Video play through rates

• US 8,209,715 B2
• Filed: 11/14/2008
• Issued: 06/26/2012
• Est. Priority Date: 11/14/2008
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Liao, Lin, Gupta, Manish, Moonka, Rajas
• Primary Examiner(s): Bui, Kieu Oanh T
• Application Number:
  US12/271,395
• Publication Number:
  US 20100125871A1
• Time in Patent Office: 1,320 Days
• Field of Search: 725/10, 725/32, 725/34, 725/35, 725/46, 725/50, 725/116, 725/134
• US Class Current: 725/34
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A request for video advertisements is received and video advertisements that can be provided in response to the request are identified. Performance features associated with the video advertisements are identified and are provided as input to a video advertisement performance model trained to estimate a play through rate for each of the video advertisements. The video advertisements are ranked based on the estimated play through rates for the video advertisements provided by the video advertisement performance model, and one or more video advertisements are provided in response to the request according to the ranking of the video advertisements.

Claims

1. A computer-implemented method, comprising:

receiving, by one or more computers, a request for video advertisements;

identifying, by the one or more computers, video advertisements that are eligible to be provided in response to the request;

obtaining, by the one or more computers, an estimated play through rate for each of the video advertisements, the estimated play through rate for each of the video advertisements being a value indicative of a likelihood that at least a predetermined amount of a duration of the video advertisement will be presented prior to occurrence of a user action that halts presentation of the video advertisement;

ranking, by the one or more computers, the video advertisements based on the estimated play through rates for the video advertisements; and

providing, by the one or more computers, one or more of the video advertisements in response to the request and according to the ranking.

2. The method of claim 1, further comprising:

obtaining data indicative of user interactions that cause presentation of the video advertisements;

for each user interaction that causes presentation of a video advertisement;

determining a presentation period for the video advertisement, the presentation period being a period between occurrence of the user interaction and occurrence of the user action that halts presentation of the video advertisement;

determining the duration of the video advertisement that was presented, the duration being determined based on a ratio of the presentation period relative to a total duration of the video advertisement;

determining whether the presentation of the video advertisement is a play through of the advertisement, wherein the presentation is determined to be a play through of the video advertisement when the duration of the video advertisement that was presented exceeds the predetermined duration; and

updating performance features of the video advertisements, the performance features for each of the video advertisements being updated based on a portion of the user interactions with the video advertisement that resulted in a presentation that was determined to be a play through of the video advertisement.

3. The method of claim 2, further comprising:

training a performance model using the performance features for the video advertisements, the performance model being trained to determine, based on features data for a video advertisement, the estimated play through rate of the video advertisement, the features data including data that are indicative of one or more visual features of the video advertisement.

4. The method of claim 3, wherein the performance model comprises a logistic regression model that includes coefficients for the visual features, the coefficient for each visual feature representing a relative weight of the visual feature for determining the estimated play through rate.

5. The method of claim 4, wherein the features data further include data that are indicative of one or more of a web property identifier;

an advertiser identifier;

a measured click through rate;

geo location data;

content of a web property associated with previous requests; and

a measured play through rate.

6. The method of claim 1, wherein ranking the video advertisements based on the estimated play through rates for the video advertisements provided by the video advertisement performance model comprises:

selecting a quality adjustment for each of the video advertisements, the quality adjustment for each of the video advertisements being selected based on the estimated play through rate for the video advertisement; and

ranking the video advertisements based, at least in part, on the selected quality adjustments.

7. The method of claim 6, wherein selecting a quality adjustment for each of the video advertisements comprises:

associating each of a plurality of quality adjustments with a corresponding range of estimated play through rates;

for each of the video advertisements;

determining the corresponding range of estimated play through rates that is inclusive of the estimated play through rate for the video advertisement; and

selecting, as the quality adjustment for the video advertisement, the quality adjustment associated with the corresponding range of estimated play through rates that is inclusive of the estimated play through rate for the video advertisement.

8. The method of claim 7, wherein ranking the video advertisements based on the estimated play through rates further comprises:

for each of the video advertisements, adjusting a bid associated with the video advertisement based on the quality adjustment for the video advertisement.

