HK1116565A

HK1116565A - Platform for advertising data integration and aggregation

Info

Publication number: HK1116565A
Application number: HK08106867.9A
Authority: HK
Inventors: 罗伯特．柯林斯; 雷扎．阿里．昌巴里; 保罗．阿包达卡
Original assignee: 雅虎公司
Priority date: 2004-07-30
Filing date: 2005-07-29
Publication date: 2008-12-24

Description

Platform for advertisement data integration and aggregation

Priority application

The present application claims priority from U.S. provisional patent application No.60/592,799 entitled "METHODS AND systems FOR USE IN A COMPUTERIZED SEARCH-BASEDDVERING MARKET" filed on 30.7.2004 AND U.S. patent application No.11/026,517 entitled "PLATFORM FOR ADVERTISING DATAEGRATION AND AGGREGATION" filed on 30.12.2004.

Copyright notice

A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the patent and trademark office patent file or records, but otherwise reserves all copyright rights whatsoever.

Technical Field

The present invention relates generally to advertising (advertising) and, more particularly, to an advertising campaign (advertising campaign) management and optimization system, method, and apparatus.

Background

The success of an advertising campaign depends on the most efficient use of the advertising budget to advertise to maximally affect the audience's behavior. For example, if a campaign is directed to selling a product, an advertiser (advertiser) may attempt to purchase an advertisement with a given budget in order to allow a maximum number of customers to purchase the product. However, determining how to efficiently and optimally spend advertising budgets and utilize such budgets to implement and manage ongoing advertising campaign(s) can pose daunting challenges to advertisers.

Increasingly, advertising campaigns include online or internet-based advertising. With the increasing use of the internet, it is natural that more advertising resources are beginning to target this wide audience. Moreover, internet-based advertising allows advertisers more opportunities to deliver more targeted, more relevant advertisements than traditional offline advertising techniques (e.g., billboards, etc.).

An increasingly important area of advertising includes sponsored listings (sponsored listing). Such a list may be presented, for example, in the form of sponsored links appearing in the results of a search performed on an internet-based search engine (e.g., Yahoo |, AskJeeves, etc.). For example, there are auction-based systems in which advertisers bid online (bid) for being included in sponsored search results for one or more particular search terms (search term), and for the ranking (ranking) or prominence of the placement of their sponsored listings in such results.

Online advertisers participating in such auction-based systems may face challenges: i.e., to manage and optimize bids that may be frequent for each of, for example, thousands or hundreds of thousands of search terms or groups of search terms. In addition, advertisers may need to manage and optimize many advertising campaigns on many disparate portals. In addition, advertisers may need to manage and optimize offline components of one or more advertising campaigns. All of this requires the skill and energy of the advertiser, which may be more appropriate for many other different business tasks.

Existing techniques for managing and optimizing advertising campaigns are inadequate to provide an efficient, effective solution to these problems.

There is a need in the art for systems and methods for managing and optimizing advertising campaigns.

Disclosure of Invention

In some embodiments, the present invention provides a method for facilitating management of an advertising campaign. The method includes one or more ad campaigns facilitation servers of an ad campaigns facilitator obtaining ad campaign information related to the ad campaigns from one or more advertisers. The method also includes one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaigns from one or more advertisers and from each of a plurality of members (affiliates) of the ad campaigns facilitator. The method also includes the one or more ad campaigns facilitation servers storing the ad campaign information and the ad campaign performance information in one or more ad campaigns databases. The one or more ad campaigns facilitation servers utilize at least a portion of the ad campaign information and at least a portion of the ad campaign performance information to facilitate managing the ad campaigns.

In one embodiment, the present invention provides a system for facilitating management of an advertising campaign. The system includes a computer network. The system also includes one or more ad campaigns facilitation servers coupled to the network of ad campaigns facilitators. The system also includes one or more ad campaign databases connected to the one or more ad campaign facilitation servers. The system also includes a plurality of members of the advertising campaign facilitator connected to the network. The system also includes a plurality of advertisers connected to the network; wherein the one or more ad campaigns facilitation servers are adapted to obtain ad campaign information from advertisers related to the ad campaigns; wherein the one or more ad campaigns facilitation servers are adapted to obtain ad campaign performance information related to the ad campaigns from the advertisers and members; wherein the one or more ad campaigns facilitation servers are adapted to store ad campaign information and ad campaign performance information in one or more ad campaigns databases; and wherein the one or more ad campaign assistance servers are adapted to utilize at least a portion of the ad campaign information and at least a portion of the ad campaign performance information to assist in managing the ad campaign.

In another embodiment, the present invention provides a computer usable medium storing program code that, when executed on a computerized device, causes the computerized device to perform a method for assisting in managing an advertising campaign. The method includes one or more ad campaigns facilitation servers of an ad campaigns facilitator obtaining ad campaign information related to the ad campaigns from one or more advertisers. The method also includes one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaign from one or more advertisers and from each of a plurality of members of the ad campaigns facilitator. The method also includes the one or more ad campaigns facilitation servers storing the ad campaign information and the ad campaign performance information in one or more ad campaigns databases. The one or more ad campaigns facilitation servers utilize at least a portion of the ad campaign information and at least a portion of the ad campaign performance information to facilitate managing the ad campaigns.

In some embodiments, the present invention provides a method for integrating ad campaign performance information from multiple disparate sources. The method includes one or more ad campaigns facilitation servers of an ad campaigns facilitator obtaining ad campaign information related to the ad campaigns from one or more advertisers. The method also includes one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaigns from the advertiser and from each of a plurality of disparate members of the ad campaigns facilitator. The one or more ad campaigns facilitation servers store ad campaign information and ad campaign performance information in an integrated manner in one or more ad campaign databases.

In one embodiment, the present invention provides a method for integrating advertisement campaign information from a plurality of disparate sources. The method includes one or more ad campaigns facilitation servers of an ad campaigns facilitator obtaining ad campaign information related to the ad campaigns from one or more disparate advertisers. The method also includes one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaigns from the advertiser and from each of a plurality of members of the ad campaigns facilitator. The one or more ad campaigns facilitation servers store ad campaign information and ad campaign performance information in an integrated manner in one or more ad campaign databases.

In another embodiment, the invention provides a method for facilitating automatic management of an advertising campaign in an auction-based search term related sponsored listings marketplace. The method includes one or more ad campaigns facilitation servers of a marketplace operator obtaining ad campaign information related to an ad campaign from one or more advertisers. The method also includes the one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaign from the one or more advertisers and from each of a plurality of disparate members of the ad campaigns facilitation servers, the ad campaign performance information including information from which the one or more per-lead return metrics can be determined. The method also includes one or more ad campaigns facilitation servers storing the ad campaign information and the ad campaign performance information in one or more ad campaigns databases in an integrated manner. One or more ad campaigns facilitation servers utilize at least a portion of the ad campaign information and at least a portion of the ad campaign performance information to facilitate automatically managing ad campaigns; wherein the one or more ad campaign assistance servers assist in automatically managing the ad campaign includes assisting in automatically implementing a bidding strategy (strategy) for advertisers in the marketplace, and including providing a user interactive interface to allow the one or more advertisers to access and modify at least a portion of the information stored in the ad campaign database.

In some embodiments, the present invention provides an apparatus for providing an interactive advertiser interface to facilitate managing one or more advertising campaigns. The apparatus includes one or more ad campaigns facilitation servers connected to a network of ad campaigns facilitators. The apparatus also includes one or more ad campaign databases connected to the one or more ad campaign facilitation servers. The apparatus also includes a plurality of members of the advertising campaign facilitator connected to the network. A plurality of advertisers connected to the network; the one or more ad campaigns facilitation servers are adapted to obtain ad campaign information related to the ad campaigns from one or more advertisers. The one or more ad campaigns facilitation servers are adapted to obtain ad campaign performance information related to the ad campaigns from the advertisers and members. The one or more ad campaigns facilitation servers are adapted to store ad campaign information and ad campaign performance information in one or more ad campaign databases. The one or more campaign assistance servers are adapted to provide one or more user-interactive applications to allow advertisers to access and manipulate advertising campaigns and advertising campaign performance information to assist in managing the advertising campaigns.

In one embodiment, the present invention provides an apparatus for providing an interactive advertiser interface to facilitate managing one or more advertising campaigns. The apparatus includes one or more ad campaigns facilitation servers connected to a network of ad campaigns facilitators. The apparatus also includes one or more ad campaign databases connected to the one or more ad campaign facilitation servers. The apparatus also includes a plurality of members of the advertising campaign facilitator connected to the network. A plurality of advertisers connected to the network; the one or more ad campaigns facilitation servers are adapted to obtain ad campaign information related to the ad campaigns from one or more advertisers. The one or more ad campaigns facilitation servers are adapted to obtain ad campaign performance information related to the ad campaigns from the advertisers and members. The one or more ad campaigns facilitation servers are adapted to store ad campaign information and ad campaign performance information in one or more ad campaign databases. The one or more campaign assistance servers are adapted to provide one or more user-interactive applications to allow advertisers to access and manipulate advertising campaigns and advertising campaign performance information to assist in managing the advertising campaigns.

In some embodiments, the present invention provides a method for assisting in optimizing an advertising campaign. The method includes one or more ad campaigns facilitation servers of an ad campaigns facilitator obtaining ad campaign information related to the ad campaigns from one or more advertisers. The method also includes one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaign from one or more advertisers and from each of a plurality of members of the ad campaigns facilitator. The method also includes the one or more ad campaigns facilitation servers storing the ad campaign information and the ad campaign performance information in one or more ad campaigns databases. With one or more ad campaign assistance servers and based at least in part on at least a portion of the ad campaign information and at least a portion of the ad campaign performance information, the method includes determining an optimal ad campaign policy for at least a first of the ad campaigns.

In one embodiment, the present invention provides a method for facilitating optimization of an advertising campaign based at least in part on a reward-per-guide metric. The method includes one or more ad campaigns facilitation servers of an ad campaigns facilitator obtaining ad campaign information related to the ad campaigns from one or more advertisers. The method also includes one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaign from one or more advertisers and from each of a plurality of members of the ad campaigns facilitator. The method also includes the one or more ad campaigns facilitation servers storing the ad campaign information and the ad campaign performance information in one or more ad campaigns databases. The method also includes calculating, with the one or more ad campaigns facilitation servers, and based at least in part on at least a portion of the ad campaign information and at least a portion of the ad campaign performance information, one or more per guide return metrics. Based at least in part on the calculated one or more per-lead reward metrics, the method includes determining an optimal ad campaign policy for at least a first of the ad campaigns.

In another embodiment, the present invention provides a computer usable medium storing program code that, when executed on a computerized device, causes the computerized device to perform a method for assisting in optimizing an advertising campaign. The method includes one or more ad campaigns facilitation servers of an ad campaigns facilitator obtaining ad campaign information related to the ad campaigns from one or more advertisers. The method also includes one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaign from one or more advertisers and from each of a plurality of members of the ad campaigns facilitator. The method also includes the one or more ad campaigns facilitation servers storing the ad campaign information and the ad campaign performance information in one or more ad campaigns databases. The method also includes determining, with one or more ad campaigns facilitation servers, and based at least in part on at least a portion of the ad campaign information and at least a portion of the ad campaign performance information, an optimal ad campaign policy for at least a first of the ad campaigns.

In some embodiments, the present invention provides a method for managing the targeted flow of directions from members of an advertising campaign facilitator to an advertiser's website. The method includes one or more ad campaigns facilitation servers facilitating presentation of targeted online advertisements to users of member's websites, the online advertisements including links that enable the users to access the advertiser's website. The method also includes one or more ad campaigns facilitation servers obtaining and storing ad campaign performance information from the members in an ad campaign database relating to the performance of the online advertisements. If a user accesses an advertiser's web site using an online advertisement. The method also includes redirecting the visitor to a website associated with the one or more ad campaigns facilitation servers to collect ad campaign performance information prior to targeting the user to the advertiser's website.

