1140 for operating advertisement distribution in response to an advertisement tag request is shown. In step 1350, client 812B sends an ad tag request to SSP 804 by notifying impression generation. The advertisement in this case can be posted / displayed to the user via the client 812B. In an embodiment, when reporting impression generation, client 812B performs pixel monitoring around a given ad slot (described above in connection with FIGS. 4A-4D). In step 1355, in response to the impression generation notification, the SSP 804 sends a bid request to the DSP to display the advertisement in the advertisement slot associated with the generated impression. In step 1360, the SSP 806 receives a bid response from the DSP and performs bid conversion (if necessary) to convert the bid amount associated with the received bid to a common cost criterion (CMP, CPC, etc.). The bid responses received can be easily compared.

  In step 1365, SSP 806 determines that the highest bid amount (at a common cost basis) is the advertising publisher 810 (whose web page provides the ad slot associated with the impression) the ad associated with the highest bid. Determine if it is higher than the lowest auction price you are willing to accept to display. If the highest bid is higher than the lowest auction price, in step 1375, the SSP 806 displays to the user the advertisement (or symbolic name of the advertisement, such as an advertisement tag) to be displayed to the user. To the client 812B. If the highest bid amount is lower than the lowest auction price, then in step 1370, SSP 806 notifies the other ad network or ad exchange that this impression has been generated and the ad network or ad exchange is in the advertising media company 810. Allows you to identify advertisements that exceed the minimum auction price. In step 1380, the advertising network or advertising exchange sends an advertisement (or symbolic name of the advertisement, such as an advertisement tag) to be displayed to the user to the client 812B if the minimum auction price is met.

<< Measurement of effectiveness of advertising campaign >>
In the present invention, the effect of an advertising campaign (which includes one or more advertisements displayed to one or more users on a web page over a period of time) to the user is the advertising campaign displayed to those users. Can be measured by using the measured advertisement viewing time for each of the one or more advertisements. One common metric used to measure effectiveness in the marketing industry is the total audience rating (GRP) achieved in television advertising campaigns. An Online Advertising Equivalent GRP (eGRP) metric that allows marketers (and advertisers) to easily compare and evaluate the value and effectiveness of advertising campaigns is very beneficial. The following discussion shows a brief background on calculating GRP and how to calculate eGRP using measured advertisement viewing times for various advertisements in an advertising campaign.

As mentioned above, GRP is an effective metric used in TV advertisements, defined by the frequency of broadcast advertisements and the arrival of distribution (the ratio of the total number of viewers to the target size of the advertisement). Is considered. The GRP can be calculated using the following equation (8).

  Here, R is the total number of potential targets of the advertisement, r is the number of target viewers realized, and f is the frequency of distribution.

In the embodiment, when the advertisement viewing time of the advertisement presented to the user group is measurable (at least for an important subset), the concept of the advertisement viewing time is incorporated into GRP, and the concept and the formula shown below (9) can be used to calculate the GRP for web content c (such as a web page).

However, the time length of the involvement of when R _c is a collection of opportunities for the delivery of content c for potential target of the number of people, V _u is a user u of the content c, t _v is the user u is viewing the content c in the opportunity v (Advertisement viewing time), r (t _v ) is the (normalized) effect of the content when the content is viewed until time t. Here, 0 <r (t _v ) <1 (ie, weight).

In an embodiment, r (t _v ) is whether a given display of a given advertisement has resulted in a CPV cost (ie, that given advertisement is viewed by a given user for at least a predetermined advertisement viewing time. Can be determined as a function). In one example, r (t _v ) can be set as r (t _v ) = 1 _{(tv> = LCPV)} . However, r (t _v ) is set to 1 when t _v > = L _CPV is satisfied, and is set to 0 when t _v is not satisfied. Further, L _CPV is the shortest total ad viewing time (ie, predetermined ad viewing time) that must be achieved before CPV costs can be collected for delivery of a given ad in the CPV model. In an embodiment, if an advertiser uses a TV advertisement of length L _{TV as} an online advertisement and a cost is incurred when a portion of the online advertisement is fully viewed (ie, achieved L _CPV ), then L _CPV = f (L _TV ). In an embodiment, if an advertiser uses a TV advertisement of length L _TV as an online advertisement and costs only when the entire advertisement is fully watched (ie, achieves L _CPV ), L _CPV = L _TV .

Also, the modified version of the GRP metric can be compared with the TV advertisement GRP metric. In an embodiment, the modified effective GRP (eGRP) can be derived using equation (10) shown below.

Where M is a collection of media to which content (eg, advertisements, advertiser-owned web pages, etc.) is distributed, U _m is a collection of users who visit media m, and V _u is distributed to user u. A collection of measured advertisement viewing times for each distribution of content, where R (M) is the total achievement of M.

  Furthermore, w (m, u) is a function of (m, u), and its value determines the weight of (m, u). Since online advertising is different from TV advertising, the true effect of online advertising on the user increases accordingly. In one embodiment, by setting w (m, u) = 1 (the weight that increases the true effectiveness of the online advertising campaign), the value of online advertising delivery is the same as the price of the TV advertising campaign. It is regarded. On the other hand, by setting w (m, u) = 0 when the user u visits the web page m by clicking on an online advertisement, at least a part of overestimation of the reach of the advertisement is avoided. (By not counting the number of advertisements of users who arrive via the advertisement and the viewing of related web pages twice).

  In an embodiment, w (m, u) is determined as a function of the ratio of the total area of ad slots that deliver advertisements to the user to the area of the user's browser. In an embodiment, w (m, u) is determined as a function of the ratio of the total viewable area of the advertisement slots that deliver advertisements to the user to the viewable area of the user's browser. The w (m, u) determined in this way is less valuable for a single delivery of an online advertisement than a single delivery of a TV advertisement when the advertisement on the TV occupies most of the display area. It corresponds to the idea.

In an embodiment where w (m, u) = 1 (online advertising distribution is equivalent to TV advertising distribution) and r (t _v ) = 1 (as described above), the estimated eGRP Has the following strengths: (1) The eGRP increases only when the advertisement is completely viewed. In this case, like a video advertisement such as a TV advertisement, the intended goal of the advertisement is probably achieved only if it is fully viewed. (2) If the same (or similar) creatives are used in both TV and online campaigns, the GRP of eGRP and TV advertising campaigns can be directly compared (number of CPV occurrences and arrival of TV ads, respectively) . The use of eGRP metrics allows for easy comparison with GRP and facilitates mass marketing, including online advertising as part of one unified marketing strategy. FIG. 14A, described below, is a visual representation of a comparison between eGRP and GRP for an advertising campaign.

In addition, for a given bid price (based on any of the stated pricing models), an achievable eGRP can be estimated. In an embodiment, the eGRP achievable for CPX-based bidding for “b” prices from advertisements can be determined using equation (11) shown below.

Where T is a content targeting attribute, and M (T) and U _m (T) are collections of media and users having specified attributes, respectively. Further, W (u, m, b) is a probability that the impression of the attribute (m, u) can be awarded by means of the bid price “b”, and f _m (u) is an estimated value of the number of times the user u visits m. , 0 ≦ r ^ (m, u) ≦ 1 is an estimation effect of one-time advertisement distribution on (m, u) (that is, the effect of online media m on user u). By entering the arrival of the TV advertisement campaign into R (M (T)), the GRP of the TV advertisement can be compared with that of the online advertisement, eGRP.

<< EgRP estimation for a given advertising campaign >>
In the campaign C eGRP calculation, it is necessary to specify the actual format of w and r. Equation (11) used to calculate eGRP shows that it can be calculated by giving an estimated size of M reached R (M). Furthermore, effective arrival r _e (m, u, v) = (m, u, v). r (tv) (value obtained by visit v to media m by user u). In addition, when the arrival of the TV advertisement is compared with the online advertisement, the arrival R _{TV of the} TV advertisement can be used as R (M). Here, the arrival of the TV advertisement includes, but is not limited to: (1) the number of households having an opportunity to be exposed to the TV advertisement; (2) the exposure to the TV advertisement estimated from the estimated number of households. The number of people who have the opportunity to be.

において配信される。 Here, tv = L _tv for r (tv). However, L _TV is the length of the TV advertisement. Here both R _TV and L _TV are given by the advertiser of campaign C. Finally, the ratio of the total area of the ad slots that deliver the advertisement for C to the area of the user's browser (which effectively captures the user's split attention in proportion to the relative display area of the advertising content) Set w (m; u). Furthermore, r (tv) can be limited to a number less than 1 (to ensure a conservative estimate of the effectiveness of an advertising campaign), but may violate this limit (ie, r ( When tv)> 1, r (tv) = 1). Using the above assumptions and values, in one example, eGRP can be determined using the following method. In this way, for user u visiting web page m, the advertisement of campaign C is an ad slot (also called a frame).

