Method and system for media initialization via data sharing

• US 20090265218A1
• Filed: 12/11/2008
• Published: 10/22/2009
• Est. Priority Date: 04/17/2008
• Status: Active Grant

• Current Assignee: Yahoo AD Tech LLC
• Original Assignee: Platform-A Incorporated
• Inventors: Holtan, Hans Marius, Luenberger, Robert Alden, Grebeck, Michael Jan, Flores, Aaron E., Amini, Ali Nasiri, Darvish, Alireza
• Granted Patent: US 8,301,497 B2
• 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 30/0243 Comparative campaigns
• CPC Class Codes: G06Q 30/0247 Calculate past, present or future revenues

Abstract

A method, apparatus, and computer-readable medium estimate media performance on advertising space inventory. The method selects at least one media cell that shares one or more common attributes with a target media cell. The method subsequently estimates mean revenue per impression (RPI) of the selected media cell, and then defines an initial estimate of a RPI of the target media cell based on the estimated RPI of the selected cell. The method computes the RPI of the target media cell by combining the initial RPI estimate for the target media cell with performance data associated with the target media cell.

Claims

1. A computer-based method for estimating media performance on advertising space inventory, comprising:

selecting at least one media cell that shares one or more common attributes with a target media cell;

estimating a mean revenue per impression of the selected media cell; and

predicting a revenue per impression of the target media cell based on the estimated mean revenue per impression of the selected media cell.

2. The method of claim 1, wherein predicting comprises:

computing a weighted average of (i) the estimated mean revenue per impression of the selected media cell and (ii) measurements of revenue per impression for the target media cell.

3. The method of claim 1, wherein estimating comprises:

predicting an error associated with the mean revenue per impression of the selected media cell.

4. The method of claim 3, wherein estimating further comprises, when predicted error falls below a threshold value:

calculating a distribution of revenue per impression for the selected media cell; and

computing the mean revenue per impression for the selected media cell based on the calculated distribution.

5. The method of claim 4, wherein computing comprises:

using an integrated maximum likelihood approach to compute the mean revenue per impression from the calculated distribution.

6. The method of claim 3, wherein estimating further comprises, when the predicted error exceeds a threshold value:

computing the mean revenue per impression of the selected media cell based on a relationship between a plurality of attributes of one or more additional media cells and an estimated mean revenue per impression of the one or more additional media cells,

wherein the estimated mean revenue per impression of the one or more additional media cells is based on a distribution of revenue per impression.

7. The method of claim 6, wherein computing comprises:

determining the plurality of factors that describe the relationship between the attributes of the one or more additional media cells and the estimated mean revenue per impression of the one or more additional media cells, wherein each of the plurality of factors is associated with a corresponding one of the plurality of attributes; and

calculating the mean revenue per impression of the selected media cell based on a logarithmic-linear combination of the plurality of factors.

8. The method of claim 7, wherein determining comprises:

using a logarithmic regression to determine the plurality of factors.

9. The method of claim 6, wherein the plurality of attributes comprise a website, a slot size, and a segmentation model associated with each additional media cell.

10. The method of claim 1, wherein estimating comprises:

adjusting the estimated mean revenue per impression of the selected media cell to account for one or more of audience leadback and advertiser leadback.

11. The method of claim 1, wherein the common attributes comprise a common tract of advertising space inventory, a common revenue model, a common slot size, a common segmentation model, a common media, and a common industry campaign.

12. An apparatus, comprising:

a storage device; and

a processor coupled to the storage device, wherein the storage device stores a program for controlling the processor, and wherein the processor, being operative with the program, is configured to;

select at least one media cell that shares one or more common attributes with a target media cell;

estimate a mean revenue per impression of the selected media cell; and

predict a revenue per impression of the target media cell based on the estimated mean revenue per impression of the selected media cell.

13. The apparatus of claim 12, wherein the processor is configured to predict the revenue per impression of the target media cell by:

computing a weighted average of (i) the estimated mean revenue per impression of the selected media cell and (ii) measurements of revenue per impression for the target media cell.

14. The apparatus of claim 12, wherein the processor is configured to estimate the mean revenue per impression of the selected media cell by:

predicting an error associated with the mean revenue per impression of the selected media cell.

15. The apparatus of claim 14, wherein the processor is further configured to estimate the mean revenue per impression of the selected media cell, when the predicted error falls below a threshold value, by:

calculating a distribution of revenue per impression for the selected media cell; and

computing the mean revenue per impression for the selected media cell based on the calculated distribution.

