System and method for estimating potential unique online users an advertisement can reach

• US 9,299,085 B2
• Filed: 09/27/2011
• Issued: 03/29/2016
• Est. Priority Date: 09/27/2011
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Lei, Ming, Jiang, Renjie, Alton, Kenneth R., Chirravuri, Varun, Quong, Russell W., Sharma, Abhinay, Zamir, Oren E., Yu, Xing, Feng, Zhengzhu
• Primary Examiner(s): Hammond, III, Thomas M
• Assistant Examiner(s): Kolosowski-Gager, Katherine
• Application Number:
  US13/246,609
• Publication Number:
  US 20150193811A1
• Time in Patent Office: 1,645 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06Q 30/0242 Determining effectiveness of advertisements
• CPC Class Codes: G06Q 30/0249 based upon budgets or funds
• CPC Class Codes: G06Q 30/0275 Auctions

Abstract

Methods, systems, and computer programs encoded on a computer storage medium include receiving, from an advertiser, advertisement criteria associated with an advertisement, the advertisement criteria comprising a first set of criteria and a budget and/or a bid, the advertisement criteria to be used in advertisement auctions for which the advertisement is to be considered for display to users performing online actions; determining a number of users for whom the advertisement was a candidate to be shown based on the first set of criteria associated with the advertisement, but to whom the advertisement was not shown based on the budget and/or bid of the advertisement during a particular period of time; and providing, in a report, information relating to the number of users.

Claims

1. A computer-implemented method comprising:

receiving, by a computer system, from an advertiser, a keyword, a web page, and advertisement criteria to be used in advertisement auctions considering one or more advertisements for display to users performing online actions, the advertisement criteria comprising a bid and a plurality of performance metrics selected from the group of;

(a) a budget associated with an ad group, (b) a quality criterion associated with at least one of the users, or (c) a frequency cap associated with the users;

determining, with a computer processor of the computer system, a first number of unique users for whom the one or more advertisements were candidates to be displayed based on the advertisement criteria, but to whom the one or more advertisements were not displayed, determining the first number of unique users comprising;

assigning a unique user id for each user performing online actions;

executing a cookie correction algorithm, wherein the cookie correction algorithm identifies non-unique users from unique users from the total number of users performing online actions;

collecting, for a predetermined period of time, auction data from a plurality of cookies associated with the users, wherein the auction data is associated with the received advertisement criteria;

determining, for the predetermined period of time an impression lost value for each user indicating whether the advertiser lost an advertisement auction, associated with the auction data, due to the advertisement criteria;

storing the auction data in a plurality of log files, each log file associated with a single user id and further comprising ad group id, criteria id, and the determined impression lost value associated with the user;

for the predetermined period of time, aggregating the plurality of log files using a two-stage process, wherein;

the first stage comprises transforming the plurality of log files into a plurality of tuples using a combination of the ad group id, criteria id, and user id of each log file as a primary key, each tuple comprising the impression lost value; and

the second stage comprises transforming the plurality of tuples to determine the first number of unique users for whom the one or more advertisements were candidates but to whom the one or more advertisements were not displayed, the first number of unique users identified using tuples having an impression lost value during the predetermined period of time;

for the predetermined period of time, determining the first number of unique users based on the aggregation, wherein the determining further comprises identifying a set of tuples of the plurality of tuples having an impression lost value indicating that the user associated with the tuple was not shown the one or more advertisements based on the advertisement criteria; and

creating, with the computer processor, a report for the predetermined period of time displaying information relating to the first number of unique users, wherein creating the report comprises;

determining and generating data reflecting a number of potential users, the number of potential users representing a sum of users that were shown the one or more advertisements and those that were not shown the one or more advertisements; and

determining and displaying, for each of the plurality of performance metrics, the first number of unique users to whom the one or more advertisements were not displayed based on the performance metric.

2. The method according to claim 1:

wherein each log file further comprises an impression value indicating whether the advertisement identified by the ad group id was served;

wherein each tuple further comprises the impression value;

wherein the second stage further comprises determining a second number of users identified by tuples having an impression value during the predetermined period of time; and

wherein providing information further comprises providing information relating to the second number of users.

