PREDICTING USER RESPONSE TO ADVERTISEMENTS

• US 20100088152A1
• Filed: 03/24/2009
• Published: 04/08/2010
• Est. Priority Date: 10/02/2008
• Status: Abandoned Application

• Current Assignee: Turn (Nexxen International Ltd.)
• Original Assignee: Turn (Nexxen International Ltd.)
• Inventors: Bennett, Dominic
• Application Number:
  US12/410,400
• Publication Number:
  US 20100088152A1
• Time in Patent Office: Days
• US Class Current: 705/10
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0217 involving input on products or services in exchange for incentives or rewards
• CPC Class Codes: G06Q 30/0245 Surveys
• CPC Class Codes: G06Q 30/0255 based on user history
• CPC Class Codes: G06Q 30/0601 Electronic shopping [e-shopping]
• CPC Class Codes: G06Q 40/12 Accounting

Abstract

A system for predicting user responses to advertisements comprises a data collection component, a segmentation component, a modeling component, a rule building component, and an ad scoring component. The data collection component receives data from cookies stored on each client and from other sources. The segmentation component organizes the data according to segments. The modeling component groups users according to segments and compares a user'"'"'s actions to the models to predicts the user'"'"'s future responses. The rule building component generates an ad campaign comprised of rules. The model or the rules are compared to a plurality of rules to generate a score. The ad with the highest combination of a score and a bid is displayed on the client.

Claims

1. A computer-implemented method for scoring advertisements performed on at least one client, said client comprising a computer-readable medium coupled to a processor, the method comprising the steps of:

a data collection component storing data received from a plurality of clients, each client associated with a unique identifier, said data comprising a uniform resource locator (URL) for each website visited by said plurality of clients, a category for each item searched, viewed, or purchased on said website, and an action performed each time said client visits said website;

a segmentation component transforming said data into segments for each category and each action;

a modeling component generating models that group a plurality of said clients according to similarities in categories and actions performed by said clients;

a server receiving an ad call from a browser, said ad call associated with a client;

said modeling component comparing said client to said models to predict said client'"'"'s similarity to said categories by comparing data associated with said client to said models;

an ad scoring component for generating an ad score for each of a plurality of advertisements based on said client'"'"'s similarity to said model and a price that an advertiser pays for display of each advertisement;

selecting an advertisement from said plurality of advertisements with a highest ad score; and

said server transmitting to said browser said selected ad.

2. The method of claim 1, further comprising the step of:

receiving with said data collection component a notice of a click-through from said client in response to displaying said ad that is most likely to result in said click-through.

3. The method of claim 2, further comprising the step of:

receiving with said data collection component a notice of a conversion by said client in response to displaying said ad that is most likely to result in said click-through.

4. The method of claim 2, further comprising the step of:

updating said segments in response to receiving said notice of said click-through from said client in response to displaying said ad that is most likely to result in said click-through.

5. The method of claim 1, further comprising the step of:

wherein said step of retrieving a list of segments associated with said client is performed in real time.

6. The method of claim 1, further comprising the step of:

receiving said client'"'"'s profile from a cookie stored on said client.

7. The method of claim 1, further comprising the step of:

multiplying said ad score by a price per impression.

8. The method of claim 1, wherein said data further comprises a profile for a user associated with said client, said profile comprising at least one of the following demographics:

age, location, income, educational status, job category, and gender.

9. A system for scoring advertisements comprising:

a processor;

a storage device in communication with said processor and storing instructions adapted to be executed by said processor;

a data collection component that stores data received from a plurality of clients, each client associated with a unique identifier, said data comprising a uniform resource locator (URL) for each website visited by said plurality of clients, a category for each item searched, viewed, or purchased on said website, and an action performed each time said client visits said website;

a segmentation component that transforms said data into segments for each category and each action;

a modeling component that generates models that group a plurality of said clients according to similarities in categories and actions performed by said clients, said modeling component comparing a client to said models to predict said client'"'"'s similarity to said categories by comparing data associated with said client to said models; and

an ad scoring component that generates an ad score for each of a plurality of advertisements based on said client'"'"'s similarity to said model and a price that an advertiser pays for display of each advertisement, said ad scoring component selecting an advertisement from said plurality of advertisements with a highest ad score and transmitting said ad to a publisher in response to an ad call.

10. The method of claim 9, further comprising the step of:

multiplying said ad score by a price per impression.

11. The method of claim 9, further comprising the step of:

charging an advertiser that provided said ad that is most likely to result in said click-through for transmitting said ad.

12. The method of claim 9, further comprising the step of:

embedding a beacon in said browser, said beacon comprising at least one rule defined by an advertiser; and

notifying said advertiser when said client performs actions that are covered by said at least one rule.

13. The method of claim 9, further comprising the step of:

receiving with said data collection component a notice of said click-through from said client in response to displaying said ad that is most likely to result in said click-through.

14. The method of claim 13, further comprising the step of:

updating said segments in response to receiving said notice of a click-through from said client in response to displaying said ad that is most likely to result in said click-through.

15. A system for predicting user behavior in response to an advertisement comprising:

a data collection component stored on at least one computer, said computer comprising a computer-readable medium coupled to a processor, said data collection component for receiving data from a plurality of clients, each client associated with a unique identifier, said data comprising a uniform resource locator (URL) for each website visited by said plurality of clients, a category for each item searched, viewed, or purchased on said website, and an action performed each time said client visits said website;

a segmentation component coupled to said data collection component, said segmentation component adapted to receive said data from said data collection component and transforming said data into segments and grouping said clients according to said segments;

a modeling component coupled to said data collection component and said segmentation component, said modeling component receiving said segments from said segmentation component and generating models to predict a user'"'"'s reaction to display of an advertisement on said client;

a rule building component coupled to said segmentation component, said rule building component for generating a series of rules for displaying advertisements on a website, said rules organized by a category, an event, an event type, a recency, and a frequency; and

an ad scoring component coupled to said modeling component, said segmentation component, and said rule building component, said ad scoring component receiving a prediction from said modeling component and scoring a plurality of ads by comparing said ads that satisfy said rules generated by said rule building component to determine an ad that is most likely to result in a click-through.

16. The system of claim 15, wherein said system comprises a distributed datastore environment.

17. The system of claim 15, wherein said ad scoring component further comprises a database for storing said plurality of ads.

18. The system of claim 15, wherein said data received by said data collection component is derived from a cookie stored on said client.

19. The system of claim 15, wherein said modeling component predicts actions for said client by comparing said client to actions of other clients grouped according to said segments.

20. The system of claim 15, wherein said data is received from a third-party server that collects data from a plurality of cookies stores on said clients.