In one embodiment, the present invention provides a system for assisting in optimizing an advertising campaign. The system includes a network and one or more ad campaigns facilitation servers connected to the network for ad campaigns facilitators. The system also includes one or more ad campaign databases, which may be served by one or more ad campaigns facilitation servers. The system also includes a plurality of members of the advertising campaign facilitator connected to the network. A plurality of advertisers are connected to the network. The one or more ad campaigns facilitation servers are adapted to obtain ad campaign information from advertisers related to the ad campaigns. The one or more ad campaigns facilitation servers are adapted to obtain ad campaign performance information related to the ad campaigns from the advertisers and members. The one or more ad campaigns facilitation servers are adapted to store ad campaign information and ad campaign performance information in one or more ad campaign databases. The one or more ad serving assist servers are adapted to determine an optimal ad campaign strategy for at least a first of the ad campaigns based at least in part on at least a portion of the ad campaign information and at least a portion of the ad campaign performance information.

In another embodiment, the invention provides a method for facilitating automatic optimization of an advertising campaign in an auction-based search term related sponsored listings marketplace. The method includes one or more ad campaigns facilitation servers of a marketplace operator obtaining ad campaign information related to an ad campaign from one or more advertisers. The method also includes the one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaign from the one or more advertisers and from each of a plurality of disparate members of the ad campaigns facilitation servers, the ad campaign performance information including information from which the one or more per-lead return metrics can be determined. The method also includes one or more ad campaigns facilitation servers storing the ad campaign information and the ad campaign performance information in one or more ad campaigns databases in an integrated manner. Utilizing one or more ad campaigns facilitation servers and based at least in part on at least a portion of the ad campaign information and at least a portion of the ad campaign performance information. The method includes automatically determining a best advertising campaign policy for at least a first of the advertising campaigns, wherein the operations of automatically determining the best advertising campaign policy include automatically determining a recommended course of action in a future time period for one or more settings for one or more parameters of the advertising campaign policy to be utilized in the future time period.

In another embodiment, the present invention provides a computer usable medium storing program code that, when executed on a computerized device, causes the computerized device to perform a method for assisting in automatically optimizing an advertising campaign in an auction-based search term related sponsored listings marketplace. The method includes one or more ad campaigns facilitation servers of a marketplace operator obtaining ad campaign information related to an ad campaign from one or more advertisers. The method also includes the one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaign from the one or more advertisers and from each of a plurality of disparate members of the ad campaigns facilitation servers, the ad campaign performance information including information from which the one or more per-lead return metrics can be determined. The method also includes one or more ad campaigns facilitation servers storing the ad campaign information and the ad campaign performance information in one or more ad campaigns databases in an integrated manner. Utilizing one or more ad campaigns facilitation servers and based at least in part on at least a portion of the ad campaign information and at least a portion of the ad campaign performance information. The method includes automatically determining an optimal ad campaign strategy for at least a first of the ad campaigns. The method further includes automatically determining an optimal ad campaign policy includes automatically determining a recommended course of action in a future time period for one or more settings for one or more parameters of the ad campaign policy to be utilized in the future time period.

Drawings

The present invention is illustrated in the accompanying drawings, which are exemplary, but not limiting, in which like references are used to refer to similar or corresponding parts, and in which:

FIG. 1 is a block diagram illustrating a distributed system according to an embodiment of the invention;

FIG. 2 is a flow diagram illustrating a method according to one embodiment of the invention;

FIG. 3 is a block diagram of a networked computer system in accordance with one embodiment of the present invention;

FIG. 4 is a block diagram illustrating automatic tag-based data tracking and collection according to one embodiment of the present invention;

FIG. 5 is a block diagram illustrating components of an ad campaign assistance program according to one embodiment of the present invention;

FIG. 6 is a block diagram of a system according to one embodiment of the invention;

FIG. 7 is a flow diagram illustrating a method according to one embodiment of the invention;

FIG. 8 is a graph of conversion rate versus time for a hypothetical search term or phrase group, according to one embodiment of the present invention;

FIG. 9 is a diagram of a hypothetical purchase cycle, in accordance with one embodiment of the present invention; and

FIG. 10 is a simplified screenshot, according to one embodiment of the present invention.

Detailed Description

In the following description of the preferred embodiment, reference is made to the accompanying drawings which form a part hereof, and in which is shown by way of illustration the specific embodiment in which the invention may be practiced. It is to be understood that other embodiments may be utilized and structural changes may be made without departing from the scope of the present invention.

Herein, the term "advertiser ad campaign sct" includes a set of one or more ad campaigns for a particular advertiser or advertising entity. The term "advertising campaign" includes one or more advertising actions or behaviors directed to achieving a common advertising goal, such as marketing or sales of a particular product, service, or content or group of products, services, or content. Two advertising campaigns are considered distinct from one another if each of the two advertising campaigns are targeted to different advertising targets.

The term "means (tact)" includes a particular form or type of advertisement. For example, in online advertising, the means may include sponsored search result listings, banner advertisements, and so forth. In offline advertising, the means may include television advertising, radio advertising, newspaper advertising, and the like. In different embodiments, a tool may be more or less broadly defined to include a subset or superset of the listed examples or other examples. For example, online advertising is an example of a broader approach than the narrower approach of sponsored search result listings.

The term "channel" includes a particular entity, organization, etc. through which advertising may be conducted. For example, in the context of online advertising, channels may include information such as MSN, CNN, Yahoo! Such as a web site or a search engine. Herein, the term "computer" includes, for example, a desktop computer, a notebook computer, or a computerized device, such as a handheld computerized device or a cellular telephone.

Here, two members, advertisers, or sources of information are considered distinct from each other if any two members, advertisers, or sources of information (e.g., an ad campaign or ad campaign performance information) use different platforms, programs, applications, hardware, software, or data storage technologies for information collection, storage, or transmission, such that the ad campaigns facilitation server 102 (shown in FIG. 1) must employ different technologies or sets of technologies for the programs or applications to receive, identify, parse, or store information from each of the two members, advertisers, or other sources.

Herein, the term "search term creation" includes bid objects related to a search term, such as a search term, a collection or a group of search terms, in an auction-based search term related sponsored listing marketplace. Originality includes any rules that specify conditions related to a search term or group that will result in authorization to display an advertisement or sponsored listing.

Some embodiments of the present invention may be used with the features or techniques described in U.S. patent application No.10/072,220, entitled "AUTOMATIC FLIGHT MANAGEMENT IN AN ONLINEARKETPLEACE" filed on 8.2.2002, the entire contents of which are hereby incorporated by reference.

Fig. 1 is a block diagram illustrating a distributed system 100 according to an embodiment of the present invention. The system 100 includes an ad campaigns facilitation server computer 102 (which in some embodiments may include multiple server computers), a plurality of members 104, 106, 108, a plurality of advertisers 110, 112, 114, a plurality of users 128, 130, 132, and a plurality of channels 116, 118, 120. The channels 116, 118, 120 shown are part of a conceptually represented means 122, the means 122 is part of a conceptually represented ad campaign 124, and the ad campaign 124 is part of a conceptually represented set of advertiser ad campaigns 126. The set of advertiser ad campaigns 126 includes other ad campaigns 127, 128, which may include other means (not shown) and channels (not shown). Other sets of advertising campaigns 118, 120 are also shown, which may include advertising campaigns (not shown), instruments (not shown), and channels (not shown).

The ad campaigns facilitation server computer 102 (hereinafter "server 102") includes a Central Processing Unit (CPU)130 and a data storage device 132. Additionally, each of the members 104, 106, 108 and advertisers 110, 112, 114 and some or all of the users 128, 130, 132 include at least one computer having a central processing unit (not shown) and a data storage device (not shown) that may include one or more browser programs, such as an Internet browser program.

Some or all of the members 104, 106, 108 may include or be connected to a database. As shown, members 104 and 108 are connected to databases 134 and 136, respectively.

Although a network is not shown, some or all of the computers may be connected by one or more computer networks, such as the Internet, as well as one or more wide area networks, local area networks, personal area networks, and the like.

Although the users 128, 130, 132 are all shown connected to the member 108, it should be noted that some or all of the users 128, 130, 132 may not be electrically connected, such as users of readers of magazines as members.

Although three users, members, advertisers, instruments, channels, ad campaigns, and sets of ad campaigns are shown for simplicity, it should be understood that fewer or more users, members, advertisers, instruments, channels, ad campaigns, and sets of ad campaigns may exist.

Each data storage device may include a different amount of RAM for storing computer programs and other data. In addition, each computer may include other components typically found in a computer, including one or more output devices (e.g., a monitor), other fixed or removable data storage devices (e.g., a hard disk, floppy disk drive and CD-ROM drive), and one or more input devices (e.g., a mouse pointing device and a keyboard).

Generally, each computer works and executes computer programs under the control of an operating system, such as Windows, Macintosh, UNIX, and the like.

Generally, the computer program of the present invention is tangibly embodied in a computer-readable medium, e.g., in one or more data storage devices attached to a computer. Under the control of an operating system, computer programs may be loaded from the data storage device into the computer RAM for subsequent execution by the CPU. The computer program includes instructions which, when read and executed by a computer, cause the computer to perform the steps necessary to execute the elements of the invention.

The data storage device 134 of the server 102 includes an ad campaign facilitator 134 and an ad campaign database 136. The ad campaign facilitator 134 broadly represents all programs, software, tools, applications, Application Program Interfaces (APIs) and other tools that may be used to implement methods according to embodiments of the present invention, including methods associated with management or optimization of ad campaigns. Although the ad campaigns facilitation program 134 is shown as being located at the server 102, in some embodiments, elements or components of the ad campaigns facilitation program 134 may be located elsewhere, such as at a computer associated with a member, advertiser, or channel, to facilitate communications between the server 102 and other entities or computers.

In some embodiments, the server 102 is owned, controlled, or operated by an advertising campaign facilitator, such as an entity or company that facilitates the planning, management, optimization, delivery, transmission, or implementation of an advertisement or advertising campaign. In some embodiments, an advertising campaign may include a sponsored search result list or link. An auction-based system or marketplace may be used by advertisers to bid on search terms or groups of terms that, when used in a search, will have their ad listings or links displayed in the display results. Advertisers may also bid for the position or prominence of their listings in the search results. In such embodiments, the event facilitator is or includes a marketplace operator, which may, for example, control, operate or manage the auction-based system, and the like.

While the server 102 may be used to facilitate placement in connection with advertisement presentation, it should be noted that in some embodiments, the server 102 (and associated ad campaign facilitator) does not place or facilitate placement of the advertisement presentation. For example, in some embodiments, the server 102 may be used to facilitate management or optimization of an advertising campaign, or to automatically facilitate management or optimization of an advertising campaign, without actually placing the presentation of the advertisement itself.

More details regarding auction-BASED systems and MARKET operators as described above, and aspects thereof, may be found IN U.S. patent application No.10/625,082 entitled "TERM-BASED condition marker" filed 7/22/2003, U.S. patent application No.10/625,000 entitled "condition variance IN a TERM-BASED condition marker" filed 7/22/2003, and U.S. patent application No.10/625,001 entitled "TERM-BASED condition marker" filed 7/22/2003, all of which are hereby incorporated by reference IN their entirety, which are filed 7/22/2003 and which belong to the same applicant as the present application. In some embodiments, systems and methods associated with ad campaign management benefit optimization in accordance with the present invention may be implemented in conjunction with the methods and systems described in these listed incorporated by reference applications.