3. The method according to claim 1:

wherein each log file further comprises a potential impression value indicating whether a user associated with the user id caused an ad server to request the advertisement from an ad index database that stores candidate advertisements;

wherein each tuple further comprises the potential impression value;

wherein the second stage further comprises determining a third number of users identified by tuples having a potential impression value during the predetermined period of time; and

wherein providing information further comprises providing information relating to the third number of users.

4. The method according to claim 1, wherein the quality criterion comprises predicted Click Thru Rate.

5. The method according to claim 1, wherein the quality criterion comprises relevancy.

6. The method according to claim 1, wherein storing the auction data is performed as a batch process once a day.

7. The method according to claim 1, wherein the user id is a pseudonymous ID of a mobile device.

8. A computer-implemented system comprising:

a receiving unit configured to;

receive, from an advertiser, a keyword, a web page, and advertisement criteria to be used in advertisement auctions considering advertisements for display to users performing online actions, the advertisement criteria comprising a bid and a plurality of performance metrics selected from the group of;

(a) a budget associated with an ad group, (b) a quality criterion associated with at least one of the users, or (c) a frequency cap associated with the users;

a determining unit configured to;

determine a first number of unique users for whom the one or more advertisements were candidate to be displayed based on the advertisement criteria, but to whom the one or more advertisements were not displayed, by;

assigning a unique user id for each user performing online actions; and

executing a cookie correction algorithm, wherein the cookie correction algorithm identifies non-unique users from unique users from the total number of users performing online actions;

a database unit configured to;

collect, for a predetermined period of time, auction data from a plurality of cookies associated with the users, wherein the auction data is associated with the received advertisement criteria; and

store the auction data in a plurality of log files, each log file associated with a single user id and further comprising ad group id, criteria id, and the impression lost value associated with the user;

an aggregating unit configured to;

aggregate, for the predetermined period of time, the plurality of log files using a two-stage process, wherein;

the first stage comprises transforming the plurality of log files into a plurality of tuples using a combination of the ad group id, criteria id, and user id of each log file as a primary key to transform the plurality of log files into a plurality of tuples, each tuple comprising the impression lost value; and

the second stage comprises transforming the plurality of tuples to determine a first number of unique users identified by tuples having an impression lost value during the predetermined period of time; and

a report generating unit configured to;

determine, for the predetermined period of time, the first number of unique users based on the aggregation by identifying a set of tuples of the plurality of tuples having an impression lost value indicating that the user associated with the tuple was not shown the one or more advertisements based on the advertisement criteria; and

generate a report for the predetermined period of time displaying information relating to the first number of unique users by;

determining and generating data reflecting a number of potential users, the number of potential users representing a sum of users that were shown the one or more advertisements and those that were not shown the one or more advertisements; and

determining and displaying, for each of the plurality of performance metrics, the first number of unique users to whom the one or more advertisements were not displayed based on the performance metric.

9. The system according to claim 8:

wherein, in the database unit, each log file further comprises an impression value indicating whether the advertisement identified by the ad group id was served;

wherein, in the aggregating unit, each tuple further comprises the impression value;

wherein, in the aggregating unit, the second stage further comprises determining a second number of users identified by tuples having an impression value during the predetermined period of time; and

wherein the report generating and displaying unit further comprises providing information relating to the second number of users.

10. The system according to claim 8:

wherein, in the database unit, each log file further comprises a potential impression value indicating whether a user associated with the user id caused an ad server to request the advertisement from an ad index database that stores candidate advertisements;

wherein, in the aggregating unit, each tuple further comprises the potential impression value;

wherein, in the aggregating unit, the second stage further comprises determining a third number of users identified by tuples having a potential impression value during the predetermined period of time; and

wherein the report generating and displaying unit further comprises providing information relating to the third number of users.

11. The system according to claim 8, wherein the quality criterion comprises predicted Click Thru Rate.

12. The system according to claim 8, wherein the quality criterion comprises relevancy.

13. The system according to claim 8, wherein, in the database unit, storing the auction data is performed as a batch process once a day.

14. The system according to claim 8, wherein, in the database unit, the user id is a pseudonymous ID of a mobile device.