Delivered in

の面積の合計とする。また、Ｓをもって時間間隔（ｔ；ｔ＋ｉ）にわたるＳ（ｔ）の平均サイズとする。ただし、ｉは、前もって決定されるある定数である。すなわち、Ｓは、次式（１２）のとおり定義できる。

Here, let S _b (t) be the viewable area of the user's browser for the user at time t. Further, an advertisement frame that can be viewed at time t with S (t)

The total area. Also, let S be the average size of S (t) over the time interval (t; t + i). However, i is a certain constant determined in advance. That is, S can be defined as in the following formula (12).

Furthermore, since there is a rich media having an advertisement frame that can be enlarged by a user's operation of placing the mouse on the advertisement, this average viewable area is considered. Here, if the average time interval is sufficiently small, the value of S can be expected to be the same as S (t) (the averaging process is omitted). Similarly, the average browser size S _b (t) can be calculated and represented by S _B. The effective advertisement viewing time V _e (t) at time t = 0 is set to zero (V _e (0) = 0). V _e (t) is periodically updated by applying the following equation (13).

の少なくとも１つが間隔（ｔ；ｔ＋１）中に視聴可能状態にあると仮定する。さらに、表示されている広告の広告視聴時間の測定中にユーザが現在表示されている広告「ｍ」を通過してスクロールしたとき、広告「ｍ」がユーザのブラウザの視聴可能領域外にある場合、測定は中断される。当該広告キャンペーンに属する別の広告「ｍ_ａ」が当該ユーザのブラウザの視聴可能領域に入った場合、視聴可能領域にあるこの新しい広告の広告視聴時間の測定がＶ_ｅ（ｍ_ａ，ｕ，ｖ）として開始される。 here,

Is in a viewable state during interval (t; t + 1). Furthermore, when the user scrolls past the currently displayed advertisement “m” while measuring the advertisement viewing time of the displayed advertisement, the advertisement “m” is outside the viewable area of the user's browser. The measurement is interrupted. When another advertisement “m _a ” belonging to the advertisement campaign enters the viewable area of the user's browser, the measurement of the advertisement viewing time of this new advertisement in the viewable area is V _e (m _a , u, v ).

に配信される広告の１つをクリックすることにより所有メディアｍ_ａに移動したとき、ｍに関するＶ_ｅ（ｍ，ｕ，ｖ）の測定は終了し、Ｖ_ｅ（ｍ_ａ，ｕ，ｖ）の測定が始まる。また、Ｖ_ｅ（ｍ_ａ，ｕ，ｖ）は，ＯＭＶＴ（前述した所有メディア広告視聴時間）に基づいて更新される。ｍ_ａはブラウザ全体を占有するので、ＯＭＶＴの効果は、低く重み付けされず、Ｖ_ｅはＯＭＶＴと同じ比率で増大する。ｕのオンライン広告視聴時間Ｖ_ｅ（ｍ，ｕ，ｖ）またはＶ_ｅ（ｍ_ａ，ｕ，ｖ）は、ｕがｍまたはｍ_ａから立ち去ったとき（およびｕがセッションを終了したとき）に、集計される。さらに、ＧＲＰと異なり、純粋に広告視聴時間に基づくｅＧＲＰは、同一広告を同一ユーザに複数回表示する低減効果を捕捉する。たとえば、所与の広告が当該ユーザに表示される都度評価される広告視聴時間は、当該広告が当該ユーザに再表示されるのにつれて低減し、かつ、当該ユーザは当該広告の再表示の都度ますますより少ない注意を払うようになる。 Similarly, the user

When moved to own the media _{m a} by clicking on one of the ads served to measurement of _{V e (m, u, v} ) relates to m is _{completed, V e (m a, u} , v) of Measurement starts. Also, V _e (m _a , u, v) is updated based on OMVT (owned media advertisement viewing time described above). Since m _a occupies the entire browser, the effect of OMVT are not weighted low, _{V e} is increased in the same proportion as OMVT. u's online ad viewing time V _e (m, u, v) or V _e (m _a , u, v) is when u leaves m or m _a (and u ends the session) Aggregated. Further, unlike GRP, eGRP purely based on advertisement viewing time captures the reduction effect of displaying the same advertisement multiple times to the same user. For example, the ad viewing time that is evaluated each time a given ad is displayed to the user decreases as the ad is redisplayed to the user, and the user reappears each time the ad is redisplayed Less attention will be paid.

  However, in GRP, such repeated display of advertisements is still counted as complete viewing by the user. That is, the GRP estimated for the TV advertisement campaign overestimates the GRP by counting all viewers (viewing or dozing) in front of the TV as viewing the advertisement. Furthermore, as described above, in the online advertisement, the later distribution of the same advertisement obtains a short advertisement viewing time when the user has already viewed the advertisement in advance.

Therefore, to compensate for eGRP and to make it comparable to (overestimated) GRP, element P _r . L _TV is added. In one embodiment, each time a user views an ad after the initial display of a given ad in campaign C, the value P _r . L _TV (0 ≦ Pr ≦ 1) is added to the advertisement viewing time V _e (m, u, v) for a given advertisement.

For example, if Pr = 0.5, it can be interpreted as follows. If the advertisement is delivered to the user twice, one delivery gets the user's attention and the advertisement is completely consumed. In the rest of the distribution, the user concentrates on content other than the advertisement, and the advertising effect is probably proportional to the advertisement viewing time (which can be further weighted according to the relative size of the advertisement to the browser). Therefore, eGRP can be defined as shown in the following equation (14) using V _e (m, u, v) obtained above.

That, EGRP is the acquired total online advertisement view time TV ads reach and TV advertising length divided by the product of the (plus additional points, p _r times the total number of impressions of a given ad) equal. If the advertisement length of the online advertisement is the same as the length of the TV advertisement, this metric is a comparison between the TV reach and the total number of viewers of the online advertisement. Even if the lengths are different, GRP and eGRP can be compared by multiplying by an appropriate factor (proportional to different lengths) in eGRP. Further, for campaigns that include multiple advertisements, in one embodiment, the eGRP can be calculated separately for each advertisement in the advertisement campaign using equation (14), and the cumulative eGRP for all of these advertisements. Note that represents the total eGRP for the advertising campaign.

  FIG. 14A shows a comparison 1400 between the GRP for a TV advertising campaign and the eGRP for an online advertising campaign. The upper graph 1402 shows the GRP achieved from the TV advertising campaign for a given market. The lower graph 1404 shows the eGRP achieved from the online advertising campaign. The horizontal axis in each graph corresponds to the number of unique targets available (ie, the number of people available in the given market that can be targeted by the advertising campaign), and the vertical axis represents the number of available targets. Represents the effective number of times each person has been exposed to the advertisement of the advertisement campaign. Here, the white box indicates the contribution that one advertisement delivery has made to increase the frequency of exposure to the advertisement. A single exposure for a TV advertisement of one viewer increases the frequency of exposure by 1, while a single exposure for an online advertisement increases it by 0.5 (in FIG. 14A). In graph 1404, the gray box indicates the contribution to achieving increased exposure frequency for the advertisement that should be attributed to the user's exposure length for that advertisement. In graph 1404, a black box indicates the time spent on a site owned by the advertiser. GRP measures exposure frequency on a per-family basis, while eGRP measures arrival on a per-person basis.

  FIG. 14B shows an overview of a system 1410 that can be used to estimate the eGRP that can be achieved by an advertising campaign. The system 1212 interacts with the databases 1414 and 1416 to collect information regarding the impressions used to advertise the campaign's advertisements by the advertising campaign and the advertisement viewing time achieved by each displayed advertisement. As described above, the total measured advertisement viewing time achieved by the display advertisement group and the total advertisement viewing time achieved by the advertisement campaign are used to determine the eGRP of the advertisement campaign.

  FIG. 14C shows a summary of a system 1420 that can be used to predict the eGRP that can be achieved with a given advertising cost. Additional details of such a system and the method used for eGRP prediction are described below.

  FIG. 14D is an illustrative example of a method 1435 that can be used to estimate eGRP achieved by an advertising campaign. At step 1432 of method 1435, a subset of media on which advertisements of the advertising campaign are displayed is selected. At step 1434 of the method 1435, a subset of users who have viewed at least one advertisement from an advertisement campaign in the media of the particular sub-subset described above, respectively, is selected. Users of this subset reached via this subset of media represent the total exposure target for the advertising campaign. This portion of the total target that was effectively exposed to the advertising campaign (eg, through the display of advertisements for the advertising campaign) represents the actual exposure achieved by the advertising campaign. As described above, this achieved exposure as a percentage of the total available target is used to determine the achieved eGRP for the advertising campaign.