16. The apparatus of claim 15, wherein the processor is configured to compute the mean revenue per impression for the selected media cell by:

using an integrated maximum likelihood approach to compute the mean revenue per impression from the estimated distribution.

17. The apparatus of claim 14, wherein the processor is configured to estimate the mean revenue per impression of the selected media cell, when the predicted error exceeds a threshold value, by:

computing the mean revenue per impression of the selected media cell based on a relationship between a plurality of attributes of one or more additional media cells and an estimated mean revenue per impression of the one or more additional media cells,

wherein the estimated mean revenue per impression of the one or more additional media cells is based on a distribution of revenue per impression.

18. The apparatus of claim 17, wherein the processor is configured to compute the mean revenue per impression of the selected media cell by:

determining the plurality of factors that describe the relationship between the attributes of the one or more additional media cells and the estimated mean revenue per impression of the one or more additional media cells, wherein each of the plurality of factors is associated with a corresponding one of the plurality of attributes; and

calculating the mean revenue per impression of the selected media cell based on a logarithmic-linear combination of the plurality of factors.

19. The apparatus of claim 18, wherein the processor is configured to determine the plurality of factors by:

using a logarithmic regression to determine the plurality of factors.

20. The apparatus of claim 17, wherein the plurality of attributes comprise a website, a slot size, and a segmentation model associated with each additional media cell.

21. The apparatus of claim 12, wherein the processor is configured to estimate the mean revenue per impression of the selected media cell by:

adjusting the estimated mean revenue per impression of the selected media cell to account for one or more of audience leadback and advertiser leadback.

22. The apparatus of claim 12, wherein the common attributes comprise a common tract of advertising space inventory, a common revenue model, a common slot size, a common segmentation model, a common media, and a common industry campaign.

23. A tangible computer readable medium comprising a set of instructions that, when executed on a processor, perform a method for estimating media performance on advertising space inventory, the method comprising:

selecting at least one media cell that share one or more common attributes with a target media cell;

estimating a mean revenue per impression of the selected media cell; and

predicting a revenue per impression of the target media cell based on the estimated mean revenue per impression of the selected media cell.

24. The computer readable medium of claim 23, wherein predicting comprises:

computing a weighted average of (i) the estimated mean revenue per impression of the selected media cell and (ii) measurements of revenue per impression for the target media cell.

25. The computer readable medium of claim 23, wherein estimating comprises:

predicting an error associated with the mean revenue per impression of the selected media cell.

26. The computer readable medium of claim 25, wherein estimating further comprises, when the predicted error falls below a threshold value:

calculating a distribution of revenue per impression for the selected media cell; and

computing the mean revenue per impression for the selected media cell based on the calculated distribution.

27. The computer readable medium of claim 26, wherein computing comprises:

using an integrated maximum likelihood approach to compute the mean revenue per impression from the calculated distribution.

28. The computer readable medium of claim 25, wherein estimating further comprises, when the predicted error exceeds a threshold value:

computing the mean revenue per impression of the selected media cell based on a relationship between a plurality of attributes of one or more additional media cells and an estimated mean revenue per impression of the one or more additional media cells,

wherein the estimated mean revenue per impression of the one or more additional media cells is based on a distribution of revenue per impression.

29. The computer readable medium of claim 28, wherein computing comprises:

determining the plurality of factors that describe the relationship between the attributes of the one or more additional media cells and the estimated mean revenue per impression of the one or more additional media cells, wherein each of the plurality of factors is associated with a corresponding one of the plurality of attributes; and

calculating the mean revenue per impression of the selected media cell based on a logarithmic-linear combination of the plurality of factors.

30. The computer readable medium of claim 29, wherein determining comprises:

using a logarithmic regression to determine the plurality of factors.

31. The computer readable medium of claim 28, wherein the plurality of attributes comprise a website, a slot size, and a segmentation model associated with each additional media cell.

32. The computer readable medium of claim 23, wherein estimating comprises:

adjusting the estimated mean revenue per impression of the selected media cell to account for one or more of audience leadback and advertiser leadback.

33. The computer readable medium of claim 23, wherein the common attributes further comprise a common tract of advertising space inventory, a common revenue model, a common slot size, a common segmentation model, a common media, and a common industry campaign.