Determining bids for television advertisement auctions

Patent metadata

Field	Value
Publication number	US-9836769-B1
Application number	US-201113159718-A
Country	US
Kind code	B1
Filing date	Jun 14, 2011
Priority date	Jun 14, 2011
Publication date	Dec 5, 2017
Grant date	Dec 5, 2017

Abstract

Methods, systems and apparatus, including computer programs encoded on a computer storage medium, for generating a target bid for an advertiser to use in an auction directed at enabling the advertiser to reach the advertiser's advertising goal (e.g., a cost-per-call goal) for a particular advertising target (e.g., a television program or demographic) based on the target's historical advertising performance.

First claim

What is claimed is:

1. A computer-implemented method performed on a processor, comprising:
receiving, by the processor, from an advertising entity, a desired value specifying an advertising goal of the advertising entity for an advertising target, wherein advertisements of the advertising entity are presented to the advertising targets; and
for each of a plurality of advertising targets:
receiving attribution data for the advertising target, the attribution data describing a measure of effectiveness of the advertising target in achieving the advertising goal during a first time period;
generating an estimated measure of effectiveness of the advertising target during a second time period based on the attribution data, wherein the second time period is after the first time period; and
automatically determining, by the processor, a target bid for the advertising target during the second time period based on the desired value and the estimated measure of effectiveness, the automatically determined target bid representing a predicted cost to achieve the advertising goal of the advertising entity for the advertisement target;
preventing, by the processor, anomalies in the attribution data from causing greater than a threshold change from a current bid for the advertising target to the automatically determined target bid for the advertising target by:
determining that the automatically determined target bid is outside of a threshold range of bids, wherein the threshold range of bids is based on the current bid and includes the current bid; and
adjusting the automatically determined target bid to match a boundary of the threshold range of bids in response to determining that the automatically determined target bid is outside of the threshold range of bids, the adjusting comprising:
for one or more increments:
 determining an incremental bid that is a value between the current bid and the boundary of the threshold range of bids;
 using the incremental bid to distribute advertisements of the advertising entity to the advertising target; and
 evaluating performance of the incremental bid in reaching the advertising goal of the advertising entity for the advertising target;
based on the performance of at least one incremental bid being different than the advertising goal of the advertising entity for the advertising target, adjusting the target bid to match the boundary of the threshold range of bids; and

using the adjusted target bid to distribute the advertisements of the advertising entity to the advertising target.

2. The computer-implemented method of claim 1 , wherein the advertising goal is a goal to achieve a specified cost-per-call resulting from a presentation of an advertisement to the advertising target.

3. The computer-implemented method of claim 2 , wherein the estimated measure of effectiveness of the advertising target is a number of calls-per-one-thousand advertisement impressions to the advertising target.

4. The computer-implemented method of claim 1 , wherein the advertising goal is a goal to achieve a specified cost-per-conversion resulting from a presentation of an advertisement to the advertising target.

5. The computer-implemented method of claim 1 , wherein the advertising goal is a goal to achieve a specified search query uplift resulting from a presentation of an advertisement to the advertising target.

6. The computer-implemented method of claim 1 , wherein generating an estimated measure of effectiveness of the advertising target comprises:
partitioning the attribution data into two or more attribution data subsets, wherein each of the attribution data subsets uniquely corresponds to one non-overlapping time segment of a plurality of non-overlapping time segments in the first time period;
generating a weighted attribution data subset for each attribution data subset of the two or more attribution data subsets based on an application of a variable weighting factor to the attribution data subset, wherein the variable weighting factor applied to an attribution data subset varies according to a recency of the non-overlapping time segment corresponding to that attribution data subset; and
aggregating the weighted attribution data subsets.

7. The computer-implemented method of claim 6 , wherein a highest weight of the variable weighting factor is applied to an attribution data subset corresponding to a non-overlapping time segment most recent in time and a lowest weight of the variable weighting factor is applied to an attribution data subset corresponding to a non-overlapping time segment most distant in time.

8. The computer-implemented method of claim 6 , wherein the plurality of non-overlapping time segments in the first time period are a first non-overlapping time segment and a second non-overlapping time segment, the first non-overlapping time segment being a most-recent two week period in the first time period and the second non-overlapping time segment being all other time in the first time period.

9. The computer-implemented method of claim 6 , wherein determining a target bid for the advertising target during the second time period comprises multiplying the desired value by the estimated measure of effectiveness.

10. The computer-implemented method of claim 1 , wherein the attribution data is based on past performance of advertisements from the advertising entity.

11. The computer-implemented method of claim 1 , wherein the attribution data is based on past performance of advertisements from the advertising entity and other advertising entities.

12. The computer-implemented method of claim 1 , further comprising:
determining whether the target bid is within a tolerance range, wherein the tolerance range is a range of bids including the current bid and the tolerance range is a subset of the threshold range of bids; and
setting the target bid at the current bid whenever the target bid is within the tolerance range.

13. The computer-implemented method of claim 1 , wherein the advertising goal is a goal to achieve a specified increase in web traffic resulting from a presentation of an advertisement to the advertising target.

14. A system, comprising:
a data processing apparatus; and
software stored on a computer storage apparatus and comprising instructions executable by the data processing apparatus and upon such execution cause the data processing apparatus to perform operations comprising:
receiving, from an advertising entity, a desired value specifying an advertising goal of the advertising entity for an advertising target, the advertising target pertaining to a category of television advertisement, wherein advertisements of the advertising entity are presented to the advertising targets; and
for each of a plurality of advertising targets:
receiving attribution data for the advertising target, the attribution data describing a measure of effectiveness of the advertising target in achieving the advertising goal during a first time period;
generating an estimated measure of effectiveness of the advertising target during a second time period based on the attribution data, wherein the second time period is after the first time period; and
automatically determining a target bid for the advertising target during the second time period based on the desired value and the estimated measure of effectiveness, the target bid representing a predicted cost to achieve the advertising goal of the advertising entity for the advertisement target;
preventing anomalies in the attribution data from causing greater than a threshold change from a current bid for the advertising target to the automatically determ

Assignees

• Murray Robert
• Gangwar Jyotsna
• Bidyuk Bozhena
• Google Inc

Inventors

• Murray Robert
• Gangwar Jyotsna
• Bidyuk Bozhena

Classifications

• G06Q30/0275 Primary

  Auctions · CPC title