  At step 1436 of method 1435, the advertisement viewing time measured for each advertisement viewed by each user (in the subset of users) within the subset media is determined. For example, the advertisement viewing time can be determined by accessing the logging server 808. In step 1438 of the method 1435, a scaling factor for each measured advertisement viewing time in step 1436 is determined. In one embodiment, the magnification of the measured advertisement viewing time is based on a relative size to the size of the web browser window when displaying the advertisement of the advertisement (which is the target of the measured advertisement viewing time) to the user. This magnification estimates the estimated effect of the advertisement for the user based on the display advertisement area (relative to the web browser) used to expose the user to the advertisement campaign.

  At step 1440 of method 1435, based on the advertisement viewing time determined at step 1436 and the associated scale determined at step 1438, the total achieved advertisement viewing time for the advertising campaign is determined. In one embodiment, the sum of all measured ad viewing times determined in step 1436 is combined with each measured ad viewing time multiplied by the associated multiplier determined in step 1438, and the total achieved ad viewing for that ad campaign. Determine the time. At step 1442 of method 1435, the total target that the total potential achievable ad viewing time is reachable in the advertising campaign (ie, the total number of times all users in the subset of users have been exposed to ads from the advertising campaign). And the total duration of potential exposure that can be achieved by each exposure of the ad from the ad campaign against the total reachable target (ie, the total length of the ad displayed on the media regardless of whether the user has seen it) As a function of At step 1444 of method 1435, the ad viewing time at which eGRP (a measure of overall exposure achieved for the advertising campaign) and the potential ad viewing time achievable (ie, the total duration of potential exposure achievable) are achieved. Is measured as a function of Equation (10) gives an example of an eGRP calculation function.

<< EGRP Prediction >>
As described above, eGRP can be estimated for a given bid price b of the advertiser. In one embodiment, a bid “b” in the form of CPX can be converted to an achievable eGRP using the following equation (15):

は、推定対象である（以下において詳述する）。また、＜ｕ＞は、同一ターゲティング属性Ｔのユーザを含むＵ_ｍのサブセットである。＜Ｕ_ｍ＞は、＜ｕ＞によるＵ_ｍの分割、すなわち＜ｕ＞のコレクションである。次に式（１５）は、以下の式（１６）で示すように再定義することができる。

However,

Is an estimation target (detailed below). <U> is a subset of U _m including users with the same targeting attribute T. <U _m > is a division of U _m by <u>, that is, a collection of <u>. Equation (15) can then be redefined as shown in Equation (16) below.

は、属性（ｍ，＜ｕ＞）のユーザの推定オンライン広告視聴時間である。この方法では、Ｗは、次の処理を用いて計算される。この処理では、Ｍ（Ｔ）を共通の要素をもたない２つの部分に分割する：Ｍ（Ｔ）＝Ｍ’ ｕ Ｍ_ａ。ここでＭ_ａは、広告主により所有されているメディア（たとえば、広告主に関連するウェブ・ページ）の集まりであり、Ｍ’は、その他のメディア（たとえば当該広告主の広告）である。Ｍ’中の各メディアｍについて、次の方法によりＷを計算する：ｂを入札することによりターゲティング属性＜ｕ＞のユーザのインプレッションを獲得する確率Ｗ（ｍ，＜ｕ＞，ｂ）を推定する。ｍから生成されたｋ個のインプレッション（ターゲティング属性＜ｕ＞）のデータを収集する。収集されたデータは、（１）各インプレッションｋの最高入札価格

および（２）２進値

（ここで、ｉ番目のインプレッションが入札価格ｂ_ｉにより落札された場合、ａ_ｉは１の値をとる）などの情報を含んでいる。１つの例では、インプレッションの競売を行うＲＴＢがこのようなデータの源である。次に、下に示す式（１７）を用いてＷを推定することができる。

However,

Is the estimated online advertisement viewing time of the user with attribute (m, <u>). In this method, W is calculated using the following process. In this process, M (T) is divided into two parts that do not have a common element: M (T) = M ′ u M _a . Here M _a, the media that are owned by the advertiser (for example, advertisers related web page) is a collection of, M 'is the other media (for example, the advertiser's ads). For each medium m in M ′, calculate W by the following method: Estimate the probability W (m, <u>, b) of acquiring the impression of the user with the targeting attribute <u> by bidding b . Collect data for k impressions (targeting attribute <u>) generated from m. The collected data is (1) the highest bid price for each impression k

And (2) binary value

(Here, if the i-th impression is successful bid by the bid price b _i , a _i takes a value of 1). In one example, an RTB that competes for impressions is the source of such data. Next, W can be estimated using equation (17) shown below.

は、ｂ＞＝ｂ_ｉのとき１をとり、そうでないとき０をとる関数である。ここで、Ｗは、各ターゲティング属性について独立に計算されたが、＜ｕ＞について平均された。さらに、メディアおよびターゲティング属性を指定してデータベースを検索することを仮定したが、検索がキャンペーンのみに関して行われる場合でも同様な方法を適用できる。 However, 0 ≦ p (b _i ) ≦ 1 is a certain constant depending on b _i ,

Is a function that takes 1 when b> = b _i and takes 0 otherwise. Here, W was calculated independently for each targeting attribute, but averaged for <u>. Furthermore, although it is assumed that the database is searched by specifying the media and targeting attributes, the same method can be applied even when the search is performed only for the campaign.

を計算する方法を示す。この推定は、前項において各ユーザについて測定したｖ_ｅに基づく。ｆ_ｍ（＜ｕ＞）を推定する１つの方法は、次のとおりである：Ｄをもって考察中のキャンペーンの期間のようなある期間の長さとする。さらに、Ｕ_ｍをもって長さＤの期間中にサイトｍを訪ねるユーザの集まりとし、ｖ_ｕをもってユーザｕ ε ＜ｕ＞による訪問の回数とする。実施形態では、ｆ_ｍ（＜ｕ＞）は、次式（１８ａ）のように推定することができる。

For each media in M _a, estimates the product of the elements W and _{f m.} The following discussion:

Shows how to calculate This estimate is based on v _e measured for each user in the preceding paragraph. One way to estimate the f m _(<u>) are as follows: the length of such a period as the period of the campaign in consideration with a D. Furthermore, U _m is a group of users who visit the site m during the period of length D, and v _u is the number of visits by the user u ε <u>. In the embodiment, f _m (<u>) can be estimated as in the following equation (18a).

は、ｖ_ｕまたはｖ_ｕのメジアン、ｖ_ｕのモードまたはｖ_ｕのメジアンの平均

のいずれかとすることができるが、それらには限定されない。実施形態では、

Where f _c is the frequency limit for this ad (eg, the maximum number of complete views of the given ad for the given user)

Is the average of v _u or v _u median, v _u mode or v _u median

However, it is not limited to them. In an embodiment,

は、メディアｍにおいて生成されたインプレッションの数および完全視聴を達成した広告を表示したインプレッションの数の関数として次式（１８ｃ）に示すように推定することができる。

Furthermore, in the embodiment,

Can be estimated as a function of the number of impressions generated in media m and the number of impressions that displayed an advertisement that achieved full viewing, as shown in equation (18c).

Where, for each pair of media (m, <u>) and the target attribute user, N (m, <u>) represents the number of impressions generated at m by the targeting attribute user <u>, _{Further, N L (m, <u} >) subset N (m, <u>) of the impressions _{L CPV} (i.e. in full view the given advertisement to be achieved for the relevant advertisements to be recorded The number of impressions (for a given advertisement displayed within the impression) that have an advertisement viewing time equal to or longer than a predetermined advertisement viewing time).

は、ターゲティング属性を有するユーザ＜ｕ＞のサブセットに関して測定された平均広告視聴時間の関数として決定することができる。実施形態では、ｅＧＲＰ予測関数の種々のパラメータは、その点までに達成された特定の広告キャンペーンに関するメディアおよびターゲティング属性を指定して行うデータベースの検索に基づいて推定することができる。実施形態では、ｅＧＲＰ予測関数の種々のパラメータは、同様な広告キャンペーンに関連するメディアおよびターゲティング属性の検索に基づいて推定することができる。ここで、広告キャンペーンの類似性は、当該広告キャンペーンの目的、キャンペーンにおいて使用される広告クリエイティブの予算、対象視聴者、性質等に基づいて決定できる。 In an embodiment,

Can be determined as a function of average ad viewing time measured for a subset of users <u> having targeting attributes. In an embodiment, various parameters of the eGRP prediction function may be estimated based on a database search that specifies media and targeting attributes for a particular advertising campaign that has been achieved up to that point. In an embodiment, various parameters of the eGRP prediction function can be estimated based on a search for media and targeting attributes associated with similar advertising campaigns. Here, the similarity of the advertisement campaign can be determined based on the purpose of the advertisement campaign, the budget of the advertisement creative used in the campaign, the target audience, the nature, and the like.