Each of the set of advertiser ad campaigns 126, 118, 120 represents a set of one or more ad campaigns for a particular advertiser (e.g., one of the illustrated advertisers 110, 112, 114). The members 104, 106, 108 represent entities, organizations, or companies associated with or affiliated with the advertising campaigns facilitator or server 102 in any manner. Members may include entities associated with the ad campaign facilitator or the server 102 in the sense of only: i.e., arranged in some manner to facilitate transmission of advertising campaign performance information to server 102; in addition, no more membership or associated entities are required to be considered members.

Advertisements may be presented by members (or their distribution pathways, portals, media, companies, etc.). Offline members include entities through which various offline advertisements may be presented in conjunction with, such as television stations, radio stations, newspaper or newspaper organizations, magazines or magazine organizations, and so forth. An online member includes an entity through which internet-based or internet-accessible advertisements may be presented in connection with, for example, such as Yahoo! A search engine such as Ask Jeeves, an e-commerce site, or other web sites such as news or content providing web sites, sports web sites.

The members may be distinct from each other. For example, the server 102 may need to employ a different program or application in order to process, reformat or otherwise translate ad campaign performance information received from disparate members and store that information in the ad campaign database 136.

Members may differ in the type of ad presentation or ad presentation medium they control. In addition, they may differ in the way or platform in which they format, store, and transmit information (including hardware, software, programs, databases, or applications used for these purposes). They may also be different in terms of any data or combination of data that it collects or stores about an advertisement, performance of an advertisement or advertising campaign, audience for an advertisement (e.g., a member's website or user of a search engine), etc.

Advertisers may include, for example, entities, individuals, companies, organizations, etc. that place advertisements with the advertising campaign facilitator to be presented by the member (e.g., advertisements in the member's newspaper or sponsored listings that appear in a search result set obtained via the member's search engine or website). In some embodiments, the advertisers may also (or some of them may) be distinct from each other.

Users (e.g., the illustrated users 128, 130, 132) are users or audiences of resources, media, distribution pathways, etc. associated with the members to which the advertisement is presented by the members. For example, the user includes a reader of a member's newspaper or a computer user using a member's search engine or browsing the member's website.

Server 102 facilitates management or optimization of an advertiser's ad campaign or set of ad campaigns, or facilitates automatic management or optimization of ad campaigns, and may facilitate placement of ad presentations by members. The server may also be used to facilitate the storage, organization, and management of information sent to the server 102 by members, including members and advertisers.

Although the illustrated members 104, 106, 108 may be offline (e.g., newspapers) or online (e.g., websites), each of the illustrated members 104, 106, 108 includes at least one computer capable of communicating with the server 102, although in some embodiments one or more members may not be electrically connected to the server 102 and may transmit, non-electronically, information that will ultimately be stored electronically in the server 102. Each of the members 104, 106, 108 can send or transmit information to the server 102. It should be noted that while the ad campaigns facilitation program 134 is shown at the server 102, it may comprise a component (e.g., a program) located elsewhere, including a program, software, or application located at or executed by a member's computer, such as an HTML tag related program, which is described further below.

The data sent from the member's computer to the server 102 may be obtained by the server and stored in an integrated manner in the ad campaign database 136, meaning that all of the data is stored together as a whole, so that the meaning of the data (including any and all subsets of the data) can be identified regardless of source. The ad campaigns facilitation program 134 may be used to parse, reformat, analyze, or otherwise process the data sent from the members as needed for integration purposes using methods known to those skilled in the art. Communication between a member and the server 102 may be facilitated by a shared or complementary program, application, or interface between the member and the server 102. For example, in some embodiments, the member's computer utilizes an Application Program Interface (API) to communicate with the server computer 102 or its programs or applications.

In some embodiments, members (e.g., members 104, 108) store data in their associated databases 134, 136, which may include ad campaign performance information and user information, among other things.

The ad campaign assistance information may include a variety of information, statistics, or metrics indicative of or suggestive of the performance or achievement of an ad, channel (or one or more ads presented through a channel, etc.), instrument, campaign, multiple campaigns, components or aspects of a campaign, etc. For example, ad campaign performance information may include information regarding the frequency with which sponsored listings cause a member's website to be presented or clicked on, or cause a user to visit a linked web page or make a purchase at a linked web page, etc.

For example, ad campaign performance may include one or more metrics that provide an indication of a value per guide (valupper lead). For example, such metrics may provide an indication of how many or a percentage of the clicks on the sponsored listing actually resulted in any sort of return to the advertiser. Such rewards may depend on the particular advertiser and the advertiser's business goals. For example, if an advertiser is attempting to sell a product, service, or content, the reward may include a purchase at the advertiser's website that is caused by or attributable to the lead. However, rewards are not limited to sales. The reward may be anything valuable to the advertiser from the actions or actions of the advertiser's website attributable to the guided visitor. Thus, the term "reward per lead" as used herein includes any type of reward caused by or attributable to a lead. Additionally, as used herein, a "per-pilot reward metric" includes any metric that provides a measure, indication, or hint regarding per-pilot rewards.

A particular advertiser may have different business goals and may specify its business goals in different ways. For example, some advertisers may specify their business goals with CPA (cost per acquisition) goals. For such advertisers, conversion rate may be used as an appropriate reward-per-guide metric. In addition, some advertisers may target business by ROAS (return on ad spend). For such advertisers, revenue per guide may be an appropriate reward metric per guide. Some advertisers may specify business goals using a mix or combination of metrics or metrics for which a mix or combination of metrics per lead reward may be appropriate.

Some embodiments of the invention are specifically described herein with reference to conversion ratios. However, it should be understood that this is illustrative and that slew rate is only one of many possible return-per-pilot metrics. Thus, embodiments of the invention described with reference to conversion rates are not limited to using metrics related to conversion rates, but may use or incorporate other or additional reward-per-guidance metrics. In addition, some embodiments of the present invention are specifically described herein with reference to business objectives expressed in terms of ROAS. It should be understood that this is also exemplary and that other or additional business objective specifications or metrics may be used in different scenarios.

In some embodiments, the present invention provides methods for facilitating automatic management or optimization of one or more advertising campaigns. This may include utilizing business rules that are specific to or specified by a particular advertiser, as well as using its business results or metrics, which may include ad campaign performance information or metrics for certain aspects thereof. In some embodiments, the present invention includes combining business rules with aggregated real-time business results or metrics thereof to facilitate automatic, dynamic, real-time management or optimization of advertising spending.

The advertiser business rules may be explicitly defined, such as by the ad campaign facilitator 134 and by ad campaign performance information (which may include, for example, ad outcome metrics), or may be implicitly defined, inferred, deduced, or obtained. Additionally, in some embodiments, business rules may be automatically modified, or modifications may be automatically recommended for review and approval by advertisers prior to implementation. In some embodiments, ad campaign performance information is automatically analyzed by the ad campaign facilitator 134, and based on this analysis, business rules may be obtained, modified, or optimized to obtain maximum advertiser interest.

Tracking and collecting of advertisement performance information may be accomplished, for example, using HTML tags for the advertiser's web site, as will be further described below with reference to FIG. 4. Ad campaign performance information may be obtained from members as well as advertisers. For example, in some embodiments, an advertising campaign

The user information may include information obtained and stored by members (or channels), including user profiles, historical user behavior information, and the like, or may be sent to the server 102 from members or other entities and stored in the ad campaign database 136. Further description of user information and its use may be found in U.S. patent application nos. 60/546,699 and 10/783,383, previously incorporated by reference.

The data, or a portion thereof, obtained and stored by the members and advertisers is sent to the server 102, converted or reformatted as necessary into a format usable and storable in the ad campaign database 136, and stored therein. Alternatively, the data may be converted or reformatted prior to transmission or otherwise manipulated to allow for appropriate storage in the ad campaign database 136. Some members or advertisers may send user profile, user behavior, or user history data directly to the server 102 without storing it in a non-volatile form in a database associated with the member, or may even send the data in a non-electronic format, such as an advertising campaign facilitator, after which the data may be converted to an electronic format and stored in the server 102.

Each of the set of advertiser ad campaigns 126, 118, 120 is associated with one of the advertisers 110, 112, 114. For example, an advertiser may wish to advertise several products for sale. An advertiser may have a set of advertising campaigns that includes campaigns associated with advertisements for each product. Each activity may utilize a number of means. For example, one means of exploitation may be to sponsor a search result list. For this approach, advertisers may utilize multiple channels. For example, an advertiser may utilize sponsored search listings in several websites or portals (e.g., Yahoo!, MSN.com, etc.).

It should be noted that a channel may be or include or be associated with a member. For example, an advertiser may place an advertisement to be presented on msn.com, which is thus the channel for advertisement presentation. Com may be a member at the same time. In addition, since the member may be a channel, the information transmitted by the member may also be transmitted by the channel, as shown in fig. 1.

The data obtained by members and advertisers may include information that may be very useful for managing or optimizing advertising campaigns. For example, advertising campaign performance or user information obtained by an advertiser or member through a user's distribution pathways, portals, or media provided by the advertiser's website or by the member may provide a rich source of information that may be used, analyzed, or mined to determine likely future performance of an advertisement in various contexts, for various users, at various times, and so forth. An advertising campaign facilitator using the server 102 is in a central location that is advantageous to obtain, collect and utilize, or facilitate utilization of data from many members and advertisers.

FIG. 2 is a flow diagram illustrating a method 200 according to one embodiment of the invention. At step 202, ad campaign information from the advertiser is obtained by the server 102 and stored in the ad campaign database 136 using the ad campaign support program 134 (shown in FIG. 1). In some embodiments, the advertising campaign information may be provided partially or completely from one or more entities other than the advertiser. The ad campaign information may include parameters or details of the ad campaign. For example, ad campaign information may include campaign goals or budget-related conditions or constraints, or may include specifying, defining, or describing the ad itself, channels, means, and so forth. For an auction-based sponsored search result listing, the advertising campaign information may include bid parameters, such as a highest or lowest bid or bid position (rank or prominence of the listing) associated with a word or word cluster, as will be described further below. Such member information may also include expressed campaign goals, quotas or objectives, such as campaign goals, quotas or objectives by metrics such as ROAS (return on advertisement spend), CPI (click per impression), or by other metrics and for individual advertisements, words or groups of words, channels, means, and the like, as will be described further below.

At step 204, ad campaign performance information is obtained by the server 102 from the members (or channels) and the advertisers (or any of the members or advertisers) using the ad campaign facilitator 134 and stored in the ad campaign database 136. The advertising campaign performance information may include a variety of information related to the historical performance of the advertising campaign, channel, means, or advertisement or group of advertisements. The advertising campaign performance information may include various types of information that indicate or suggest how an advertisement or an advertisement presented through a particular channel, etc., is or may affect the behavior of a user or customer effectively. For example, such as Yahoo! Such members may collect performance information regarding a particular sponsored search result list. This information may include the number or percentage of viewers clicking on the link, or who purchased or purchased the product at the advertiser's website because of the listing, and so on.

In some embodiments, to facilitate tracking and collection of performance information for various advertising campaigns, HTML tags are inserted into the advertiser's website or a different page thereof (as will be described in more detail with reference to FIG. 4). In this case, the tagging may be facilitated by the ad campaign support program 134 and the tagging program or application may be considered part of it, wherever it is located and by whom it is used. Further, ad campaign performance information and other information in the ad campaign database 136 may be updated periodically or continuously as new or updated information is obtained.

It should be understood that obtaining ad campaign information and ad campaign performance information includes any necessary reformatting or conversion of the data by methods known to those skilled in the art to achieve the purpose of obtaining and storing data from disparate sources and disparate members.

Although not included in method 200, in some embodiments, user information is also obtained from members or advertisers. The user information may include user profile information, user behavior information, and the like. Such information may be useful, for example, for targeting users for advertising, such as described in detail in U.S. patent application nos. 60/546,699 and 10/783,383, previously incorporated by reference.