9. The method of claim 1, wherein the predetermined amount is 50%.

10. A system comprising:

one or more computers; and

a computer-readable medium coupled to the one or more computers having instructions stored thereon which, when executed by the one or more computers, cause the one or more computers to perform operations comprising;

receiving a request for video advertisements;

identifying video advertisements that are eligible to be provided in response to the request;

obtaining an estimated play through rate for each of the video advertisements, the estimated play through rate for each of the video advertisements being a value indicative of a likelihood that at least a predetermined amount of a duration of the video advertisement will be presented prior to occurrence of a user action that halts presentation of the video advertisement;

ranking the video advertisements based on the estimated play through rates for the video advertisements; and

providing one or more of the video advertisements in response to the request and according to the ranking.

11. The system of claim 10, wherein the operations further comprise:

obtaining data indicative of user interactions that cause presentation of the video advertisements;

for each user interaction that causes presentation of a video advertisement;

determining a presentation period for the video advertisement, the presentation period being a period between occurrence of the user interaction and occurrence of the user action that halts presentation of the video advertisement;

determining the duration of the video advertisement that was presented, the duration being determined based on a ratio of the presentation period relative to a total duration of the video advertisement;

determining whether the presentation of the video advertisement is a play through of the advertisement, wherein the presentation is determined to be a play through of the video advertisement when the duration of the video advertisement that was presented exceeds the predetermined duration; and

updating performance features of the video advertisements, the performance features for each of the video advertisements being updated based on a portion of the user interactions with the video advertisement that resulted in a presentation that was determined to be a play through of the video advertisement.

12. The system of claim 11, wherein the operations further comprise:

training a performance model using the performance features for the video advertisements, the performance model being trained to determine, based on features data for a video advertisement, the estimated play through rate of the video advertisement, the features data including data that are indicative of one or more visual features of the video advertisement.

13. The system of claim 12, wherein the performance model comprises a logistic regression model that includes coefficients for the visual features, the coefficient for each visual feature representing a relative weight of the visual feature for determining the estimated play through rate.

14. The system of claim 13, wherein the features data further include data that are indicative of one or more of a web property identifier;

an advertiser identifier;

a measured click through rate;

geo location data;

content of a web property associated with previous requests; and

a measured play through rate.

15. The system of claim 10, wherein ranking the video advertisements based on the estimated play through rates for the video advertisements provided by the video advertisement performance model comprises:

selecting a quality adjustment for each of the video advertisements the quality adjustment for each of the video advertisements being selected based on the estimated play through rate for the video advertisement; and

ranking the video advertisements based, at least in part, on the selected quality adjustments.

16. The system of claim 15, wherein selecting a quality adjustment for each of the video advertisements comprises:

associating each of a plurality of quality adjustments with a corresponding range of estimated play through rates;

for each of the video advertisements;

determining the corresponding range of estimated play through rates that is inclusive of the estimated play through rate for the video advertisement; and

selecting, as the quality adjustment for the video advertisement, the quality adjustment associated with the corresponding range of estimated play through rates that is inclusive of the estimated play through rate for the video advertisement.

17. The system of claim 16, wherein ranking the video advertisements based on the estimated play through rates further comprises:

for each of the video advertisements, adjusting a bid associated with the video advertisement based on the quality adjustment for the video advertisement.

18. A system, comprising:

one or more computers that include a prediction engine and a scoring engine;

the prediction engine being operable to;

receive a score request;

determine, based on features data for a video advertisement, an estimated play through rate for the video advertisement, the estimated play through rate being a value indicative of a likelihood that at least a predetermined amount of a duration of the video advertisement will be presented prior to occurrence of a user action that halts presentation of the video advertisement, the features data including data that are indicative of one or more visual features of the video advertisement; and

provide the estimated play through rate to the scoring engine;

the scoring engine being operable to;

identify video advertisements that are eligible to be provided in response to a request for video advertisements;

rank the video advertisements based on the estimated play through rates for the video advertisements; and

provide one or more of the video advertisements in response to the request and according to the ranking.

19. The system of claim 18, wherein the prediction engine comprises a logistic regression model that includes coefficients for the visual features, the coefficient for each visual feature representing a relative weight of the visual feature for determining the estimated play through rate.

20. The system of claim 19, wherein the features data further include data that are indicative of one or more of a web property identifier;

an advertiser identifier;

a measured click through rate;

geo location data;

content of a web property associated with previous requests; and

a measured play through rate.