の積がｆ_ｃ（所与のユーザについて到達されるべき完全視聴の回数の上限）より、たまたま大きい場合、ｅＧＲＰを計算するとき、

の積はｆ_ｃに設定することができる。かかる置換は、ｅＧＲＰが広告ターゲティングを行う場合の所与の広告に関する最大完全視聴回数の効果を反映することおよび所与の予算によるｅＧＲＰの推定の精度を向上することを可能とし、それにより広告主は前もって当該広告キャンペーンのＲＯＩ（投資収益率）をより正確に推定できるようになる。さらに、広告は、ＴＶ広告よりオンライン広告において所与のユーザにより頻繁に配信される傾向があるが、オンライン広告キャンペーンにおける所与のユーザに対する広告掲載の回数は、所与の広告の最大完全視聴回数を制限することにより、ＴＶ広告が所与のユーザにより視聴されると思われる回数のそれに抑制することができる。かかる最大完全視聴回数に基づいて推定されたｅＧＲＰは、ＴＶ広告とオンライン広告の比較を可能にする。ｅＧＲＰ推定は、ＴＶ広告のより広い視聴者到達を把握し、かつ、オンライン・キャンペーンのより高い頻度の同一サブセット視聴者到達の効果を制限するからである。 In an embodiment,

Than the product is f _{c (upper} limit for the number of full view to be reached for a given user), if happens large, when calculating the EGRP,

The product can be set to f _c. Such a substitution allows to reflect the effect of the maximum number of full views for a given ad when eGRP does ad targeting and improve the accuracy of eGRP estimation for a given budget, thereby Will be able to estimate the ROI (return on investment) of the advertising campaign more accurately in advance. In addition, advertisements tend to be delivered more often by a given user in online advertising than TV advertising, but the number of times an ad is served to a given user in an online advertising campaign is the maximum number of full views of a given advertisement. By limiting the number of TV advertisements, the number of times that a TV advertisement is expected to be viewed by a given user can be suppressed. The eGRP estimated based on the maximum number of complete viewing times enables a comparison between a TV advertisement and an online advertisement. This is because eGRP estimation captures the wider audience reach of TV advertisements and limits the effectiveness of the higher frequency same subset audience reach of online campaigns.

は、次式（１９）に示すようにｕ ε ＜ｕ＞にわたるｖ_ｅ（ｍ；ｕ）の重み付け平均として推定することができる。

further,

Can be estimated as a weighted average of v _e (m; u) over u ε <u> as shown in the following equation (19).

は、＜ｕ＞に属するユーザによる訪問の合計回数である。さらに、ｖ_ｕは、２つの意味で定義される：ユーザｕによる訪問の回数またはユーザｕの訪問の集まり。さらに、＜ｕ＞に関して平均されたｆ_ｍおよび

を導くことができる。さらに、メディアおよびターゲティング属性を指定するデータベースの検索を仮定しているが、キャンペーンのみに関して検索が行われる場合にも同様な方法を適用することができる。 However,

Is the total number of visits by users belonging to <u>. Furthermore, v _u is defined in two ways: the number of visits by user u or the collection of visits by user u. Further, _{f m} and is averaged over <u>

Can guide you. Furthermore, although a database search specifying media and targeting attributes is assumed, a similar method can be applied when a search is performed only for a campaign.

Using the estimated values of the above, it is possible to estimate the product of f _m and W for each m _a in M _a using the following equation (20).

のそれぞれの平均を＜ｕ＞についてとった場合、予測公式は、次式（２１）に示すように定義することができる。

Here, _cm is the click-through rate (CTR) of the medium m. Aggregation is performed for m. Some ad on where m is linked to m _a (i.e., linked to owned media). W, _{f m} and

When the average of each is taken for <u>, the prediction formula can be defined as shown in the following equation (21).

Here, v _m is the predicted total online advertisement viewing time of the medium m, and n _m is the expected number of impressions generated from the medium m.

  FIG. 14C shows an overview of a system 1420 that can be used to predict the eGRP that can be achieved for a given ad spend. Server 1430 predicts eGRP based on data available via databases 1426 and 1428. Database 1426 is maintained by a server 1424 that collects measured advertisement viewing times for various advertisements that are placed on various advertisement slots on various web pages. Utilizing the collected data, advertisement viewing that can be achieved by utilizing the database 1426 to display advertisements to users with certain targeting attributes in various media with certain targeting attributes (eg, as desired by an advertising campaign). Can be determined. Similarly, the database 1428 is maintained by a server 1422 that collects data regarding various impressions (ie, a given advertisement display for a given advertisement slot). Using the collected data, the database 1428 can be used to determine data related to previous impressions such as the bid price at which the impression was sold, attributes related to the impression, and the like. As described above, the server 1430 can predict eGRP based on collected data from the databases 1426 and 1428.

  FIG. 14E is an illustrative example of a method 1445 that can be used to predict possible eGRP that can be achieved for an advertising campaign based on a given bid amount. The method 1445 identifies potential advertisement slots that can be purchased and used to deliver advertisements from the advertisement campaign to one or more targets (ie, users). Given the identified ad slot as a basis and then using the predicted achievable overall exposure and the predicted achieved overall exposure, similar to the method for estimating the eGRP achieved in method 1435 Calculate the expected eGRP that can be achieved for the bid amount of “b”.

  At step 1446 of method 1445, a subset of media that can display advertisements for the advertising campaign is selected. This selection can be made based on attributes related to the media such as contents, properties, audience appeal, etc. desired by the advertisement campaign. Subsets of users who each viewed at least one advertisement (for any campaign) of the identified subset of media in step 1448 of method 1445. In some embodiments, this subset of users may be further filtered to include only a subset of users having one or more attributes of viewers that the campaign wishes to reach.

  At step 1450 of method 1445, for each user in this subset of users, each media in this subset of media using a given bid price “b” (for displaying ads from that advertising campaign). Determine the probability of winning an ad slot at and the frequency of visits of that user to each media in this subset of media. In one embodiment, the probability of acquiring an ad slot in a given media for a given bid amount can be determined based on the pre-sale price of that ad slot.

  In step 1452 of method 1445, for each user in this subset of users, the scale of each potential ad view that reduces the effect on the user of the viewed advertisement (when calculating the total exposure) is determined. In step 1454 of method 1445, the probability of acquiring the ad slot determined in step 1450, the frequency of available ad slots (based on the frequency of user visits determined in step 1450), and the determined in step 1452 As a function of the scaling factor, the total predicted achievable target of the bid price “b” is determined. At step 1456 of the method 1445, a total predicted available target on the identified available media that may have displayed the advertisement from the advertisement campaign is determined. In step 1458 of method 1445, predicted eGRP (a measure of predicted total exposure that can be achieved for the advertising campaign with a given bid amount “b”) as a function of total predicted obtainable target and total predicted available target is measured. Is done. Equation (11) is an example of a predicted eGRP calculation function.

  FIG. 14H is an illustrative example of an interaction 1480 between various participants evaluating eGRP in one embodiment of a system used to evaluate eGRP. One of the participants, for example, the advertiser, requests an eGRP calculation (eGRP achieved for an ongoing advertising campaign or eGRP estimate for a given advertising campaign budget) using a computer system such as a front-end computer To do. This request includes information about the TV advertising campaign, such as the ads that are run on the TV used in the estimation or prediction of eGRP for the online advertising campaign, the number of spots, and the GRP achieved. A calculation server executing a method for estimating and predicting eGRP receives an eGRP evaluation request, and subsequently makes a query to a database storing information related to an online advertising campaign in order to calculate eGRP. A database that interacts with a logging server that tracks the user's various behaviors and the performance of the advertising campaign (such as the amount of time the posted advertisement was achieved) can search online advertising campaign data It is managed in the form. In response to the query, the database returns various log data regarding the online advertising campaign to the calculation server. Using the eGRP estimation / prediction method described above, the calculation server calculates the eGRP of the online advertising campaign from the information obtained from the database, and returns the eGRP to the computer system.

  FIG. 14I is an explanatory example of a user interface (UI) displayed on the computer system. Advertisers or advertising agencies can use this to interact with a calculation server that estimates or calculates eGRP. This UI can be used to obtain an estimate of the eGRP before starting the advertising campaign, or the resulting achieved eGRP during or after the advertising campaign. UI 1485 shows the interface before entering information about online campaigns, while UI 1490 displays the various information after entering the required information and pressing the “Calculate” buttons 1486, 1491 in the UI. Showing the interface. The UI 1485 gray boxes 1488, 1489 are automatically filled based on the information provided after pressing the “Calculate” button.