At step 206, the obtained information, including ad campaign information, ad campaign performance information, and possibly other information (e.g., user information), is analyzed using the ad campaign facilitator 134 to facilitate determining or determining an optimal ad campaign strategy. Herein, a "best" ad campaign policy includes any ad campaign policy determined to be best or superior to other policies, determined to be possibly best, predicted or expected to be best or possibly best, etc. In some embodiments, the optimization is performed against parameters or combinations of parameters specified by an advertiser, automatically or partially automatically provided by an ad campaign facilitator, or otherwise provided.

Additionally, an "ad campaign policy" includes any process or some aspect or combination of processes of action (e.g., including changing or not changing current settings or policies) or behavior related to an ad campaign. The ad campaign policy may include a recommendation regarding the course of action related to one or more aspects or parameters of the ad campaign, and may include an immediate course of action or a set of parameters, or a set of courses of action or parameters in a specified time window. For example, in the context of an auction-based search result list scenario, the best advertising campaign strategy may include recommendations for bids and bid hiding rates, which may be combined with an auction or marketplace for search terms or groups of terms, which may in turn be combined with sponsored listings.

At step 208, the ad campaign facilitator 134 is used to facilitate managing or managing an ad campaign (or set of ad campaigns) for or on behalf of an advertiser. In some embodiments, the ad campaign support program 134 facilitates the automatic management of an ad campaign or set of campaigns. As used herein, "managing" includes any of a variety of activities related to supervising one or more advertising campaigns or certain aspects thereof, and making or implementing actions or behavior decisions related to one or more advertising campaigns or certain aspects thereof. For example, in some embodiments, advertisers are provided with one or more user-interactive computer applications to allow access, manipulation, and searching of information in an ad campaign database (e.g., relational database searches) related to the performance of one or more ad campaigns, or certain aspects thereof. For example, an advertiser may specify parameters related to the performance of an advertising campaign by requesting to view, obtain reports on, obtain summaries of, or even download information regarding the performance of a particular advertisement, a particular advertising channel, a particular campaign or campaign element, and so forth. In the context of an auction-based sponsored search results list, this may include obtaining a summary of advertising performance or advertising campaign performance in conjunction with certain means or channels based on a particular search term or group of terms. The ad campaign database 136 may contain a large accumulation of information about ad campaign performance from disparate member and advertiser sources and is therefore useful in this regard.

Ad campaign management may also include managing or automatically managing the ads themselves, such as by deleting or introducing new ads or listings, modifying or changing ads, etc., all of which may be stored in the ad campaign database 136.

Additionally, ad campaign management may include adding a campaign or set of campaigns from new advertisers, or determining information related to the use of the ad campaign facilitator 134, such as which advertisers logged in at a given time, and the like. Such actions may be limited to individuals associated with or employed by the advertising campaign facilitator, or to an administrator of the server 102, for example.

Managing an ad campaign may also include implementing or automatically implementing an ad campaign policy or action. For example, in the context of an auction-based sponsored search results list, this may include implementing a bidding strategy.

In some embodiments, ad campaign management may include implementing or automatically implementing the determined optimal ad campaign policy. The optimal ad campaign strategy may be determined automatically or partially automatically using an ad campaign facilitator. Once determined, the ad campaigns facilitation program may be used to automatically implement or partially automatically implement such a strategy. Examples and embodiments regarding this in the context of an auction-based sponsored search results list are described below.

It should be noted that in some embodiments, it is either ad campaign management or ad campaign optimization that is assisted, but not both.

It is also noted that in some embodiments, advertising campaigns may be facilitated for or on behalf of other entities in addition to advertisers, such as advertising companies associated with advertisers.

Many of the following description relates to embodiments of the present invention relating to sponsored search result listings, auction-based sponsored search result listings markets, and related contexts. However, it should be understood that the methods and systems described in this context may also be applied in a variety of other contexts, including other online contexts as well as (in some cases) offline contexts.

In some embodiments, advertisers place HTML tags on relevant web pages of their websites to allow automatic tracking of ad performance or user behavior information to be sent to the server and stored in the ad campaign database 136. For example, HTML tags may be used to track a user's visits to, interactions with, or purchases made at advertiser's websites as a result of the user clicking on sponsored links associated with the advertiser.

FIG. 3 is a block diagram of a networked computer system 300 according to one embodiment of the invention. As shown, the Internet 302 connects one or more 324 with a plurality of website-based members 304, 306, 308, a plurality of website-based advertisers 310, 312, 314, and a plurality of users 318, 320, 322. The marketplace operator server 324 may be or include one or more ad campaigns facilitation servers 102 (shown in fig. 1). Members 304, 306, 308 are shown as including MSN304, Yahoo 306, and new york times 308, and include associated websites or search engines. Advertisers 310, 312, 314 are shown to include a product advertiser 310, a service advertiser 312, and a content advertiser 314. Advertisers 310, 312, 314 include advertiser websites at which visitors or customers may perform actions such as purchasing products, services, or content. Visitors to an advertiser's web site include directions obtained from, for example, sponsored listings (targeted directions) and other traffic.

The users 318, 320, 322 are presented with advertiser advertisements, such as sponsored listings, while visiting the web page of one of the members 304, 306, 308. In some embodiments, the marketplace operator utilizes the marketplace operator server 304 to assist in the placement of advertiser advertisement presentations.

Communications between the members 304, 306, 308 and the marketplace operator server 324, and between the advertisers 310, 312, 314 and the marketplace operator server 324, may be facilitated using the APIs 336, 338, 340, 342, 344, 346. In some embodiments, an API, such as an XML-based API, may provide an interface with an ad campaign database to allow for changes related to ad listings themselves or related to bids, for example, or orders or offers in search term related auctions 326 to be provided.

As shown, a marketplace operator server 324 is used to provide or assist in providing the virtual marketplace 316 (or set of virtual marketplaces). The marketplace 316 may include a search term related auction 326 in conjunction with a sponsored search result list that will be presented to users of the member search engine along with search results when the users use a particular search term, group of search terms, etc. in a search. The marketplace 316 may also include an offer exchange for facilitating placement of advertisements between members and advertisers, including suggestions and matches corresponding to member and advertiser offers. Additional features and details regarding the marketplace 316 and its components, including the offer exchange 328, may be found in U.S. application nos. 60/546,699 and 10/783,383, previously incorporated by reference.

The marketplace operator server 316 also includes an ad campaign facilitator and ad campaign database that is used to provide the ad campaign facilitator 330 to, for example, advertisers 310, 312, 314. As shown, these tools include an ad campaign optimization tool 332 and an ad campaign management tool 334.

FIG. 4 is a block diagram illustrating automatic tag-based data tracking and collection according to one embodiment of the present invention. In general, according to some embodiments, tags and labels may be used to facilitate automatic tracking of metrics including or about leads obtained via sponsored listings and further user actions including conversions generated by such leads and revenue generated by such conversions. This information is valuable for advertisers or other website operators to evaluate or analyze or allow evaluation or analysis of the performance of sponsored listings and thus enact strategies with respect to their sponsored listings or bids. Additionally, in some embodiments, the collected information may be used by an advertising campaign assistance program (e.g., including a bid optimizer and bid manager as shown in FIG. 5) according to some embodiments of the present invention to perform such analysis and policy making.

Some embodiments of the invention utilize or may incorporate features and techniques such as HTML tagging, data tracking, and related techniques described in the following patent applications: U.S. patent application No.09/832,434, filed on day 10/4/2001 and entitled "SYSTEM AND METHOD FOR MONITORING USER WITH A WEBDOMAIN" and U.S. patent application No.09/587,236, filed on day 2/6/2000 and entitled "SYSTEM AND METHOD FOR MONITORING USER WITH Web PAGES", both OF which are hereby incorporated by reference in their entirety.

Internet-based traffic 410 is shown visiting advertiser's web page 404. Traffic 410 includes a lead 402, which is a hit on web page 404 due to a user clicking on the advertiser's sponsored search result list, as well as other non-lead traffic 412. After accessing the initial web page 404, the visitor may then click on the link to go to another page or pages associated with the web site, such as the illustrated pages 406 and 408. At some point, the user may, for example, place the goods in a shopping cart or actually make a purchase. The process by which a user goes deep into an advertiser's (or other entity's) website eventually reaching the end of a purchase (in some cases) is referred to as a funnel 414. As shown, a tab 416 is included on the advertiser's web page (or selected such page).

In some embodiments, the HTML tags 416 facilitate automatic tracking, collection, and use of traffic and collection of information that is subsequently sent to the server 102, e.g., via the internet, and stored in the ad campaign database 136. With tags, the lead can be distinguished from other traffic, and depending in part on the configuration of the advertiser's web page, the tracked information sent to the server 102 can include the number, frequency and time of hits on various web pages, the deepest stage of entry into the funnel for a particular lead, whether shopping was made, whether purchases were made, the type or amount of purchases, and other information. In some embodiments, advertisers are helped mark or equip their websites or webpages via applications provided with the ad campaign facilitator 134 (shown in FIG. 1).

In some embodiments, new pages added to a site are automatically appropriately marked after an advertiser (or other website operator) initially arms.

In some embodiments, the tag facilitates communicating the transaction ID value to the server 102. The transaction ID value is a unique value generated as a result of a user's action (e.g., a shopping action) at the advertiser's web site. The transaction ID value can facilitate distinguishing between multiple stores and transition events occurring within a single browser session. For example, if a second transition event for the same amount of revenue within a single browser session is detected, it may not be apparent whether such a purchase actually occurred or whether the visitor just refreshed or returned to the web page with the transition tag. However, generating a new transaction ID value for the second transaction confirms that the second translation has occurred. In embodiments that do not use a transaction ID value, the assumed limit may be utilized: one shopper and one conversion per browser session.

In some embodiments, tagging includes placing a generic tag in the header of all web pages. In addition, the conversion label is placed over the generic label on the transaction completion page (e.g., thank you page or purchase confirmation page). The generic tags are comprised of code for capturing any customer-specific information associated with the tracked HTML pages. The generic tag calls a piece of JavaScript called Instrumentation Script (Instrumentation Script) and marks out the page that the advertiser wishes to track. In some embodiments, the length of the equipment script is about 6 KB. Additionally, in some embodiments, user actions are instrumented script collections and sent to the server 102 with 1x1.gif image requests. The equipment script (which may be part of the ad campaign facilitator 134) is provided from the server 102 (or one of multiple servers 102 that may be located in many different geographic locations, including perhaps locations around the world). The equipment script is downloaded into the visitor's browser only when the first page of the session is loaded. After the first load, the browser caches the script, eventually generating a cookie. The script will not be downloaded again unless the user flushes its browser cache.

The generic tag also identifies and collects statistics for the page in which it is embedded. When the browser leaves the marked-up page, the instrumented script is paused and no more data is collected due to the inherent security aspects of JavaScript. Once the arm script is activated within the browser, the collected data is sent via a 1X1 pixel gif image request.

For each page view, the instrumented script returns two data packets: one packet is returned when the page is loaded and one when the page is unloaded (i.e., when the visitor moves to the next page). For each page, a total of approximately 500 to 800 bytes are sent. Each data transmission occurs entirely in the background with no impact on the visitor, even for those visitors with slow modem connections. In some embodiments, each data transmission takes on average about 0.21 seconds to reach server 102. If data transmission fails to occur, the arm script is paused and no more data is collected.

In some embodiments, additional tags are utilized. For example, a shopper tag may be used to indicate that a visitor has visited a page indicating that the advertiser believes the visitor is a shopper. In the absence of a shopper tag, a default rule can be used that specifies that a site visitor transitioning from an unprotected page to a protected page indicates that the visitor is a shopper.

