Providing data and analysis for advertising on networked devices
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for providing data and analysis for advertising on networked devices. One of the methods includes creating a vector of identifiers representing an ad opportunity. The method includes linking data attributes that describe the ad opportunity to the identifiers. The method includes expressing the data attributes following predefined scheme of hierarchy. The method includes linking a taxonomy describing data attributes. The method includes obtaining outcome measurements of ad events associated with the ad opportunity. The method also includes associating user interaction events with the ad with at least one of the identifiers or data attributes associated with the identifier.
1. A computer-implemented method for providing analysis for advertising transactions comprising:
creating, in a processor of a computer, profiles for a plurality of ad requests by aggregating inventory metrics using one or more identifiers according to an ad model, wherein:
the ad model provides a common language for conducting analysis and exchanging data between different sources;
the creating comprises reducing a necessary size of one or more databases by computing the inventory metrics at a finest granularity of a number of ad requests per pair of (i,j), wherein i is an inventory of advertising impressions and j is an advertisement presented to a user, and remaining inventory metrics are aggregated based on the pair;
the aggregated remaining inventory metrics are stored on a per request basis in the one or more databases during run time;
identifying, using the processor in the computer, a plurality of inventory vectors based on the profiles, wherein the plurality of inventory vectors uniquely identify an ad opportunity at a particular time;
creating, in the computer using the processor, a profile database, of the one or more databases, of ad performance comprising performance metrics for each inventory vector, associated data attributes, and ad identifiers;
providing, using the processor in the computer, predictions of ad performance metrics as function of time, index of inventory attributes, and the data attributes; and
providing, using the processor in the computer, a user interface to enable a user to query one or more metrics associated with the one or more identifiers.
2. The computer-implemented method of claim 1 , wherein the inventory metrics comprise at least one of number of request, click through rate, conversions rate, prices, bid floors and the time horizon.
3. The computer-implemented method of claim 1 , wherein the predictions comprise at least one of bid prices as function of index of inventory; a winning rate as a function of price and index of inventory, inventory identifiers, index of identifiers, and time; click through rate, conversion rate, and life time value.
4. The computer-implemented method of claim 3 , wherein the predictions are associated with a measure of accuracy and confidence level.
5. The computer-implemented method of claim 3 , wherein one or more data attributes are assigned a value, and incremental value in using the data.
6. The computer-implemented method of claim 1 , wherein the one or more identifiers separate the aggregated inventory metrics at a level of individual users, media, and advertisement interaction facts resulting in a user database, a media database, and an interaction fact database, of the one or more databases.
7. The computer-implemented method of claim 6 wherein the user database comprises user identifiers that are linked to identifiers associated with the user in settings of a device or application.
8. A non-transitory computer readable medium, encoded with computer program instructions that when executed by one or more computers cause the one or more computers to perform operations for providing data for providing analysis for advertising transactions comprising:
creating profiles for a plurality of ad requests by:
aggregating inventory metrics using one or more identifiers according to an ad model, wherein the ad model provides a common language for conducting analysis and exchanging data between different sources; and
the creating comprises reducing a necessary size of one or more databases by computing the inventory metrics at a finest granularity of a number of ad requests per pair of (i,j), wherein i is an inventory of advertising impressions and j is an advertisement presented to a user, and remaining inventory metrics are aggregated based on the pair;
the aggregated remaining inventory metrics are stored on a per request basis in the one or more databases during run time;
identifying a plurality of inventory vectors based on the profiles, wherein the plurality of inventory vectors uniquely identify an ad opportunity at a particular time;
creating a profile database of ad performance comprising performance metrics for each inventory vector, associated data attributes, and ad identifiers;
providing predictions of ad performance metrics as function of time, index of inventory attributes, and the data attributes; and
providing a user interface to enable a user to query one or more metrics associated with the one or more identifiers.
9. The non-transitory computer readable medium of claim 8 , wherein the inventory metrics comprise at least one of number of request, click through rate, conversions rate, prices, bid floors and the time horizon.
10. The non-transitory computer readable medium of claim 8 , wherein the predictions comprise at least one of bid prices as function of index of inventory; a winning rate as a function of price and index of inventory, inventory identifiers, index of identifiers, and time; click through rate, conversion rate, and life time value.
11. The non-transitory computer readable medium of claim 10 , wherein the predictions are associated with a measure of accuracy and confidence level.
12. The non-transitory computer readable medium of claim 10 , wherein one or more data attributes are assigned a value, and incremental value in using the data.
13. The non-transitory computer readable medium of claim 8 , wherein the one or more identifiers separate the aggregated inventory metrics at a level of individual users, media, and advertisement interaction facts resulting in a user database, a media database, and an interaction fact database, of the one or more databases.
14. The non-transitory computer readable medium of claim 13 , wherein the user database comprises user identifiers that are linked to identifiers associated with the user in settings of a device or application.
15. A system for providing analysis for advertising transactions comprising one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
creating profiles for a plurality of ad requests by:
aggregating inventory metrics using one or more identifiers according to an ad model, wherein the ad model provides a common language for conducting analysis and exchanging data between different sources;
the creating comprises reducing a necessary size of one or more databases by computing the inventory metrics at a finest granularity of a number of ad requests per pair of (i,j), wherein i is an inventory of advertising impressions and j is an advertisement presented to a user, and remaining inventory metrics are aggregated based on the pair;
the aggregated remaining inventory metrics are stored on a per request basis in the one or more databases during run time;
identifying a plurality of inventory vectors based on the profiles, wherein the plurality of inventory vectors uniquely identify an ad opportunity at a particular time;
creating a profile database of ad performance comprising performance metrics for each inventory vector, associated data attributes, and ad identifiers;
providing predictions of ad performance metrics as function of time, index of inventory attributes, and the data attributes; and
providing a user interface to enable a user to query one or more metrics associated with the one or more identifiers.
16. The system of claim 15 , wherein the inventory metrics comprise at least one of number of request, click through rate, conversions rate, prices, bid floors and the time horizon.
17. The system of claim 15 , wherein the predictions comprise at least one of bid prices as function of index of inventory; a winning rate as a function of price and index of inventory, inventory identifiers, index of identifiers, and time; click through rate, conversion rate, and life time value.
18. The system of claim 17 , wherein the predictions are associated with a measure of accuracy and confidence level.
19. The system of claim 17 , wherein one or more data attributes are assigned a value, and incremental value in using the data.
20. The system of claim 15 , wherein the one or more identifiers separate the aggregated inventory metrics at a level of individual users, media, and advertisement interaction facts resulting in a user database, a media database, and an interaction fact database, of the one or more databases.
21. The system of claim 20 wherein the user database comprises user identifiers that are linked to identifiers associated with the user in settings of a device or application.