  In FIG. 14I, a client (eg, an advertiser or advertising agency) can select one of the regions provided in drop-down menus 1487, 1492. The client may also assemble the page so that the client can select multiple non-overlapping areas from a number of possible choices. Further, these non-overlapping areas can be freely selected, thereby allowing a certain district (eg, country) to be divided into a number of different sets of non-overlapping areas. For example, when estimating or predicting eGRP for the US market, the US market can be divided in various ways: (1) state divisions, (2) programs for a given TV station. Is divided according to whether or not it is viewable in a certain area.

  FIG. 14J is an explanatory example of a user interface (UI) displayed on the computer system. Advertisers or advertising agencies can use this to interact with a calculation server that estimates or calculates eGRP. UI 1495 gives the achieved (or predicted) eGRP at any given time depending on the portion of the online advertising campaign budget spent.

<< Bid in the form of cost per desired eGRP >>
As described above, the advertiser can make an advertising bid based on the desired eGRP. The following discussion provides a way to determine the bid price “b” needed to achieve the desired eGRP. As shown in Expression (15), eGRP is generally defined as in the following Expression (22).

In the above equation (22), b is a bid price in the form of CPX. However, W is a monotonically increasing function with respect to b (ie, increasing b increases the winning bid rate W to be the same or increase regardless of (m, u). An appropriate eGRP for b can be determined by binary search eGRP (b) using the following method, which determines a sufficiently large upper bound b _u of b, which represents the current candidate bid price with b _s And let b _l = 0, and let G denote the advertiser's desired eGRP, where b _s = (b ₁ + b _u ) / 2 and eGRP (b _s ) are calculated. If eGRP (b _s ) <G, then let b _l : = b _s , otherwise b _u : = b _s .

が成り立つまで、境界を再設定する上記のステップを繰り返す。最終繰り返しにおいて到達した入札価格ｂ_ｓが所望ｅＧＲＰを達成する提案入札価格である。ここで、最終繰り返しは、所望ＧＲＰと決定されたＧＲＰの差が許容値ｄより小さくなるステップである。さらに、Ｒ_ＴＶおよびＬ_ＴＶは、関数１８により計算されるｅＧＲＰにおける定数であるから、この提案入札価格「ｂ」は、所望オンライン広告視聴時間を達成する等価提案入札価格である。さらに、オンライン広告視聴時間の予測において広告フレームのサイズを考慮しない場合、提案入札価格「ｂ」は、所有メディアを含むメディアから得た所望広告視聴時間から導かれたものとみなすことができる。 For a given d (allowable value)

Repeat the above steps to reset the boundary until. Bid price b _s has been reached in the final repetition is proposed bid price to achieve the desired eGRP. Here, the final repetition is a step in which the difference between the desired GRP and the determined GRP is smaller than the allowable value d. Furthermore, since R _TV and L _TV are constants in eGRP calculated by function 18, this suggested bid price “b” is an equivalent suggested bid price that achieves the desired online advertisement viewing time. Further, when the size of the advertisement frame is not considered in the prediction of the online advertisement viewing time, the proposed bid price “b” can be regarded as derived from the desired advertisement viewing time obtained from the media including the owned media.

In another embodiment, the cost per desired eGRP (in CPX form) can be derived as shown in equation (23).

  Where G represents the desired cost per GRP, L represents the length of time for the ad to be delivered, and R is the total reach of all media that match the targeting attribute specified by the advertiser Represents.

  FIG. 14F illustrates a method 1460 that can be utilized to determine the cost of achieving the desired eGRP for an advertising campaign. Further, advertisers (and campaign managers) can use method 1460 to bid in the form of the desired eGRP. This can then be translated into costs that can be used to achieve the desired eGRP. In method 1460, a binary search of the eGRP prediction function (described above) is performed until a cost value (the predicted eGRP value at that cost value approaches the desired eGRP value) is identified. This cost value is therefore equal to the cost required to achieve the desired eGRP. In step 1462, upper and lower bid amounts (ie, costs required to achieve the desired eGRP) are determined, and a search for bid amounts required to achieve the desired eGRP is searched within this limit. Is done.

  In step 1464, the upper and lower limits are averaged and the average bid price is used to determine a predicted eGRP that can be achieved at the average bid price. If, at step 1466, the difference between the desired eGRP and the predicted eGRP is less than the tolerance “d”, then in step 1468, this average bid amount is equal to the cost required to achieve the desired eGRP. Otherwise, in step 1470, it is examined whether the predicted eGRP is over-achieved with respect to the average bid amount (ie, exceeds the desired eGRP). If so, in step 1474, the upper limit bid amount is set to the average bid amount and step 1464 is repeated. Otherwise, in step 1472, the lower limit bid price is set to the average bid price and step 1464 is repeated.

FIG. 14G shows an example of eGRP calculated based on the various parameters described above. Table 1 shows the studied examples and the resulting eGRP. Here, AVT is an abbreviation for average advertisement viewing time, and AOMVT is an abbreviation for average owned media advertisement viewing time. Here, the number of TV viewers defines the arrival of TV. Here, the arrival of the TV is estimated by (number of viewership zones) × (number of viewers per viewership zone). In the following, the number of viewing stations is set to 15 million (the number of families in the Kanto region), and the number of viewers per viewing station is set to 2.5. Therefore, the estimated number of TV viewers is 37.5 million. The advertisement length of TV advertisement is set to 15 seconds, and the cost per GRP of TV advertisement is set to 100,000 yen. Therefore, this advertiser acquires 1000 GRP in TV advertisement (with a budget of 100 million yen). The budget for online advertising is 30 million yen. Purchased impressions are distributed equally to each medium. Set _Pr to 0.5. Here, if a user visited the site, which is owned by the advertiser by the click-through, P _r is not added, only when an ad is delivered to the user, to add a P _r.

  FIG. 14K is an illustrative example of an interaction 1496 between various participants that determines the cost of achieving the desired eGRP for an advertising campaign in one embodiment of the system. One of the participants, for example, the advertiser, requests calculation of the cost (or bid amount) to achieve the desired eGRP using a computer system such as a front-end server. This request includes information on TV advertisement campaigns such as advertisements to be played on TV, the number of spots, GRP to be achieved, and the like. These are used in the estimation or prediction of eGRP for online advertising campaigns. A computing server that performs the method 1460 to estimate the cost of achieving the desired eGRP receives the bid amount evaluation request, and then uses an online advertising campaign (which includes the current online advertising campaign and the bid amount to calculate the bid amount). Query a database that stores information related to (or similar to or part of).

  A database that interacts with a logging server that tracks the user's various behaviors and the performance of the advertising campaign (such as the amount of time the posted advertisement was achieved) can search online advertising campaign data It is managed in the form. In response to the query, the database returns various log data regarding the online campaign to the calculation server. Using the cost estimation method 1260, the compute server recursively queries the database to recalculate the achievable eGRP for a given cost until the computed eGRP falls within the desired eGRP range. The estimated cost for the calculated eGRP is returned to the computer system as the bid amount necessary to achieve the desired eGRP.

  FIG. 14L is an example of a user interface (UI) displayed on the computer system. Advertisers or advertising agencies can use this to interact with a calculation server that estimates the cost (or bid amount) to achieve the desired eGRP. This UI can be used to obtain an estimate of the cost of achieving the desired eGRP before starting an advertising campaign. UI 1497 shows the interface before entering information about the online campaign, while UI 1498 displays the various information after entering the required information and pressing the “Calculate” button 1494 in the UI. Shows the interface. The gray box 1499 of UI 1497 is automatically filled based on the information provided after pressing the “Calculate” button.

  In FIG. 14L, a client (eg, an advertiser or advertising agency) can select one of the regions provided in the “Region where TV advertisement was aired” drop-down menu. The client may also assemble the page so that the client can select multiple non-overlapping areas from a number of possible choices. Further, these non-overlapping areas can be freely selected, thereby allowing a certain district (eg, country) to be divided into a number of different sets of non-overlapping areas. For example, when estimating the cost of a desired eGRP in the US market, the US market can be divided in various ways (and the cost is estimated as the sum of the costs of each divided region): : (1) Division by state, (2) Division by whether or not a program of a given TV broadcasting station is viewable in a certain area, etc.

  FIG. 15 illustrates a computer system that can be utilized, for example, to implement a platform server (eg, 114 of FIG. 1), a web server (eg, 125 of FIG. 1), or other computing devices as indicated in the above disclosure. 2 is a high level block diagram illustrating an example of a 1500 architecture. FIG. Computer system 1500 includes one or more processors 1505 and memory 1610 connected via interconnect 1525. The interconnect 1525 is an abstract representation of both of these connected by one or more individual physical buses, point-to-point connections, appropriate bridges, adapters or controllers. Thus, the interconnect 1525 can be, for example, a system bus, a peripheral component interconnect (PCI) bus, a HyperTransport or industry standard architecture (ISA) bus, a small computer peripheral interface (SCSI) bus, a universal serial bus ( USB), IIC (I2C) bus, or Institute of Electrical and Electronics Engineers (IEEE) standard 694 bus (sometimes referred to as “Firewire”).