In some embodiments of the present invention, in the context of an auction-based search result list, the ad campaign assist program 134 is used to optimize and manage bidding strategies in the auction, bidding by advertisers in connection with search terms, word groups, and the like.

In one embodiment of the present invention, the ad campaigns facilitation program 134 comprises a set of software and programming tools that comprise applications accessible by advertisers via the Internet. The software toolset is provided by an advertising campaign facilitator, which is also a market operator of the auction-based sponsored search result listing marketplace.

FIG. 5 is a conceptual block diagram illustrating an ad campaign assistant program 502 and some of its conceptual components or modules according to one embodiment of the present invention. The ad campaign facilitator 502 comprises a set of software and programming tools available to advertisers via the internet, referred to as marketing console tools 504. Marketing console tool 504 includes a search optimizer tool 506 (or simply search optimizer 506). Search optimizer 506 includes, among other things, a bid optimizer program 508 (or simply bid optimizer 508), a bid manager program 510 (or simply bid manager 510), and a bid hiding engine 512. Although the bid hiding engine is shown separate from the bid optimizer 508 and bid manager 510, in some embodiments, the bid hiding engine may be part of the bid optimizer 508 or bid manager 510 or both, or may be partially or completely separate therefrom.

In some embodiments, the search optimizer 506 or components thereof may include or allow configuration by a user (e.g., advertiser) to allow the user to calibrate or set the tool according to the user's specific and unique business goals. For example, the user may make specific decisions about how to mark their web page (as described in more detail previously with reference to FIG. 4) to accommodate the user's business logic and business goals.

The advertiser uses the marketing console tool 504 to assist in optimizing, managing, or both optimizing and managing an advertising campaign or set of advertising campaigns. Marketing console tool 504 may automatically assist these actions after being provided any necessary parameters or ad campaign information by the advertiser, or may partially automatically assist these actions after being provided decision making input from the advertiser, or may assist the advertiser in analyzing ad campaign performance to optimize an ad campaign and assist advertiser management, including making decisions and implementing ad campaign management strategies.

Search optimizer 506 may also include a user interactive interface program 514 to allow, for example, a user to access and change information stored in the ad campaign database (more details regarding user interfaces are given with reference to FIG. 10).

It should be noted that while the roles of the bid optimizer 510 and bid manager 512, as their names suggest, may include assistance or performance in optimizing and performing an advertising campaign, respectively, their roles are not limited to such functionality, may not perform all aspects of such functionality themselves, and their roles incorporating such functionality may overlap or partially overlap.

In some embodiments, as mentioned with reference to FIG. 5, a marketplace operator provides a virtual marketplace (which may include many marketplaces) or the like that can help advertisers obtain targeted guidance. An internet user may indicate what they are looking for each time they use a search engine. Both advertisers and internet users benefit when product information related to a search is supplied.

The marketplace operator may, for example, be associated with a worldwide network of search engine members (and possibly other members as well) participating in the marketplace, including Yahoo | and MSN, as well as other more localized portals and search engines. Two important features of the market operator network for participating members are the relevance of the results and the time required to complete the search request.

In some embodiments, when an internet user performs a search, the portal sends a request to the marketplace operator server to retrieve paid search results (or listings) that are or are proven to be likely to be relevant to the user's search. In parallel with the request for paid results, the portal sends a separate request to an "algorithmic" search engine to retrieve results found from the internet ranked by relevance. The algorithmically determined list is displayed in order of relevance, and the accounting results are displayed in order of bid position, relevance, or both. For paid search results, the marketplace operator hosts an auction for each search phase and ranks the results based on bids.

The marketplace operator can ensure relevance of the advertiser's listing or some of them through a rigorous human editorial review that is conducted before the listing can participate in the auction. Editorial review may be used, for example, to ensure that a sponsored listing sufficiently corresponds to an associated search term or group of terms, such as to ensure that the title, description in the listing, or content of the linked web page corresponds. In some embodiments, editorial review may be limited to such search terms or groups of terms: such search terms or groups of terms are used most frequently and generate the most traffic (or "high speed" terms, discussed in more detail below) and are therefore considered important enough to justify efforts and expenses. While human editorial review can be expensive and time consuming, it can be the only way to ensure high relevance in sponsored listings, which can motivate greater confidence in users of such links and users of the websites or search engines that provide them.

The market auctions in each market are continuously or frequently updated. Advertisers with listings authorized to participate in the auction may make any, frequent changes to their bids, as well as having listings online and offline. When a set of search results is requested by a member, the current or up-to-date status of the auction determines the list to be served. If an Internet user clicks on one of the listings supplied by the marketplace operator, the HTTP request goes to the marketplace operator server, the advertiser is billed for the click, and the Internet user's browser is redirected to the relevant page on the advertiser's website. For example, in some embodiments, the advertiser may be billed $0.01 higher than the next lower bid in the auction, ranging from a minimum of $0.10 to a maximum of the advertiser's bid. In the case of a tie (equal bid amounts from multiple advertisers), the listings may be ranked in the order in which the bids occur. When a bid tie occurs, all but the list of last bids are paid the full bid amount for each click.

Some market auctions are stable, while others have many advertisers who continually strive to position, join bid battles, etc. Some advertisers do not change their bids as often, while others change their bids as often as possible.

Bid changes can be implemented in different ways. In some embodiments, bid changes are implemented manually by a marketplace operator web application, or with a software program that automates this process through an API, such as an XML-based API that may allow communication with a marketplace operator server and changes data in a database (e.g., the ad campaign database 136 shown in FIG. 1).

In some embodiments, when an advertiser changes a bid associated with a listing, the new state of the auction must be available to all computers (or servers) that supply search results for that marketplace. As mentioned above, the response time of search feeds may be critical, so the computers to which these results are to be fed are replicated around the world as close as possible to the server requesting the advertisement of the search results in order to minimize network latency. The distributed nature of search offers places a burden on the marketplace operator infrastructure, i.e., the replication of all bid updates to all relevant search offer sites in near real-time. Replication of bid updates has a measurable cost in terms of infrastructure, bandwidth, and labor to support the system.

Due to the costs, system requirements, and potential delays associated with replication or over-replication, in some embodiments, advertisers are limited by the total number or frequency of bid updates associated with the advertiser, the advertiser's active set, or one or more costs thereof. For example, advertisers may be limited to not being able to exceed a certain number of bid updates per bid topic (e.g., search term or group) per day. Advertisers may also be limited in a cumulative manner, such as to bid updates that cannot exceed a certain total number (or "aggregate") or frequency per day for a certain number of bid topics, or to bid updates that cannot exceed a certain average bid update number or frequency per day for a certain number of advertisements, and so forth. In some embodiments, the advertiser pays for the updates, or the available updates may be based on the advertiser's spending. Since updates can be a limited and valuable resource, it would be sensible for advertisers to distribute available bid updates differently for different search terms or for search terms originality.

For example, an advertiser may wish to use a higher bid update rate for more important or valuable search term originations or for search term originations in a less stable market, and compensate by using a lower bid update rate for less important or less valuable search terms or groups or for search terms or groups in a less stable market. In some embodiments, the bid optimizer 408 determines a bid update period based on such factors, for example. This may result in a more reasonable, optimized, or maximized approach than utilizing a uniform update rate for all lists without regard to value. A less frequently updated list may offset a more frequently updated list. For example, the list limits may be cumulative, such that if an advertiser uses an underbound amount for one or more lists, the advertiser may be allowed to use the excess amount for one or more other lists as long as the cumulative limit is not exceeded. Methods for calculating, determining, or evaluating value are described further below.

One technique that may be useful to advertisers or other bidders in an auction-based scenario such as that described above is referred to as bid hiding (or highest bid hiding). Bid hiding is a technique that can be used manually, for example, by the advertiser itself, where the advertiser can use the ad campaign facilitator 134 in this regard. However, in some embodiments, bid hiding may be automatically used by, for example, bid manager 510 or bid optimizer 508, or both.

Bid hiding may include policies used by bidders for listings in a listing auction. For example, assume that a bidder has offered or is ready to offer a certain highest bid, or the highest bid that the bidder is willing or is likely to submit. However, bidders may wish to avoid exposing this highest bid to other bidders during a listing auction. The winning bidder may be billed $0.01 more per click than the next lowest bidder in the auction, but this is not necessarily the amount the winning bidder actually bids. Exposing the highest bids of bidders can be detrimental to bidders, for example, by subjecting bidders to malicious bidding strategies. Such a malicious policy may include the second bidder placing a price just below the first bidder's highest bid to ensure that the amount actually charged by the first bidder (assuming the first bidder wins the listing) will be based on the first bidder's highest bid. Further, exposing the highest bid allows potential competitors to know that the bidder is willing to bid, which the bidder may not wish to be.

Bid hiding or maximum bid hiding is a technique in which the amount bid by a bidder is only equal to the amount that the bidder would expect to be billed if the bidder submitted the bidder's maximum bid, which, as described above, would be below the bidder's maximum bid. A system regulator (which may be a program or software module, for example, included in the ad campaigns facilitation server) may be used in conjunction with the auction, which limits the amount of updates per listing day for each advertiser, where the update period is the time between the highest bid hidden updates.

For example, assume that the marketplace operator has exposed auction status, including all of the highest bids and advertisements associated with each bid (even if the click is billed $0.01 higher than the next lower bidder). Bid hiding attempts hide an advertiser's highest bid by just bidding in an amount that is exactly equal to the amount that it would be expected to be billed if it submitted its highest bid to the auction. This not only protects the highest bid from the competitor's view, but also prevents some malicious bidding strategies, such as bid prices $0.01 below the competitor's bid, so they will pay their highest bid for each click.

In some embodiments, the bid optimizer 508 may include a program, software, or one or more applications configurable by an advertiser user that may be used to determine a desirable or optimal bid for an advertiser on a listing (e.g., paid search results). The configuration by the user may include, for example, the user setting goals and constraints. The constraints may include a highest bid and a lowest bid. The targets may be associated with the list and may be specified in terms of one or more metrics related to the performance of the list. The bid optimizer 408 can analyze recent analysis data in conjunction with metrics and specify bid recommendations that are predicted by the bid optimizer to achieve a goal or to be as close to a goal as possible. The bid optimizer 408 can provide recommendations for a list that can include a highest bid and an update period, which can be the time between highest bid hiding updates.

In some embodiments, a bid update rate adjuster (e.g., a program or software module that may be part of an advertising campaign assist program) is used to limit the replication costs to market operators, but also limits the ability of advertisers to control their placement in auctions that are of the greatest importance to their business. Some embodiments of the present invention thus provide a solution to this problem by aligning the cost structure of the market operator with the business goals of the advertiser.

One approach is for the marketplace operator to bill the advertiser for bid updates. This would cover the costs associated with replication for market operators and encourage advertisers to efficiently use bid updates. This may result in a reasonable decision by the advertiser as to the true value of each bid update. This approach may not be practical in some situations for several reasons, including considering that auction participants should not be billed for participation alone (which may be considered as opposed to a business model that pays for performance)

In some embodiments, the bid update frequency is adjusted for a listing based on the value provided to the advertiser by the listing; the greater the value, the more frequent the bid updates. It is generally the case that a small portion of the listings provide the majority of the value to any given advertiser, and therefore a reduction in the bid update frequency of many low-value listings is used to offset the significant increase in the bid update frequency of high-value listings. The benefits are enormous for advertisers, while the total number of bid updates (and thus costs) is kept constant or reduced.

A first embodiment of the present invention is deployed with specialized access to XML-based APIs to keep the bid update rate regulator from being enabled-the value-based bid update rate is controlled within a bid hiding engine that is part of the bid manager 410. An alternative embodiment is for the regulator to be modified to implement a value-based bid update rate.

