Generating Audience Metrics Including Affinity Scores Relative to An Audience
A social networking system receives a selection of user characteristics defining a benchmark audience and a target audience, and generates audience metrics that compare the audiences across a set of user characteristics. These user characteristics include demographics, interests, purchasing activity, and actions on the social networking system. The audience metrics are provided to an advertiser who may select additional user characteristics to refine the benchmark or target audiences. The audience metrics may include an affinity score that compares the audience metrics for a particular type of interaction, and may normalize the frequency of interactions relative to interactions of the audience as a whole. Advertisers may use the defined audiences to establish targeting criteria for an advertisement, and may use existing targeting criteria to seed the selection of an audience.
1 . A method comprising:
receiving a benchmark audience of users of an online system and a target audience of users of the online system;
for each user interest of a plurality of user interests, generating an affinity score for the user interest by:
computing a first frequency of occurrence of the user interest for the users of the target audience,
computing a second frequency of occurrence of the user interest for the users of the benchmark audience,
comparing the first and second computed frequencies, and
generating the affinity score based on the comparison of the first and second frequencies;
selecting one or more outlier user interests from the user interests based on the generated affinity scores; and
providing the selected outlier user interests and the generated affinity score associated with the selected outlier user characteristics for display to a user device.
2 . The method of claim 1 , wherein generating the affinity score further comprises:
normalizing the affinity scores based on a frequency of user interests associated with the target audience relative to a frequency of user interests associated with the benchmark audience.
3 . The method of claim 1 , wherein the user interests comprise pages interacted with by the users of the online system.
4 . The method of claim 1 , wherein the user interests comprise events interacted with by the users of the online system.
5 . The method of claim 1 , wherein the one or more outliers are selected based on a ranking of the plurality of user interests by affinity scores.
6 . The method of claim 1 , wherein the affinity score is generated using the following equation:
A
C
=
T
C
I
T
T
C
⋃
B
C
I
T
⋃
I
B
wherein
Ac is the affinity score for user interest C relative to target audience T and benchmark audience B;
T C is the number of users in the target audience with user interest C;
B C is the number of users in the benchmark audience with user interest C;
I T is the total number of interactions of an interaction type corresponding to the user interest performed by the target audience; and
I B is the total number of interactions of the interaction type performed by the benchmark audience.
7 . The method of claim 1 , further comprising generating a relevancy score for each user interest in the plurality of user interests, the relevancy score indicating the affinity score adjusted for the number of users in the target audience associated with the user interest.
8 . The method of claim 7 , wherein the relevancy score is used to select the one or more outlier user interests.
9 . The method of claim 7 , wherein the relevancy score is generated using the following equation:
R
C
=
T
C
T
*
(
A
C
-
1
)
wherein
R C is the relevancy score for a user interest C;
Tc is number of users in the target audience T with user interest C;
Tis the number of users in the target audience; and
Ac is the affinity score for user characteristic C.
10 . A non-transitory computer-readable medium comprising instructions that when executed by a processor cause the processor to perform steps of:
receiving a benchmark audience of users of an online system and a target audience of users of the online system;
for each user interest of a plurality of user interests, generating an affinity score for the user interest by:
computing a first frequency of occurrence of the user interest for the users of the target audience,
computing a second frequency of occurrence of the user interest for the users of the benchmark audience,
comparing the first and second computed frequencies, and
generating the affinity score based on the comparison of the first and second frequencies;
selecting one or more outlier user interests from the user interests based on the generated affinity scores; and
providing the selected outlier user interests and the generated affinity score associated with the selected outlier user characteristics for display to a user device.
11 . The non-transitory computer-readable medium of claim 10 , wherein generating the affinity score further comprises: normalizing the affinity scores based on a frequency of user interests associated with the target audience relative to a frequency of user interests associated with the benchmark audience.
12 . The non-transitory computer-readable medium of claim 10 , wherein the user interests comprise pages interacted with by the users of the online system.
13 . The non-transitory computer-readable medium of claim 10 , wherein the user interests comprise events interacted with by the users of the online system.
14 . The non-transitory computer-readable medium of claim 10 , wherein the one or more outliers are selected based on a ranking of the plurality of user interests by affinity scores.
15 . The non-transitory computer-readable medium of claim 10 , wherein the affinity score is generated using the following equation:
A
C
=
T
C
I
T
T
C
⋃
B
C
I
T
⋃
I
B
wherein
Ac is the affinity score for user interest C relative to target audience T and benchmark audience B;
T C is the number of users in the target audience with user interest C;
B C is the number of users in the benchmark audience with user interest C;
I T is the total number of interactions of an interaction type corresponding to the user interest performed by the target audience; and
I B is the total number of interactions of the interaction type performed by the benchmark audience.
16 . The non-transitory computer-readable medium of claim 10 , the steps further comprising generating a relevancy score for each user interest in the plurality of user interests, the relevancy score indicating the affinity score adjusted for the number of users in the target audience associated with the user interest.
17 . The non-transitory computer-readable medium of claim 16 , wherein the relevancy score is used to select the one or more outlier user interests.
18 . The non-transitory computer-readable medium of claim 16 , wherein the relevancy score is generated using the following equation:
R
C
=
T
C
T
*
(
A
C
-
1
)
wherein
R C is the relevancy score for a user interest C;
Tc is number of users in the target audience T with user interest C;
Tis the number of users in the target audience; and
Ac is the affinity score for user characteristic C.