  The processor 1505 includes, for example, a central processing unit (CPU) that controls the overall operation of the host computer. In certain embodiments, processor 1505 does this by executing software or firmware stored in memory 1510. Is processor 1505 one or more programmable general purpose or special purpose microprocessors, digital signal processors, programmable controllers, application specific integrated circuits (ASICs), programmable logic devices (PLDs), etc., or a combination of these devices? Or including them.

  Memory 1510 is or includes main memory of computer system 1500. Memory 1510 represents any form of random access memory (RAM), read only memory (ROM), flash memory (as described above), etc., or a combination of these devices. In use, memory 1510 includes, among other things, a set of machine language instructions that, when executed by processor 1505, cause processor 1505 to perform operations that implement embodiments of the present invention.

  Also connected to the processor 1505 via 1525 is a network adapter 1515. The network adapter 1515 is a device that provides the computer system 1500 with the capability to communicate with remote devices such as storage clients and / or other storage servers, such as an Ethernet adapter or a fiber channel adapter. is there.

Claims (63)

Receiving an advertisement displayed in an advertising portion of a scrollable web page of an advertising publisher by a server comprising a processor, the advertising being displayed for a predetermined duration in the advertising portion And the steps that are
Sending, by the server, the advertisement for display within the advertising portion of the advertising publisher's web page;
Measuring a user's total attention duration for the advertisement displayed on the advertisement portion of the advertisement publisher's web page by the server, wherein the advertisement portion is a web page in response to scrolling of the web page; - exist in the viewable area or viewing area outside of the browser, the total attention duration of the user, the duration of the advertisement has been displayed in a predetermined state in the web browser the viewable area of the The total attention duration of the user for the advertisement is less than or equal to the predetermined duration of the advertisement;
Determining by the server a cost incurred for displaying the advertisement as a function of the user's total attention duration for the advertisement.   Cost per view (CPV) billing where the cost incurred to display the advertisement determines the cost of displaying the advertisement as a function of both the total attention duration of the user for the advertisement and a predetermined minimum display duration The method of claim 1, wherein a cost incurred for displaying the advertisement is zero based on a model and when the total attention duration of the user is less than the predetermined minimum display duration. Sending an advertisement to be displayed to a user by a server comprising a processor, wherein the advertisement is displayed in an advertisement portion of a scrollable web page of the advertising media company, and the web page of the advertising media company is the user The advertisement portion is present in a viewable area or outside a viewable area of a web browser in response to scrolling of the webpage ; and
By the server, the method comprising: receiving the cost of displaying the advertisement to the user, the cost is calculated as a function of the total care duration of the user with respect to the advertisement, the total attention duration, the web Measuring as a function of the duration that the advertisement was displayed in a predetermined state within the viewable area of the browser.   Cost per view (CPV) billing where the cost incurred to display the advertisement determines the cost of displaying the advertisement as a function of both the total attention duration of the user for the advertisement and a predetermined minimum display duration The method of claim 3, wherein the cost incurred for displaying the advertisement is zero based on a model and when the total attention duration of the user is less than a predetermined minimum display duration. At least one memory storing computer-executable instructions;
At least one processor configured to access the at least one memory and execute the computer-executable instructions to perform a series of operations, the operations being performed by an advertising media company scrollable web page; Receiving an advertisement to be displayed in an advertising portion, wherein the advertisement is intended to be displayed in the advertising portion for a predetermined duration;
Sending the advertisement for display within the advertising portion of the advertising publisher's web page;
Measuring a user's total attention duration for the advertisement displayed on the advertisement portion of the advertisement publisher's web page, wherein the advertisement portion is viewed by a web browser in response to scrolling of the web page present outside region within or viewable area, wherein the total note duration of the user, measured as a function of the duration in which the advertisement was displayed in a predetermined state in the web browser the viewable area of the The total attention duration of the user for the advertisement is less than or equal to the predetermined duration of the advertisement;
Determining a cost incurred for displaying the advertisement as a function of the total attention duration of the user for the advertisement.   Cost per view (CPV) billing where the cost incurred to display the advertisement determines the cost of displaying the advertisement as a function of both the total attention duration of the user for the advertisement and a predetermined minimum display duration 6. The system of claim 5, wherein the cost incurred for displaying the advertisement is zero based on a model and when the total attention duration of the user is shorter than the predetermined minimum display duration. Wherein a step of identifying at least 1 advertisements from the plurality of advertisement by the server, determined as the function of the particular minimum auction prices for the bid amount and given impressions relating to the advertising of the advertisement from the plurality of ad it is, and the step is a proposed bid amount of money that the minimum auction price will be paid with respect to the display of the advertisement in the given impression,
Wherein a step of determining each ad relevance scores of advertisements identified above reviews the 1 by the server, the advertisement relevance score of the advertisement and the bid amount extra regarding the advertisement and the bid amount relating to the advertising is determined as a function of both, is determined as a function of acceptable display the total number of the advertisement to the total number and the user of the prior display of the advertisement to the user regarding the advertisement the advertising relevance score further said impressions Steps,
Classifying the one or more identified advertisements by the server based on the advertisement relevance score for each of the one or more identified advertisements;
Identifying the advertisement to be displayed in the impression from the one or more classified advertisements by the server, wherein the identified advertisement displayed in the impression is a function of the advertisement relevance score for the advertisement The method of claim 1, further comprising: It said preceding display ads is determined as a function of both the predetermined minimum duration of the display time for the advertisement and the total note duration of the user relative to the display of the advertisement, the total attention duration display time of the user wherein when a predetermined shorter than the shortest duration is assumed that the preceding display of the advertisement does not occur a method according to claim 7. Wherein when preceding the total number of display of the advertisements for the user about the impression is greater than the total number of display acceptable in the advertisement for the user, the advertisement relevance score of the advertisement is zero, wherein Item 8. The method according to Item 7. Said measured as the total attention duration of the duration of time that the attention of the user is focused on the display of the advertising function of the user with respect to advertising, the total attention duration of the user with respect to the advertisement said The method of claim 8, wherein the method is less than or equal to the predetermined duration of the advertisement. The server is
Part to the a the advertisement portion of said via a web-browser user can watch, the advertisement viewable portion of the attention of the user is greater than the minimum prescribed threshold are concentrated in the advertisement,
The part the user can watch one of the web browser to display the advertisement to the user or,
A interaction with the web browser of the user, based on one or more criteria including the interaction that occurs at least once in each of the one or more predetermined time period, the attention of the user with respect to the advertisement The method of claim 10, wherein the method is determined. And determining the ad relevance score for each of a plurality of advertisements by the server, displays the advertisement-related score is allowed of the advertisement to the total number and the user of the prior display of the advertisement to the user about the impression Steps determined as a function of the total number of times,
Identifying the advertisement to be displayed in the impression from the plurality of advertisements by the server, wherein the identification of the advertisement from the plurality of advertisements displayed in the impression is the advertisement relevance score for the advertisement The method of claim 1, further comprising: determining as a function. It said preceding display ads is determined as a function of both the predetermined minimum duration of the display time for the advertisement and the total note duration of the user with respect to the display of the advertisement, the total note duration of the user display If less than the predetermined minimum duration of time is assumed to the preceding display of the advertisement does not occur a method according to claim 12. Wherein when the preceding display the total number of the advertisement for the user about the impression the advertisement acceptable the total number is larger than the display of for the user, the advertisement relevance score of the advertisement is zero, claim 12. The method according to 12. Said measured as the total attention duration of the duration of time that the attention of the user is focused on the display of the advertising function of the user with respect to advertising, the total attention duration of the user with respect to the advertisement said The method of claim 13, wherein the method is less than or equal to the predetermined duration of the advertisement. The server is
Part to the a the advertisement portion of said via a web-browser user can watch, the advertisement viewable portion of the attention of the user is greater than the minimum prescribed threshold are concentrated in the advertisement,
The part the user can watch one of the web browser to display the advertisement to the user or,
A interaction with the web browser of the user, based on one or more criteria including the interaction that occurs at least once in each of the one or more predetermined time period, the attention of the user with respect to the advertisement 16. The method of claim 15, wherein the method is determined. Said action is
Be to identify 1 or more ads from multiple ad, said the specific of the ads from multiple ad is determined as a function of the minimum auction price on the bid amount of money and a given impression on the advertisement, the and it is proposed bid amount of money that the lowest auction price will be paid with respect to the display of the advertisement in the given impression,
Comprising: determining a respective ad relevance scores of advertisements identified above 1 wherein, determining the advertisement relevance score as a function of both the bid amount premium relating to the advertisement and the bid amount relating to the advertising It is a to be determined as a function of the acceptable total number of display is of the advertisement to the preceding display the total number and the user of the advertisement the relative user regarding advertising relevance score further the impression of the advertisement,