In this context, there are many possible definitions for the "value" of a list, including the rate that advertisers spend on the list, and the revenue rate of the advertisers generated by leads from the list. In some embodiments, the "value" is calculated using the bid optimizer 508.

In some embodiments, the value of the list is determined based on the spending rate S of the list. Studies have shown that in some cases 90% of the advertisers spend are concentrated in the 1% list. This means that, for example, if all of these lists are bid updated at the maximum update rate and the bid update rate of the least expensive 99% list is to be reduced by half, the bid update rate of the most expensive 1% list can be increased to 100 times the previous bid update rate without increasing the total number of bid updates.

In a first embodiment, the following formula is used:

(1)R＝min(max(M×S，R_min)，R_max)

where R is a value-based bid update rate in minutes between bid updates of the listing.

S is the rate of spending the advertiser has recently on the list toIs a unit. If the list does not result in a click, and thus no cost, then it can be used

M is the cost in dollars required per bid update. M may be a constant value, e.g., M ═ 2.00, or it may be dynamically updated to reflect changes in bid update costs. R_minIs the minimum time allowed between bid updates in minutes. In the first embodiment, a constant R is used_minBut different constants may be used or they may be changed dynamically. R_maxIs the maximum time allowed between bid updates in minutes. In the first embodiment, a constant R is used_max1020, different constants may be used or may be dynamically changed.

To determine S, the "most recent" action associated with the list is reviewed. In this context, recently, data sets should be reviewed far enough in time to collect a large enough number to be relatively stable, but not so wide as to hide the changes in recently spent rates. The period of review may be defined as D (in minutes) and the cost C (in dollars) of the advertiser at time of period D. Thus, the

In some embodiments, it is desirable (but not necessary) to limit D so that an unlimited amount of data need not be considered. In the first embodiment, D_maxThe maximum value of (a) is 30 days. There are several strategies for determining the relevant data sets to consider. For example, one approach is to look back far enough to capture a certain amount of cost, such as C ≧ 10. The disadvantage of this strategy is that the cost per click is constant. Another approach is to review a fixed period, for example three days. The disadvantage of this strategy is that it is not sensitive to high frequency variations in the rate of the flowers. Another approach is to look back far enough to capture a certain number of clicks, for example at least 100. This is the policy used in the first embodiment. .

In some embodiments, the bid optimizer 408 is a predictive, budget-aware optimizer that optimizes the spending of a limited budget on a pay-for-placement network. The infrastructure to support prediction-based optimization is unusual. A bid optimization is provided in a shorter term and a retrospective control loop optimizer is used to recommend the highest bid.

In some embodiments, the user interface provides the advertiser with the option to perform recommended changes on behalf of the advertiser. The user may set up an account to accept recommendations automatically when they change, or manually.

In some embodiments, implementations support various matching schemes or selections, such as a matching scheme that requires entry of one or more exact search terms to cause a list to be presented, or a matching scheme that only requires one or more terms to appear somewhere in a search, and so forth.

In some embodiments, advertisers configure bid optimizer 408 by setting goals and constraints. For example, in some embodiments, the user specifies a target CPA (cost per acquisition). The user also specifies a maximum CPA that is used (in conjunction with the CPA target) to determine whether the offer was successful. Optionally, the user may also specify a maximum of two constraints: the highest bid, the lowest bid (some embodiments may include two additional constraints: highest position and lowest position). These goals and constraints may be specified at the following levels: global defaults (e.g., across the entire active set), active defaults, and originals. These levels form a hierarchy: if no value is specified at the level of originality, then use the value from the activity level; if no value is specified at the activity level, a global default value is used.

In some embodiments, the target is required, so only two states are available: either a value or "inheritance" (unavailable for global default inheritance). The constraint is optional and may have one of three states: a value, "inherit," or "none" (except "inherit" is not available at the global level). The goals (and analysis data) direct the bid optimizer 408 to select recommendations and are used to determine how to evaluate the success of the offer. In some embodiments, all optimizations and evaluations are done at the promised level.

The constraints (and the recommendations and current market state of the bid optimizer 408) guide the bid updates of the bid manager 410. At the importation, the highest bid constraint and the lowest bid constraint of the list with the current bid below $0.10 will be set to the current bid. All other lists will inherit the constraint value at the entrance.

In some embodiments, the bid optimizer 408 reviews over time (up to 30 days) to find analytical data related to impressions, leads, conversions, costs, revenues, and the like. First, it samples the analysis data over a period of time long enough to cover at least 10 transitions. If zero transitions are found, it does the same procedure, but the time of review covers at least 1000 boots. If zero leads are found, it attempts to cover at least 10,000 impressions. The time period covering the required number of events (transitions, leads or impressions) is called an aggregation period. Based on the analysis data, the bid optimizer makes recommendations for each listing and updates the recommendations.

In some embodiments, the recommendation for the list consists of the highest bid and an update period (the highest bid hides the time between updates-referencing the bid manager to see how that value is used). Each listing receives recommendations based on the analysis data for the listing and market dynamics.

In some embodiments, a bid recommendation for a list is checked/updated when at least one of the following conditions is met: (1) at least 20% of the aggregation period has elapsed since the last examination; (2) if zero transitions are found in the aggregation period, at least 20% of the time required to spend the target CPA since the last check has elapsed. In other words, if the target CPA is $10, the aggregation period is 100 hours, and the cost during the aggregation period is $100, then the time required to spend the target CPA is 10 hours-so the rule will trigger a check every 2 hours; (3) at least one day has passed since the last examination.

In some embodiments, the update period is determined according to the following formula (proportional to the spending rate), where the recommendation is updated with the matching first rule for each list. Hereinafter, "Impr" means "impression", "Conv" means "conversion", and "CPA" means "cost per acquisition".

Table 1:

is the list online?	Impr	Guiding	Conv	CPA	Cost in aggregation period	Movement of
is the list online?	Impr	Guiding	Conv	CPA	Cost in aggregation period	Movement of	Whether or not	0	*	*	*	*	Recommending current bids and disabling bid hiding
*	*	*	＞0	Target	*	Not changing recommendations	Whether or not	0	*	*	*	*	Recommending current bids and disabling bid hiding
*	*	*	＞0	Target	*	Not changing recommendations	*	*	*	≥10	< object	*	Raising bids to$0.10 higher than the next higher position in the exact match market
*	*	*	≥10	Target > target	*	Lowering bids to $0.01 lower than the next lower position in the exact match market	*	*	*	≥10	< object	*
*	*	*	≥10	Target > target	*		*	*	*	＜10	< object	*	Raise bid by $0.01
*	*	*	＜10	Target > target	*	Lower bid by $0.01	*	*	*	＜10	< object	*	Raise bid by $0.01
*	*	*	＜10	Target > target	*	Lower bid by $0.01	*	*	＞0	0	N/A	< target CPA	Raise bid by $0.01
*	*	＞0	0	N/A	CPA ≧ target	Lower bid by $0.01	*	*	＞0	0	N/A	< target CPA	Raise bid by $0.01
*	*	＞0	0	N/A	CPA ≧ target	Lower bid by $0.01	*	＞0	0	0	N/A	*	Raise bid by $0.01
Is that	0	0	0	N/A	*	Lower bid by $0.01	*	＞0	0	0	N/A	*	Raise bid by $0.01

In some embodiments, it is assumed that the conversion rate is the same for all bid positions for a given offer.

In some embodiments, the bid manager 410 always performs the highest bid hiding by attempting to bid $0.01 higher than the next lower bid.

In some embodiments, the recommendations for the list by the bid optimizer 408 consist of the highest bid and a bid update period. The bid manager 410 checks/updates the bids of the list at the end of each update period. In some cases, unscheduled checking/updating of bids for a list is implemented. These are the cases: (1) the constraint changes and the current bid violates the new constraint. The highest priority of these bid updates; (2) the recommended highest bid changes.

Whenever a bid of the list is checked/updated (scheduled or unscheduled), the next check for the bid is scheduled based on the recommended update period. The bid manager 410 checks market status, recommendations, and constraints each time it manages bids for a listing. It limits the recommended highest bid with constraints, including market state for location-based constraints, to generate the highest bid. If the constraints can be met, the market state is checked to see if there are any competing bids that are equal to the highest bid. If so, the current bid is the highest bid. If not, the market state is checked for the highest competing bid that is less than the highest bid. If such a bid is found, the current bid price is $0.01 higher than the bid. If no lower bid is found, the current bid is the lowest bid. If the previous current bid is equal to the new current bid, no update is required. In either case, the next update time is set to the current time plus the recommended update period.

In some embodiments, system regulators are used to limit bid update rates and market state check rates, which can reduce replication load.

It is noted that retrospective control loop optimization is susceptible to interaction between the rate of convergence of the bid optimizer 408 and the rate of change of the system being controlled. For example, assume that the conversion rate doubles from noon to midnight in a 24-hour period due to the change in the population surfing the Web during the day. If the control loop is able to measure over a short, recent period of time (say a few hours) and converge quickly, the daily period will be tracked fairly well. However, if there is a severe mismatch, the control loop will raise the bid when the conversion rate is decreasing and lower the bid when the conversion rate is increasing. If the control loop were to review several days to evaluate current performance, the daily period would not significantly affect recommendations and bids would remain relatively stable without tracking the daily period.

In some embodiments, different kinds of bid changes are controlled separately. For example, in some embodiments, an automatically recommended change of less than $0.05 is automatically made, but an explicit approval is obtained for any larger amount. In some embodiments, bid increases are automatic, but bid decreases are not.

Determining the statistical significance of the ratio metric involves several considerations. In general, it is desirable to measure enough of the resulting events to determine the ratio (error bars). For example, when 100 transition events are seen, it is well understood what the lead to transition rate is, even if the rate is very small. However, assume that after 100 leads are measured, a transition is seen. In this case, it is not possible to say with great confidence what the ratio is. However, a range may be set therefor; for example, it is believed that the ratio is much less than 75%. It is necessary to characterize how many resulting events need to be measured for confidence in the ratio estimate, and how confidence in the maximum ratio is as a function of the number of source events measured.

In some embodiments, the configurable parameters include the data retention period, the number of impressions/leads/transitions required for N-statistical significance, and the delay and recommendation step size between successive recommendation updates.

The delay should be expressed as a function of time to achieve the N impressions, leads or transitions. This allows high inventory offers to have a tighter control loop. There should be a maximum delay so that the offer to get no/lower traffic can also be updated with the recommendation. If the day split is complete, the delay should be expressed in such a way that: this way, the delay is also meaningful when data is collected only for a given fraction of every 24 hours or every 7 days.

The recommended step size may be adaptive and may be gap-aware. It may also be subpenny to slow the rate of change.

Fig. 6 is a block diagram of a system 600 according to one embodiment of the invention. As shown, the system 600 includes a search optimizer 602, a marketplace 604, and an advertiser website 606, the search optimizer 602 may be part of an advertising campaign assistance program, and the marketplace 604 may be provided by or assisted by a marketplace operator. Search optimizer 602 includes a bid manager 616 and a bid optimizer 618. Search optimizer 602 also includes databases including constraint database 608, recommendation database 610, target database 612, and analysis data database 614. The databases 608, 610, 612, 614 may be part of an ad campaign database. The data flow is shown to include objective information sent to the bid optimizer, recommendation information sent from the bid optimizer 618 to the recommendation database 610, and constraints and recommendation information sent to the bid manager. Other illustrated data flows include auction status information sent from the marketplace 604 to the bid manager 616 and the bid optimizer 618, bid update information sent from the bid manager to the marketplace 604, referral (guidance) and cost and impression data sent from the marketplace 604 to the advertiser's website 606, and click stream information sent from the advertiser's website 606 to the analytics data database 614. The illustrated information flow is not intended to be comprehensive or limiting.

