Continuous value-per-click estimation for low-volume terms

• US 8,117,066 B1
• Filed: 07/09/2008
• Issued: 02/14/2012
• Est. Priority Date: 07/09/2007
• Status: Active Grant

• Current Assignee: Marin Software
• Original Assignee: Marin Software
• Inventors: Walcott, Wister, Reichling, Florian, Barnett, Joseph L., Betz, Martin S.
• Primary Examiner(s): Poinvil, Frantzy
• Application Number:
  US12/170,378
• Time in Patent Office: 1,315 Days
• Field of Search: 705 10- 44
• US Class Current: 705/14
• CPC Class Codes: G06Q 30/0241 Advertisements
• CPC Class Codes: G06Q 30/0242 Determining effectiveness of advertisements
• CPC Class Codes: G06Q 30/0275 Auctions
• CPC Class Codes: G06Q 40/12 Accounting

Abstract

An estimated conversion rate for a desired advertisement term is calculated. A total number of conversion events for terms having a conversion rate is determined for each of a plurality of conversion rates, and data is generated that describes a first distribution function representing the conversion rates and the associated numbers of conversion rates. Additionally, data describing a second distribution function representing the probability of a given number of conversion events occurring is also generated. Based at least in part on the first distribution function and the second distribution function, the estimated conversion rate for the desired advertisement term is computed.

Claims

1. A computer-implemented method of computing an estimated conversion rate for a desired advertisement term, the advertisement term being a member of a folder of related advertisement terms, the method comprising:

performing, by a computer processor, steps comprising;

storing statistics for the folder of related terms, the statistics comprising a number of clicks and a number of conversion events for each of the terms;

determining, for each conversion rate of a plurality of conversion rates, a total number of conversion events for terms having the conversion rate;

generating data describing a first distribution function that represents a distribution of the total number of conversion events over the plurality of conversion rates;

generating data describing a second distribution function that represents probabilities of different numbers of conversion events occurring;

computing the estimated conversion rate for the desired advertisement term based at least in part on the data describing the first distribution function and the data describing the second distribution function; and

storing the estimated conversion rate.

2. The computer-implemented method of claim 1, wherein the determining, generating, and computing steps are performed responsive at least in part to a given statistic of the statistics for the advertisement term being below a threshold.

3. The computer-implemented method of claim 2, wherein the given statistic is the number of clicks for the advertisement term and the threshold is 100.

4. The computer-implemented method of claim 2, the steps further comprising computing the estimated conversion rate based only on the stored statistics for the advertisement term, responsive at least in part to the given statistic of the statistics for the advertisement term exceeding the threshold.

5. The computer-implemented method of claim 1, wherein the first distribution function is a beta distribution describing a distribution of the total number of conversion events over the plurality of conversion rates.

6. The computer-implemented method of claim 1, wherein the second distribution function is a binomial distribution describing probabilities of different numbers of conversion events occurring for a given number of clicks of the advertisement term.

7. The computer-implemented method of claim 1, the steps further comprising computing an estimated optimal bid for an advertiser to submit for an advertisement associated with the advertisement term based at least in part on the computed estimated conversion rate.

8. The computer-implemented method of claim 7, the steps further comprising computing an adjusted bid by increasing the estimated optimal bid responsive to at least one of a minimum bid imposed by an advertising publisher to which the adjusted bid is submitted, and remaining funds in an advertising budget of the advertiser.

9. The computer-implemented method of claim 1, the steps further comprising obtaining the stored statistics for the plurality of related terms from an advertising publisher.

10. A computer program product for computing an estimated conversion rate for a desired advertisement term, the advertisement term being a member of a folder of related advertisement terms, the computer program product comprising:

a computer-readable storage medium containing computer program code that when executed by a computer processor performs steps comprising;

storing statistics for the folder of related terms, the statistics comprising a number of clicks and a number of conversion events for each of the terms;

determining, for each conversion rate of a plurality of conversion rates, a total number of conversion events for terms having the conversion rate;

generating data describing a first distribution function that represents a distribution of the total number of conversion events over the plurality of conversion rates;

generating data describing a second distribution function that represents probabilities of different numbers of conversion events occurring;

computing the estimated conversion rate for the desired advertisement term based at least in part on the data describing the first distribution function and the data describing the second distribution function; and

storing the estimated conversion rate.

11. The computer program product of claim 10, wherein the determining, generating, and computing steps are performed responsive at least in part to a given statistic of the statistics for the advertisement term being below a threshold.

12. The computer program product of claim 11, wherein the given statistic is the number of clicks for the advertisement term and the threshold is 100.

13. The computer program product of claim 11, the steps performed by the computer program code further comprising computing the estimated conversion rate based only on the stored statistics for the advertisement term, responsive at least in part to the given statistic of the statistics for the advertisement term exceeding the threshold.

14. The computer program product of claim 10, the steps performed by the computer program code further comprising obtaining the stored statistics for the plurality of related terms from an advertising publisher.

15. A computer system configured to compute an estimated conversion rate for a desired advertisement term, the advertisement term being a member of a folder of related advertisement terms, comprising:

a term statistics database comprising storing statistics for the folder of related terms, the statistics comprising a number of clicks and a number of conversion events for each of the terms; and

an information analysis module executed by a computer processor and configured to;

determine, for each conversion rate of a plurality of conversion rates, a total number of conversion events for terms having the conversion rate;

generate data describing a first distribution function that represents a distribution of the total number of conversion events over the plurality of conversion rates;

generate data describing a second distribution function that represents probabilities of different numbers of conversion events occurring; and

compute the estimated conversion rate for the desired advertisement term based at least in part on the data describing the first distribution function and the data describing the second distribution function; and

store the estimated conversion rate within the term statistics database.

16. The system of claim 15, wherein the first distribution function is a beta distribution describing a distribution of the total number of conversion events over the plurality of conversion rates.

17. The system of claim 15, wherein the second distribution function is a binomial distribution describing probabilities of different numbers of conversion events occurring for a given number of clicks of the advertisement term.

18. The system of claim 15, further comprising a bid optimization module configured to compute an estimated optimal bid for an advertiser to submit for an advertisement associated with the advertisement term based at least in part on the computed estimated conversion rate.

19. The system of claim 18, wherein the bid optimization module is further configured to compute an adjusted bid by increasing the estimated optimal bid responsive to at least one of a minimum bid imposed by an advertising publisher to which the adjusted bid is submitted, and remaining funds in an advertising budget of the advertiser.

20. The computer system of claim 15, wherein to determine, generate and computer functions are performed responsive at least in part to the number of clicks for the advertisement term being below 100.