Categorizing the one or more identified advertisements based on the advertisement relevance score for each of the one or more identified advertisements;
Identifying an advertisement to be displayed in the impression from the one or more classified advertisements, wherein the identified advertisement to be displayed in the impression is determined as a function of the advertisement relevance score for the advertisement. The system of claim 5 further comprising: It said preceding display ads is determined as a function of both the predetermined minimum duration of the total note duration and the display for advertising time of the user with respect to the display of the advertisement, the total note duration of the user display time above is shorter than a predetermined minimum duration is assumed that the preceding display of the advertisement does not occur, according to claim 17 systems. The ad relevance score of the advertisement if the preceding display the total number of the advertisements for the user about the impressions is more than the total number of display acceptable of the advertisements for the user is zero, claim 17 The system described in. Said measured as the total attention duration of the duration of time that the attention of the user is focused on the display of the advertising function of the user with respect to advertising, the total attention duration of the user with respect to the advertisement said The system of claim 17, wherein the system is less than or equal to the predetermined duration of advertisement. The processor is
Part to the a the advertisement portion of said via a web-browser user can watch, the advertisement viewable portion of the attention of the user is greater than the minimum prescribed threshold are concentrated in the advertisement,
The part the user can watch one of the web browser to display the advertisement to the user or,
A interaction with the web browser of the user, based on one or more criteria including the interaction that occurs at least once in each of the one or more predetermined time period, the attention of the user with respect to the advertisement 21. The system of claim 20, wherein the system is determined. And determining the total achieved advertisement view time by the advertisement campaign by the server, the total achieved advertisement view time of the display time for the duration and the advertisement of the advertisement viewing time measured for displaying advertisements for the advertising campaign Determined as a function of both a predetermined minimum duration and determined as a function of a user's total attention duration for the advertisement in which the duration of the measured advertisement viewing time is displayed, and the total attention duration time is measured as a function of the duration of time that attention of the user is focused on the display of the advertisement, the total note duration of the user is less than the predetermined minimum duration of the display time for the advertising advertisement viewing to be measured for the display of the advertisement on the advertising campaign in the case A step in which the duration between is zero,
Wherein a step of determining the advertisement total achievable advertisement viewing time by the campaign by the server, the total achievable advertisement view time is calculated as a function of the duration of the display time feasible for the display of the advertisement, the A possible display duration is determined as a function of the length of the advertisement;
Further comprising the step of determining the total exposure achieved by the advertisement campaign by the server as a function of the total achieved advertisement view time and the total achievable advertisement view time, The method of claim 1. Further comprising the method of claim 22 the step of determining the effective gross rating of the advertising campaign by the server (EGRP) as a function of the total exposure achieved by the advertisement campaign. By the server, said the eGRP of the advertising campaign is compared to the total audience rating of TV advertising campaign (GRP), the total exposure and the comparison is achieved by the television advertising campaign and the total exposure to be achieved by the advertising campaign be utilized to determine the difference between, the television advertising campaign, include displaying the advertisement on the advertising campaign in one or more of the audience that can be reached via the TV, the display time for the advertisement wherein the predetermined minimum duration is determined as a function of the length of the advertisement to be displayed in the television advertising campaign, the method of claim 23. The television advertising the GRP campaign is calculated as a television total number of functions of receiving the advertisement from the total number and the television advertising campaign television said displaying advertisements from the television advertising campaign, according to claim 24 the method of. The server is
A portion of the advertisement that is viewable by the user via the web browser, wherein the user's attention is concentrated on the advertisement when the viewable portion of the advertisement is greater than a minimum predetermined threshold;
A portion of the web browser that displays the advertisement to the user that the user can view, or interaction with the user's web browser, at least once in each of one or more predetermined time periods. 23. The method of claim 22, wherein the user's attention to the advertisement is determined based on one or more criteria including an interaction that occurs. The time duration during which the one or more criteria that determine the user's attention to the advertisement are met is measured, and the measured duration is the total attention duration of the user for the advertisement. 26. The method according to 26.   28. The method of claim 27, wherein the measurement of the duration is stopped when at least one of the one or more criteria that determine the user's attention to the advertisement is not met.   29. The method of claim 28, wherein the measurement of the duration is resumed when the one or more criteria that determine the user's attention to the advertisement are met. Determining a plurality of advertisements for an advertising campaign by the server;
For each of the plurality of advertisements, the method comprising the steps of determining the total achieved advertisement view time of the advertisement by the server, continue CPV time the total achieved advertisement view time is measured for the display of the advertisements for the user is calculated as a function of both the predetermined minimum duration of the display time and with respect to time and the advertisement, the duration of the advertisement view time to be the measurements are determined as a function of total attention duration of the user with respect to the advertisement, the total Note duration attention of the user is measured as a function of the duration of time that was focused on the display of the advertisement, the total note duration of the user is less than the predetermined minimum duration of the display time for the advertising step wherein the duration of said measured for the display of advertisement advertisement view time is zero when there And,
By the server, and determining the overall achievement advertisement view time of the advertising campaign, the steps of the overall achievement advertisement view time is calculated as a function of each of the total achieved advertisement view time of the plurality of advertisements,
For each of the plurality of advertisements, the method comprising the steps of determining the advertisement total achievable advertisement view time by the server, the total achievable CPV said display duration of display time feasible for the time the ad a step of being calculated as a function, the display duration feasible is calculated as a function of the length of the advertisement,
And determining the overall achievable advertisement view time of the advertisement campaign by the server, the step of the overall achievable advertisement view time is calculated as a function of each of the total achievable advertisement view time of the plurality of ads When,
By the server, further comprising the step of determining the total exposure achieved by advertising campaign as a function of the overall achievement advertisement view time and the overall achievable advertisement view time, The method of claim 1. Further comprising the step of determining the effective gross rating of the advertising campaign by the server (EGRP) as a function of the total exposure achieved by the advertising campaign, the method of claim 30. By the server, the eGRP of the advertising campaign is compared to the total audience rating of TV advertising campaign (GRP), is achieved the total exposure to be achieved and by the television advertising campaign by the advertising campaign by using the comparison to determine the difference between the total exposure, the television advertising campaign, include displaying the advertisement on online advertising in one or more of the audience that can be reached via TV, the predetermined display time for the advertisement shortest duration is determined as a function of the length of the said advertisement to be displayed in the television advertising campaign, the method of claim 31.   33. The GRP of the television advertisement campaign is calculated as a function of the total number of televisions that have displayed advertisements from the television advertisement campaign and the total number of televisions that have received advertisements from the television advertisement campaign. Method. The server is
Part to the a the advertisement portion of said via a web-browser user can watch, the advertisement viewable portion of the attention of the user is greater than the minimum prescribed threshold are concentrated in the advertisement,
A portion of the web browser that displays the advertisement to the user that the user can view, or interaction with the user's web browser, at least once in each of one or more predetermined time periods. The method of claim 30, wherein the user's attention to the advertisement is determined based on one or more criteria including an interaction that occurs. The duration of the time that the one or more criteria for determining the user's interest is satisfied for the advertisement is measured, and this measured duration with the total note duration of the user with respect to the advertisement, claims 34. The method according to 34.   36. The method of claim 35, wherein the measurement of the duration is stopped when at least one of the one or more criteria that determine the user's attention to the advertisement is not met.   40. The method of claim 36, wherein the measurement of the duration is resumed when the one or more criteria that determine the user's attention to the advertisement are met. The operation is
Comprising: determining the total achieved advertisement view time by the advertisement campaign, the total achieved advertisement view time a predetermined minimum continuous display time for the duration and the advertisement of the advertisement viewing time measured for displaying advertisements for the advertising campaign is determined as a function of both time and the duration of the measured advertisement view time is determined as a function of the total care duration of the user to the advertisement, the total attention duration attention of the user of the advertisement measured as a function of the duration of time that has been focused on the display, the display of the ad to the ad campaign when the total note duration of the user is less than the predetermined minimum duration of the display time for the advertising The duration of the ad viewing time measured for is assumed to be zero,
The method comprising determining the advertisement total achievable advertisement viewing time by the campaign, the total achievable advertisement view time is calculated as a function of the duration of the display time feasible for the display of the advertisement, which can be the realization The duration of the display time is calculated as a function of the length of the ad,