As described above, in embodiments of an auction-based sponsored search results listing environment, the prominence or ranking of listings may be important to ad performance and thus relevant to ad campaign optimization. Ranking is important to advertisers because it determines the quality of placement of their listings on the page displayed to the user. Although the details may vary depending on the member (search engine), a typical layout is as follows. The highest ranked list appears at the top of the page, the next list appears at the right tail, and the additional list appears at the bottom of the page (typically not visible if not scrolled). Listings ranked below about the first five will appear on subsequent search results pages.

There is a strong correlation between the ranking and both the number of impressions and the click-through rate (per impression click), which provides the advertiser with the opportunity to pay more per click (get higher ranking) for more visitors to visit their website. The result is that the advertiser needs or should or has determined on behalf of the advertiser how much the advertiser should be willing to bid for each listing based on the advertiser's business goals and the quality of the traffic generated by the listings on its website.

In the embodiment shown and described with reference to FIG. 6, a conceptual distinction is maintained between bid management and bid optimization. In this embodiment, bid management includes an exact determination of what bid to submit to the auction at any given time, where the determination is based on the highest bid willing to submit and other bids exposed in the auction. One common bid management strategy is bid hiding, which includes bid in an amount exactly equal to what is willing to pay for each click, as described above. In this embodiment, bid optimization includes determining the maximum amount that is willing to be paid per click for a listing at any given time. It should be noted that the distinction between bid management and bid optimization is applicable only to certain embodiments, including the embodiment shown and described with reference to FIG. 6. Other embodiments do not necessarily include such a distinction.

The task of bid optimization can be daunting to advertisers. Advertisers must measure traffic quality for each listing by tracking the behavior of individual users on the website and associating the results with the listing that directs users to the site. Both user behavior and auction dynamics may change constantly, and advertisers may have thousands of listings to manage. The difficulties associated with optimizing paid search bids in conjunction with the importance of paid search channels to advertisers have led to the development and importance of Search Engine Management (SEM) providers. The SEM utilizes a combination of bid management experience and software tools to assist advertisers in performance measurement, bid management, and bid optimization.

One aspect of the optimization problem is simply due to the large number of lists, which can be solved using software automation. Another aspect of the problem is the distribution of traffic on the list. In a study sample of advertiser account campaigns for market operators, 90% of advertiser spending was found to be concentrated on a list of only 1% over a one month period. The fact that most of the traffic is skewed to a small list means that there is a small number of "high speed" lists. The high speed list generates enough conversions to enable a clear assessment of performance for the business objectives. However, the problem of too much data arises. The large accumulation of data from high speed words creates a huge "inertia" that reduces the impact of current bid changes on measured performance.

Most lists are "low speed". Here, the problem is that the search terms associated with these lists are extremely specific to and relevant to a few searches. And also tend to have less competition in low speed auctions and therefore tend to be lower cost per click. The specificity of low speed lists often results in higher conversion rates than more generalized high speed lists. While there is great value in the low speed words, there is not enough performance data to be able to unambiguously assess performance against business goals. This means that the optimization method for high speed words does not work for low speed words. In summary, advertisers have many listings to manage, and all listings often have either too much or insufficient performance data.

As shown in FIG. 6, search optimizer 602 includes user-interactive Web application(s) to help advertisers automate bid management and bid optimization. The web application allows advertisers to configure automated collection filtering and aggregation of analytics data, as well as view analytics data in a set of reports. In addition, the web application allows advertisers to specify optimized business performance goals and bid constraints. Optimization objective types include or are expressed or indicated as cost-per-acquisition (CPA), Return On Advertisement Spending (ROAS), and constraint-only (optimization not based on performance), among others. The constraint types for bid management may include lowest bid, highest bid, lowest position, highest position, and so on.

The optimization component in the system 600 shown in FIG. 6 is a bid optimizer 618. The bid optimizer 618 generates a recommendation consisting of the highest bid and a value-based bid hiding rate. Recommendations are based on cost and impression data from the marketplace 604 or marketplace operator, click stream data from the advertiser's web site 606, performance goals set by the advertiser, and the current status of the auction. In the illustrated embodiment, the recommendation consists of the highest bid and a bid hiding update frequency.

The bid management component of the system 600 shown in FIG. 6 is a bid manager 616. As shown, the bid manager 616 manages the actual bids in the auction to conform to recommendations in the context of constraints and changing states of the auction. The bid manager 616 updates bids for the listing based on the recommended bid hiding rate (if necessary). Whenever a list is considered, the recommended bid is limited by the current bid associated with the lowest position and highest position constraints. Bids are also limited by the minimum and maximum bid constraints. Finally, the bids are also constrained by any limitations imposed by the auction itself.

In some embodiments, the bid optimizer 618 generates recommendations consisting of the highest bid and a value-based bid hiding rate (or refresh rate). The bid hiding rate is proportional to the rate the advertiser spends on the list.

FIG. 7 is a flow diagram illustrating a method 700 according to one embodiment of the invention. In some embodiments, the bid optimizer is implemented as a control loop optimizer style, although other implementations are also contemplated. The illustrated method 700 is performed by a control loop style bid optimizer. The illustrated method 700 is a primary control loop performed by some embodiments of a bid optimizer. As shown, at step 702, the bid optimizer determines current recommendation values, including a recommended highest bid and a bid hiding rate (or refresh rate). At step 704, the bid optimizer waits a specified period of time to allow the currently recommended and utilized value to have sufficient effect. After waiting a specified period of time at step 706, the method 700 returns to step 704 where the bid optimizer determines a new current recommendation value, including a new recommended maximum bid and a bid hiding rate.

In some embodiments, one or more algorithms or programs are used to determine a recommended highest bid or a recommended bid hiding rate. One feature or strategy of such an algorithm according to some embodiments is to use a variable amount of recent analysis data to evaluate performance proportional to the "speed" of the list. The policy is to view just enough data, or wait long enough to view just enough data to achieve sufficient confidence (in a statistical sense), or to determine or judge a sufficient amount of confidence, for example, by a market operator, to evaluate the recent performance of the listing. For example, in some embodiments, if 10,000 transitions have been measured, it may not be necessary to consider all 10,000 to determine c.p.a; the last 10 conversions may be sufficient. The advantage of looking at just enough data is that it maximizes the effect of the current situation, thus allowing a better decision to be made.

Another feature or policy used by some embodiments of the bid optimizer is sensitivity to the type and quality of analysis data available for the listing. With this strategy, the more statistically significant the performance assessment is, the more aggressively the recommendation bid is changed. An advantage is that more aggressive methods may be employed when more reliable data is available, and more conservative methods may be employed when the data is less conclusive.

Another feature or strategy used by some embodiments of the bid optimizer is the following method for optimizing low speed listings. The sensitivity to the type and quality of the analysis data allows to distinguish between low speed lists and to apply different recommendation algorithms. In particular, words that have not recently been converted and therefore are not amenable to CPA or ROAS calculations are of interest. The strategy is to slowly raise the bid until the last cost on the list is more than a certain threshold, and then slowly lower the bid. For lists with CPA targets, the target is used as the spending threshold. For lists with ROAS targets, the active measured CPA containing the list is used. If this is not available, the measured CPA of the advertiser's web site is generally used. If this is not available, the nominal value of the threshold is used. Another option used in some embodiments is to allow the advertiser to configure the threshold as another control parameter. The strategy is to raise the bid in an attempt to get more traffic in anticipation of a transition; when the listing bid falls below the lowest bid, it typically takes a little more than the target CPA, so even if a transition is obtained at that point, a lower bid will still be recommended. In other words, the more it takes to exceed the CPA target without a transition, the more confident (statistically) that the CPA target cannot be achieved for the list.

Another feature or strategy used by some embodiments of the bid optimizer is to use a variable refresh rate proportional to "speed". From one perspective, it is desirable to maximize refresh rate, as it determines the convergence rate of the bid optimizer and the ability of the bid optimizer 508 to track high frequency changes in performance. However, if the refresh rate is too fast, the current setting may not have an opportunity to affect performance and thus the bid optimizer tends to override the optimal setting. As such, a high refresh rate may be advantageous in terms of bid optimization because it enables better accuracy or "granularity" in the analysis of rapidly changing performance and corresponding changes to settings. However, if the refresh rate is too high, sufficient time may not have elapsed to accurately evaluate the setting impact.

It is therefore desirable to utilize a refresh rate window that is balanced to be large enough to produce sufficient statistical significance in evaluating the impact of settings, but small enough to respond sufficiently agilely to changing performance. In some embodiments, the refresh interval is set to 20% of the interval over which performance analysis data is considered or the shorter of the day, which has been found to be a good overall balance in most lists and situations. However, in some embodiments, the window is computed in a more complex manner in order to be self-optimizing.

However, it has been observed that the rate of transition and rate of change of the rate for a particular search term or group of terms may change dynamically depending on the week or time of the associated search, for example (the rate of transition is specified in this example as transition divided by lead). For example, if a search occurs late at night or on a particular day or days of the week, the search engine user investigating new car prices is much less likely to purchase. This may result in a sharp change in conversion rate and conversion rate change frequency or speed depending on the week and time of day.

It has also been observed that the shopping cycle may vary dramatically for different products. The purchase period may represent the amount of time between a lead to initially visit the website and a lead to generate a conversion (e.g., by purchasing an advertised product). For example, a purchaser typically waits a long time, such as one or two weeks, before purchasing his or her car for investigation, as opposed to, for example, a purchaser of a book being likely to act immediately or within one or two days. Further, the peak amount of time between lead acquisition and purchase may vary for different products, services, content, etc. The purchase period may affect or break the association of the boot with the transition and thus may skew the transition rate if the refresh rate window is too small.

For the above reasons, in some embodiments, the refresh rate is optimized or balanced based at least in part on factors including observed changes in the conversion rate and rate of change of the conversion rate, a particular purchase period, or other factors. For example, a larger window may be utilized during days or times when the rate of change of conversion is expected to be low or the purchase period is long, while a shorter window may be utilized during days or times when the rate of change of conversion is expected to be high. In addition, expected change in conversion rate based on week or time of day (or other factors, such as holidays, seasons, current events, etc.) may be considered in determining the optimal settings.

FIG. 8 is a graph of conversion rate versus time for a hypothetical search term or phrase group, according to one embodiment of the present invention. Fig. 8 shows an example of how the conversion rate and the conversion rate change speed (or rate) change based on the week or time of day. As shown, the conversion rate peaks and remains relatively stable over a period of hours on friday centered at about 8pm, as shown by data point 802. By the time of data point 804 (approximately 12am), the slew rate drops off rapidly. By data point 806 (approximately Saturday 5am), the conversion rate is at the low point of Saturday and is again relatively stable. By data point 808 (approximately 8pm for sunday), the conversion rate reached the peak for sunday and the peak was higher than the peak for friday. In some embodiments, the bid optimizer 508 is programmed to analyze data including information about historical and expected transition rate over time (which may be frequently updated) and consider this data in determining settings including, for example, the highest bid and refresh rate.

FIG. 9 is a graph 900 of a hypothetical purchase cycle, plotted as a number of transitions versus elapsed time since boot acquisition for two different products, product A (cycle shown in solid lines) and product B (cycle shown in dashed lines). As shown, for product a, a very high initial peak occurs at data point 902 immediately after the boot acquisition. This is followed by a sharp drop, reaching a low point at data point 904 at approximately the end of day 1, a slow rise to a lower second peak at data point 906 at approximately day 4, and a very slow drop to zero or nearly zero at data point 908 at approximately day 9.

