Systems, methods, and articles of manufacture to measure online audiences
Methods and apparatus to monitor media content at a content display site are described. An example method includes assigning a first set of weights to each of a plurality of panelists based on a set of collected characteristics for each person of a subset of unmeasured persons at unmeasured locations, the subset of unmeasured persons being the unmeasured person for which the collected characteristics are known, selecting a subset of the panelists based on the first set of weights, each panelist of the subset of panelists selected to be representative of an unmeasured person of the unmeasured persons, re-weighting the subset of the panelists based on estimated characteristics of the unmeasured persons at the unmeasured locations to generate a second set of weights, generating a virtual panel including measurement data collected for the subset of the panelists as virtual panelists, the measurement data of the virtual panel representing the media consumption activity of the unmeasured persons at the unmeasured locations, and scaling the measurement data of the virtual panel collected from the subset of the panelists to estimate media consumption at the unmeasured locations.
1. A method to generate a virtual panel, the method comprising:
assigning a first set of weights to each of a plurality of panelists based on a set of collected characteristics for each person of a subset of unmeasured persons at unmeasured locations, the subset of unmeasured persons being the unmeasured persons for which the collected characteristics are known;
selecting a subset of the panelists based on the first set of weights, each panelist of the subset of panelists selected to be representative of an unmeasured person of the unmeasured persons;
re-weighting the subset of the panelists based on estimated characteristics of the unmeasured persons at the unmeasured locations to generate a second set of weights;
generating a virtual panel including measurement data collected for the subset of the panelists as virtual panelists, the measurement data of the virtual panel representing the media consumption activity of the unmeasured persons at the unmeasured locations; and
scaling the measurement data of the virtual panel collected from the subset of the panelists to estimate media consumption at the unmeasured locations.
2. The method of claim 1 , wherein the unmeasured persons access a web page from a first unmeasured location.
3. The method of claim 2 , wherein the unmeasured location is a type of location at which no panel measurement data has been collected.
4. The method of claim 1 , further including estimating an audience, the estimated characteristics of the panelists being based on the estimated audience.
5. The method of claim 4 , wherein the estimated characteristics are based on a difference between the estimated audience and a measured audience.
6. The method of claim 1 , wherein the selecting of the subset of the panelists includes randomly selecting the subset of panelists using the first set of weights assigned to each person of the subset of panelists as selection probabilities for the corresponding panelist.
7. The method of claim 1 , further including calculating the estimated characteristics based on at least one of a demographic target, an activity target, an audience target, or an entity target.
8. The method of claim 7 , wherein the calculating of the estimated characteristics includes comparing online activities of the panelists to census-based page view data.
9. An apparatus comprising:
a sample weighter to assign a first set of weights to each of a plurality of panelists based on a set of collected characteristics for each person of a subset of unmeasured persons at unmeasured locations, the subset of unmeasured persons being the unmeasured persons for which the collected characteristics are known;
a sample selector to select a subset of the panelists based on the first set of weights, each panelist of the subset of panelists selected to be representative of an unmeasured person of the unmeasured persons;
a sample re-weighter to re-weight the subset of the panelists based on estimated characteristics of the unmeasured persons at the unmeasured locations to generate a second set of weights;
a virtual panel generator to generate a virtual panel including measurement data collected for the subset of panelists as virtual panelists, the virtual panel representing the media consumption activity of the unmeasured persons at the unmeasured locations; and
an activity fulfiller to scale the measurement data of the virtual panel collected from the selected panelists to estimate media consumption at the unmeasured locations.
10. The apparatus of claim 9 , wherein the unmeasured persons access a web page from an unmeasured location.
11. The apparatus of claim 10 , wherein the unmeasured location is a type of location at which no panel measurement data has been collected.
12. The apparatus of claim 9 , wherein the sample weighter is to:
obtain an estimated audience; and
calculate the estimated characteristics of the panelists based on the estimated audience.
13. The apparatus of claim 12 , wherein the sample weighter is to calculate the estimated characteristics based on a difference between the estimated audience and a measured audience.
14. The apparatus of claim 9 , wherein the sample selector is to select the panelists in the subset by randomly selecting the panelists using the first set of weights assigned to corresponding panelists as selection probabilities for the corresponding panelists.
15. The apparatus of claim 9 , wherein the sample weighter is to calculate the estimated characteristics based on at least one of a demographic target or an activity target.
16. The apparatus of claim 15 , wherein the sample weighter is to calculate the estimated characteristics by comparing online activities of the panelists to measured page view data.
17. A computer readable medium comprising instructions that, when executed, cause a processor to:
assign a first set of weights to each of a plurality of panelists based on a set of collected characteristics for each person of a subset of unmeasured persons at unmeasured locations, the subset of the unmeasured persons being the unmeasured persons for which the collected characteristics are known;
select a subset of the panelists based on the first set of weights, each panelist of the subset of panelists selected to be representative of an unmeasured person of the unmeasured persons;
re-weight the selected panelists based on estimated characteristics of the unmeasured person at the unmeasured locations to generate a second set of weights;
generate a virtual panel including measurement data collected for the subset of panelists as virtual panelists, the measurement data of the virtual panel representing the media consumption activity of the unmeasured persons at the unmeasured locations; and
scale the measurement data of the virtual panel collected from the selected panelists to estimate media consumption at the unmeasured locations.
18. The computer readable medium of claim 17 , wherein the unmeasured persons access a web page from an unmeasured location.
19. The computer readable medium of claim 18 , wherein the unmeasured location is a type of location at which no panel measurement data has been collected.