6. The system of claim 5, further comprising: determining a total exposure achieved by the advertising campaign as a function of the total achieved advertisement viewing time and the total achievable advertisement viewing time. Further comprising The system of claim 38 said motion determines the effective gross rating of the advertising campaign (EGRP) as a function of the total exposure achieved by the advertisement campaign. By said processor, said eGRP of the ad campaign is compared with gross rating television advertising campaign (GRP), is achieved the total exposure achieved with by the television advertising campaign by the advertising campaign using the comparison to determine the difference between the total exposure, the television advertising campaign, include displaying the advertisement on online advertising in one or more of the audience that can be reached via TV, the predetermined display time for the advertisement shortest duration is determined as the advertising function of the length of the displayed in the television advertising campaign, system of claim 39.   41. The GRP of the television advertisement campaign is calculated as a function of the total number of televisions that have displayed advertisements from the television advertisement campaign and the total number of televisions that have received advertisements from the television advertisement campaign. system. A portion of the advertisement that is viewable by the user via the web browser, wherein the user's attention is concentrated on the advertisement when the viewable portion of the advertisement is greater than a minimum predetermined threshold;
A portion of the web browser that displays the advertisement to the user that the user can view, or interaction with the user's web browser, at least once in each of one or more predetermined time periods. 40. The system of claim 38, wherein the user's attention to the advertisement is determined based on one or more criteria including an interaction that occurs. The duration of the time that the one or more criteria for determining the user's interest is satisfied for the advertisement is measured, and this measured duration with the total note duration of the user with respect to the advertisement, claims 45. The system according to 42.   44. The system of claim 43, wherein the measurement of the duration is stopped when at least one of the one or more criteria that determine the user's attention to the advertisement is not met.   45. The system of claim 44, wherein the measurement of the duration is resumed when the one or more criteria that determine the user's attention to the advertisement are met. And determining the impressions Tsu Deployment suitable for displaying advertisements for advertising campaigns by the server, determined based on a match between the characteristics of the characteristics and the impressions required suitability is by the advertising campaign of said impressions And steps
Determining, by the server, a probability of acquiring the impression with a given bid amount;
Determining a predicted advertisement viewing time for the impression by the server, wherein the predicted advertisement viewing time is calculated as a function of an average duration of advertisement viewing time achieved by one or more equivalent impressions; and the step is a function of the total attention duration of the user with respect to the advertisement given of advertisement viewing time is displayed on the impressions,
By the server, determining a predicted frequency of generation of one or more equivalents of impression by the user,
The server determines a predicted perfect audience rating as a ratio of a total number of one or more equivalent impressions that have achieved an advertisement viewing time longer than a predetermined minimum advertisement viewing time and the total number of one or more equivalent impressions. Steps,
By the server, said method comprising the steps of: determining the total prediction ad viewing time of the advertising campaign, the probability and the impressions Tsu said predicted ad viewing time of the function of Deployment the total predicted ad viewing time to win the impressions Tsu Deployment Calculating the total predicted ad viewing time as a function of the predicted frequency and the predicted perfect audience rating;
By the server, and determining the total availability advertisement viewing time by the advertising campaign, the steps of the total available advertisement view time is calculated as a function of the length of the advertisement,
The method of claim 1, further comprising: determining, by the server, a total exposure achievable by the advertising campaign as a function of the total predicted advertisement viewing time and the total available advertisement viewing time. By said server, further comprising the step of determining the predicted effective gross rating of the advertising campaign (EGRP) as a function of the total exposure achievable by the advertising campaign, the method of claim 46. The equivalent of impressions with the total number and related sale price of less than or equal to the given of the bid price of the equivalent of impressions Tsu Deployment probability is Uriwatasa so far to win the given the impression by the given of the bid amount of money It is determined as a function of the total number of Tsu Deployment, the equivalent impressions Tsu Deployment having an associated characteristics required by the advertiser a method according to claim 46. Given achieve the duration of the advertisement viewing time is calculated as a function of the total attention duration of the user with respect to the advertisement in a given equivalent impressions, the total attention duration attention of the users are concentrated in the display of the advertisement 48. The method of claim 46, wherein the method is measured as a function of time duration. The given bid amount is received from an advertiser displaying an advertisement for the advertising campaign, the given bid amount is used to acquire an impression to display the advertisement to a user, and the advertisement of the advertisement to the user 48. The method of claim 46, wherein display achieves exposure of the advertising campaign. The server is
It said a the advertisement portion of that can be the viewing users via a web browser, part of the advertisement viewable portion of it is the user's attention to the case is larger than the minimum prescribed threshold are concentrated in the advertisement,
The part the user can watch one of the web browser to display the advertisement to the user or,
A interaction with the web browser of the user, based on one or more criteria including the interaction that occurs at least once in each of the one or more predetermined time period, the attention of the user with respect to the advertisement 48. The method of claim 46, wherein the method is determined. The duration of the time that the one or more criteria for determining the user's interest is satisfied for the advertisement is measured, and this measured duration with the total note duration of the user with respect to the advertisement, claims 51. The method according to 51. The method according to which the said measuring duration when at least one is not satisfied of the one or more criteria for determining the attention of the user is canceled, claim 52 for the advertisement. The duration the when the one or more criteria for determining the attention of the user with respect to the advertisement is satisfied measured is restarted, the method according to claim 53. Said action is
Comprising: determining the impressions Tsu Deployment suitable for displaying advertising relating to an ad campaign is determined adequacy of the impressions Tsu Deployment is based on a match between the characteristics relating to the impressions and characteristics required by the ad campaign And
And determining the probability of obtaining the impressions Tsu Deployment by a given bid amount of money,
Comprising: determining a predictive advertisement view time of the impressions Tsu Deployment, the prospective ad viewing time is calculated as a function of the average duration of the advertisement view time achieved by one or more equivalents of the impression, given and is a function of the total note duration of a given user with respect to the advertisement advertisement view time is displayed on the impression,
Determining a predicted frequency of generation of one or more equivalents of impressions Tsu Deployment by the user,
Determining the predicted perfect audience rating as a ratio of the total number of one or more equivalent impressions that achieve an ad viewing time longer than a predetermined minimum ad viewing time to the total number of one or more equivalent impressions;
Said was to determine the total prediction ad viewing time of the advertising campaign, is calculated as a function of the predicted ad viewing time of the probability and the impressions Tsu Deployment the total predicted ad viewing time to win the impressions Tsu Deployment, The total predicted ad viewing time is calculated as a function of the predicted frequency and the predicted complete audience rating;
Comprising: determining a total available advertisement viewing time by the advertising campaign, and said total available advertisement view time is calculated as a function of the length of the advertisement,
6. The system of claim 5, further comprising: determining a total exposure achievable by the advertising campaign as a function of the total predicted advertisement viewing time and the total available advertisement viewing time. Further comprising The system of claim 55 said motion determines predicted gross rating of the advertising campaign (EGRP) as a function of the total exposure achievable by the advertising campaign. By the given of the bid price of the equivalent of impressions Tsu Deployment with the total number and related sale price of less than or equal to the bid price of the given of equal probability to win the impression has been Uriwatasa so far impressions Tsu Deployment It is determined as a function of the total number, the equivalent of impressions Tsu Deployment having an associated characteristics required by the advertiser, the system according to claim 55. Given achieve the duration of the advertisement viewing time is calculated as a function of the total attention duration of the user with respect to the advertisement in a given equivalent impressions, the total attention duration attention of the users are concentrated in the display of the advertisement 56. The system of claim 55, measured as a function of duration of time. The given bid amount is received from an advertiser displaying an advertisement relating to the advertising campaign, and the given bid amount is utilized to acquire an impression to display the advertisement to a user, the advertisement to the user; 56. The system of claim 55, wherein display of achieves exposure of the advertising campaign. Part to the a the advertisement portion of said via a web-browser user can watch, the advertisement viewable portion of the attention of the user is greater than the minimum prescribed threshold are concentrated in the advertisement,
The part the user can watch one of the web browser to display the advertisement to the user or,
A interaction with the web browser of the user, based on one or more criteria including the interaction that occurs at least once in each of the one or more predetermined time period, the attention of the user with respect to the advertisement 56. The system of claim 55, wherein the system is determined. The advertisement duration of time that the one or more criteria for determining the attention of the user is satisfied with measurement on the duration that is measured is the sum Note the duration of the user with respect to the advertisement, wherein Item 60. The system according to Item 60. Wherein the duration the when one or more of the criteria of at least one but not fulfilled measurement is aborted, the system according to claim 61 to determine the interest of the user with respect to the advertisement. Wherein the measurement of the duration is restarted when one or more criteria are met, the system according to claim 62 to determine the interest of the user with respect to the advertisement.

Images