For product B, a lower initial peak occurs at data point 910 immediately after the pilot acquisition, followed by a slightly sharp drop, reaching a low point at data point 912 approximately at day 2. This is followed by a gradual rise, reaching a second peak at data point 914, approximately day 6, and finally slowly dropping to zero or nearly zero at data point 916, approximately day 13.

As shown in fig. 9, the purchase period may vary greatly between advertised products, services, content, etc., including peaking and dropping at different times, increasing or decreasing conversion rate changes at different times, and dropping to zero or nearly zero at different times. In some embodiments, this information, which may include statistics, curves, and models and frequent updates based on historical purchase cycle information for various types of products, may be provided to a bid optimizer 508, and the bid optimizer 508 determines settings based at least in part on this information. For example, a larger refresh window may be determined for a longer purchase period to ensure that the lead is accurately associated with the relevant transition.

FIG. 10 is a simplified screenshot 100 according to one embodiment of the present invention. In some embodiments, marketing console 1002 includes a user interactive interface provided by a Web application or set of Web applications, accessible via the internet, and available to advertisers (or other entities controlling advertising campaigns, or the administrator of marketing console 1002 itself). Marketing console 1002 may be used for a number of purposes to assist in the management and optimization of an advertising campaign. Marketing console 1002 is accessible via the internet, and access may be protected by many means known in the art, including password protected access.

In some embodiments, marketing console 1002 may be used by advertisers to assist in ad campaign management and optimization, which may include, for example, managing listings associated with an auction-based search term-related sponsored search results listing marketplace. For example, advertisers may use a marketing console to access ad campaign information and ad campaign performance information, search information, analysis information, procurement reports, summaries, and the like, stored in a relational ad campaign database. The advertiser may also utilize the marketing console 1002 to change listings or bidding strategies, which are updated in the advertising campaign database. Additionally, marketing console 1002 may be used to perform comparisons of the performance of components of an advertising campaign, such as the performance of particular listings, search term originals, channels, instruments, and so forth.

While marketing console 1002 is described in the context of an auction-based search term-related sponsored listing, it should be understood that in some embodiments, marketing console may also be used for offline or non-sponsored search advertising campaigns and advertising campaign performance, or online and offline advertising campaign information combinations.

Marketing console 1002 takes advantage of and facilitates a large number of advertising campaigns and advertising campaign performance information stored in an advertising campaign database (e.g., advertising campaign database 136 shown in fig. 1). As shown in fig. 10, one such tool is a search optimizer 1004. In general, the search optimizer 1004 may be used to access advertising campaigns and advertising campaign performance data, provide summaries, reports, and obtain exportable spreadsheet data or files to be used external to the marketing console 1002.

The user may interact with the search optimizer 1004 to specify parameters for customized collection, searching, presentation, analysis, and reporting of data. For example, a user may specify a particular aspect of an advertising campaign, or a particular term, or both, and request corresponding data or summaries. The user may specify, for example, a channel or means, a particular search term or origin and duration, and request summary information. Search optimizer 1004 may access and use information in the relational advertisement campaign database in response to user requests. The ad campaign database includes data collected from potentially many disparate sources, including information from many members as well as information from the advertiser's web site itself, which may be utilized by the search optimizer 1004. The search optimizer 1004 may also be used by advertisers to modify their ad campaign information in the ad campaign database.

As shown, the user may enter parameters of the request or search in a parameters area 1006 and obtain results in a results area 1008. In the illustrated example, the user has made a request and a result has been provided that indicates a user preference for the Yahoo! A collection of search terms or keywords in a search engine. A chart 1012 is provided that includes a list 1008 of keywords and a row 1010 that includes metric or analysis data associated with the keywords, which may be expressed in a variety of ways, including performance metrics (such as CPA, ROAS, etc.), percentages, and so forth. For example, the user may obtain results that allow for a comparison of performance between different members, different originals, and the like. Of course, a wide variety of information and ways of organizing information are possible and available to the user.

The use of marketing console 1002 thus provides advertisers with a convenient and simple way to access customized reports or analytics on campaign information, with the advantage that a large amount of data from a variety of disparate sources is available.

As shown, the user may select any one of a series of tool groupings 1014. As shown, a configuration management tool group is selected. It should be remembered that the screenshot 1000 has been simplified so as not to show details that may include sub-groups of tools and other features.

In some embodiments, a user may utilize a search optimizer to specify a user "watch lists". The watch list may include information about particular selected terms, such as the performance of the most important search terms of the advertiser being tracked, allowing easy and immediate access to critical data.

In some embodiments, the search optimizer may be used to select an "auto-accept mode" in which the user specifies that the recommendations of the bid optimizer are to be automatically implemented, or a mode in which the recommendations are presented to the user for acceptance before being implemented, or a manual mode that bypasses the bid optimizer. In some embodiments, an auto-accept mode may be used for some situations or for some words, while a different mode may be used for other situations or for other words.

The information accessed by the search optimizer 1004 can include indications of settings such as bid settings and refresh rates, and can provide indications of which settings were automatically implemented or last changed and which settings were manually implemented or last changed.

Marketing console 1002 may also provide access to billing and pricing information in conjunction with a marketplace operator.

In some embodiments, the marketing console may also be used by a manager or agent of the marketplace operator. Such users may use the marketing console for purposes such as tracking other users 'usage of the marketing console (and displaying reports, etc.), tracking usage of the market operator's server computer, troubleshooting software or hardware problems, and the like.

Claims

1. A method for facilitating management of an advertising campaign, the method comprising:

one or more ad campaigns facilitation servers of an ad campaigns facilitator obtain ad campaign information from one or more advertisers related to the ad campaigns;

the one or more ad campaigns facilitation servers obtaining ad campaign performance information related to the ad campaigns from the one or more advertisers and from each of a plurality of members of the ad campaigns facilitation bureaus;

the one or more ad campaigns facilitation servers store the ad campaign information and the ad campaign performance information in one or more ad campaigns databases; and is

The one or more ad campaign facilitation servers utilize at least a portion of the ad campaign information and at least a portion of the ad campaign performance information to facilitate managing ad campaigns.

2. The method of claim 1, wherein the one or more ad campaign assistance servers assist in managing ad campaigns comprises assisting in implementing a bidding strategy for advertisers in an auction-based search term-related sponsored listings marketplace.

3. The method of claim 2, comprising facilitating implementation of bids related to originality of the search term.

4. The method of claim 3, comprising the one or more servers using a highest bid hiding policy for implementing bids.

5. The method of claim 4, comprising the one or more ad campaigns facilitation servers automatically implementing a bidding strategy for advertisers.

6. The method of claim 5, wherein the one or more ad campaigns facilitation servers utilize information stored in the ad campaigns database to determine an optimal bidding strategy for advertisers.

7. The method of claim 6, wherein the one or more ad campaigns facilitation servers automatically implement the determined optimal bidding strategy.

8. The method of claim 1, comprising providing a user interactive interface with the one or more ad campaigns facilitation servers to allow the one or more advertisers to access and modify at least a portion of the information stored in the ad campaigns database to facilitate managing the ad campaigns.

9. The method of claim 8 wherein providing the user interactive interface comprises providing the advertiser with the ability to search for ad campaign information and ad campaign performance information, obtain an analysis of ad campaign information and ad campaign performance information, and obtain summary information about ad campaign information and ad campaign performance information.

10. The method of claim 9, wherein providing the user interactive interface comprises providing the advertiser with a monitored list of information related to the originality of the advertiser-selected search term.

11. The method of claim 1 wherein obtaining advertising campaign performance information from each of a plurality of members comprises obtaining advertising campaign performance information from each of a plurality of distinct members.

12. The method of claim 10 wherein obtaining advertising campaign performance information from each of a plurality of distinct members comprises obtaining advertising campaign performance information from at least one online member and at least one offline member.

13. The method of claim 1 wherein obtaining advertising campaign performance information from one or more advertisers comprises obtaining advertising campaign performance information from a plurality of disparate advertisers.

14. The method of claim 1, comprising storing the ad campaign information and the ad campaign performance information in the ad campaign database in an integrated manner.

15. The method of claim 1, wherein at least one of the members interacts with the one or more ad campaigns facilitation servers using one or more application program interfaces.

16. The method of claim 1, wherein the one or more ad campaigns facilitation servers comprise a plurality of geographically distributed ad campaigns facilitation servers, and comprising replicating changes in ad campaign information and ad campaign performance information across several of the ad campaigns facilitation servers to facilitate information synchronization.

17. The method of claim 1, wherein the one or more ad campaigns facilitation servers are of a marketplace operator, and the marketplace operator operates an auction-based search term-related sponsored listings marketplace for advertiser participation.

18. The method of claim 17, comprising utilizing a promise exchange engine to assist in providing the marketplace.

19. The method of claim 1, comprising the one or more ad campaigns facilitation servers facilitating instrumenting advertiser websites with HTML tags to facilitate automatic collection of ad campaign performance information to be collected by the one or more ad campaigns facilitation servers and stored in the one or more ad campaigns databases.

20. A system for facilitating management of an advertising campaign, the system comprising:

a computer network;

one or more ad campaigns facilitation servers connected to ad campaigns facilitators of the network;

one or more ad campaign databases connected to the one or more ad campaign facilitation servers;

a plurality of members of the advertising campaign facilitator connected to the network; and

a plurality of advertisers connected to the network;

wherein the one or more ad campaigns facilitation servers are adapted to obtain ad campaign information from the advertisers related to the ad campaigns;

wherein the one or more ad campaigns facilitation servers are adapted to obtain ad campaign performance information related to the ad campaigns from the advertisers and the members;

wherein the one or more ad campaigns facilitation servers are adapted to store the ad campaign information and the ad campaign performance information in one or more ad campaigns databases;

and wherein the one or more ad campaign facilitation servers are adapted to facilitate management of ad campaigns utilizing at least a portion of the ad campaign information and at least a portion of the ad campaign performance information.

21. The system of claim 20, wherein the one or more ad campaigns facilitation servers are adapted to provide a user interactive interface to allow the advertiser to access and modify at least a portion of the information stored in the ad campaigns database to facilitate managing the ad campaigns.

22. The system of claim 20, wherein the one or more servers use a bid manager program to automatically implement a bidding strategy for advertisers.

23. The system of claim 20, wherein the bid manager hides a highest bid for implementing a bidding strategy.

24. The system of claim 22, wherein the one or more servers utilize a bid optimizer program to determine an optimal bidding strategy utilizing information stored in the advertisement campaign database.

25. The system of claim 20, wherein the plurality of members comprises a plurality of distinct members.

26. The system of claim 25, wherein the plurality of distinct members includes at least one online member and at least one offline member.

27. The system of claim 22, wherein the plurality of advertisers includes a plurality of disparate advertisers.

28. The system of claim 20, wherein the one or more ad campaigns facilitation servers are adapted to store the ad campaign information and the ad campaign performance information in the ad campaigns database in an integrated manner.

29. The system of claim 20, wherein at least one of the members interacts with the one or more ad campaigns facilitation servers using one or more application program interfaces.

30. The system of claim 20, wherein the one or more ad campaigns facilitation servers comprise a plurality of geographically distributed ad campaigns facilitation servers, and comprising replicating changes in ad campaign information and ad campaign performance information across several of the ad campaigns facilitation servers to facilitate information synchronization.

31. The system of claim 20, wherein the one or more ad campaigns facilitation servers are marketplace operators and the marketplace operators operate an auction-based search term-related sponsored listings marketplace for advertiser participation.

32. The system of claim 31, wherein the one or more ad campaigns facilitation servers utilize a promise exchange engine to facilitate providing the marketplace.

33. The system of claim 20, wherein the network comprises the internet.

34. A computer usable medium storing program code that, when executed on a computerized device, causes the computerized device to perform a method for assisting in managing an advertising campaign, the method comprising: