Systems and methods for promoting groups
Systems, methods, and non-transitory computer readable media configured to receive an initial targeting for an advertisement. The advertisement can promote a group of a social networking system. A refined targeting for the advertisement can be generated. The advertisement based on the refined targeting can be delivered.
1. A computer-implemented method comprising:
receiving, by a computing system, an initial targeting for an advertisement based at least in part on a selected set of users, wherein the advertisement promotes a group of a system and includes an interaction event;
determining, by the computing system, clusters associated with the selected set of users, wherein the clusters are associated with demographics and interests of users of the system and include at least one user of the selected set of users and at least one other user of the system, wherein the clusters are determined based on a machine learning model of the computing system trained based on training data that includes the demographics and the interests of the users of the system;
generating, by the computing system, one or more suggested targetings based at least in part on a selection of a cluster of the determined clusters that satisfies a threshold size value;
providing, by the computing system, the one or more suggested targetings;
receiving, by the computing system, a refined targeting for the advertisement based at least in part on a selection of at least one of the one or more suggested targetings;
generating, by the computing system, ranks for the interaction event associated with the advertisement with respect to the users of the refined targeting based at least in part on indications of willingness of the users to join groups, wherein the indications of willingness include a quantity of group join requests sent by the users within a selected period of time, wherein the generating comprises:
providing, by the computing system, to the machine learning model, user feature data, including at least an indication of willingness to join groups, for the users of the refined targeting;
providing, by the computing system, to the machine learning model, group feature data, including at least one of: a name of the group, a cover photo of the group, a description of the group, tags of the group, locations associated with the group, topics discussed in the group, and feature data of users who have joined the group, for the group of the system;
receiving, by the computing system, from the machine learning model, a prediction of likelihood of satisfying an interaction event with respect to the group for each user of the users of the refined targeting; and
ranking, by the computing system, the users of the refined targeting based on the prediction of likelihood of satisfying the interaction event associated with the advertisement for each of the users of the refined targeting; and
delivering, by the computing system, the advertisement based at least in part on the refined targeting and the ranks for the interaction event associated with the advertisement, wherein the advertisement is delivered to the users of the refined targeting that satisfy a threshold rank.
2. The computer-implemented method of claim 1 , wherein the delivering the advertisement comprises:
placing, by the computing system, the advertisement in news feeds of the users of the refined targeting based at least in part on the ranks for the advertisement with respect to the users.
3. The computer-implemented method of claim 1 , wherein the
interaction event includes at least one of j oining the group, creating a post for the group, or creating a comment for the group.
4. The computer-implemented method of claim 1 , wherein:
the user feature data includes one or more of social data for the users, indications of kinds of groups the users join, or indications of kinds of pages the users follows; and
the group feature data includes the user feature data of users who have joined the group.
5. The computer-implemented method of claim 1 , wherein the generating the refined targeting comprises:
changing, by the computing system, the initial targeting for the advertisement to include some or all of a targeting which is characteristic of a reference set of users of the system.
6. The computer-implemented method of claim 5 , further comprising:
providing, by the computing system, to the machine learning model, for each user of one or more users of the reference set of users, user feature data of the user;
receiving, by the computing system, from the machine learning model, the clusters, wherein the clusters include users with similar user feature data;
selecting, by the computing system, one or more of the clusters; and
employing, by the computing system, the demographics and the interests of the selected one or more clusters as the targeting which is characteristic of the reference set of users.
7. The computer-implemented method of claim 5 , wherein the reference set of users comprises existing users of the group promoted by the advertisement, users of a group other than the group promoted by the advertisement, or individually-selected users of the system.
8. The computer-implemented method of claim 1 , further comprising:
receiving, by the computing system, an initial budget for the advertisement; and
generating, by the computing system, a refined budget for the advertisement,
wherein the delivering of the advertisement is further based at least in part on the refined budget.
9. The computer-implemented method of claim 8 , wherein the generating the refined budget comprises:
providing, by the computing system, to a machine learning model, feature data for the advertisement;
receiving, by the computing system, from the machine learning model, a predicted budget for the advertisement; and
changing, by the computing system, the initial budget to be the predicted budget.
10. The computer-implemented method of claim 9 , wherein the feature data for the advertisement comprises one or more of a size of the group or a growth rate of the group.
11. A system comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform:
receiving an initial targeting for an advertisement based at least in part on a selected set of users, wherein the advertisement promotes a group of the system and includes an interaction event;
determining clusters associated with the selected set of users, wherein the clusters are associated with demographics and interests of users of the system and include at least one user of the selected set of users and at least one other user of the system, wherein the clusters are determined based on a machine learning model of the computing system trained based on training data that includes the demographics and the interests of the users of the system;
generating one or more suggested targetings based at least in part on a selection of a cluster of the determined clusters that satisfies a threshold size value;
providing the one or more suggested targetings;
receiving a refined targeting for the advertisement based at least in part on a selection of at least one of the one or more suggested targetings;
generating ranks for the interaction event associated with the advertisement with respect to the users of the refined targeting based at least in part on indications of willingness of the users to join groups, wherein the indications of willingness include a quantity of group join requests sent by the users within a selected period of time, wherein the generating comprises:
providing, to the machine learning model, user feature data, including at least an indication of willingness to join groups, for the users of the refined targeting;
providing, to the machine learning model, group feature data, including at least one of: a name of the group, a cover photo of the group, a description of the group, tags of the group, locations associated with the group, topics discussed in the group, and feature data of users who have joined the group, for the group of the system; and
receiving, from the machine learning model, a prediction of likelihood of satisfying an interaction event with respect to the group for each user of the users of the refined targeting; and
ranking the users of the refined targeting based on the prediction of likelihood of satisfying the interaction event associated with the advertisement for each of the users of the refined targeting; and
delivering the advertisement based at least in part on the refined targeting and the ranks for the interaction event associated with the advertisement, wherein the advertisement is delivered to the users of the refined targeting that satisfy a threshold rank.
12. The system of claim 11 , wherein the delivering the advertisement comprises:
placing the advertisement in news feeds of the users of the refined targeting based at least in part on the ranks for the advertisement with respect to the users.
13. The system of claim 12 , wherein the
interaction event includes at least one of j oining the group, creating a post for the group, or creating a comment for the group.
14. The system of claim 11 , wherein the generating the refined targeting comprises:
changing the initial targeting for the advertisement to include some or all of a targeting which is characteristic of a reference set of users of the system.
15. The system of claim 11 , wherein the instructions, when executed by the at least one processor, further cause the system to perform:
receiving an initial budget for the advertisement; and
generating a refined budget for the advertisement, wherein the delivering of the advertisement is further based at least in part on the refined budget.
16. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform a method comprising:
receiving an initial targeting for an advertisement based at least in part on a selected set of users, wherein the advertisement promotes a group of a system and includes an interaction event;
determining clusters associated with the selected set of users, wherein the clusters are associated with demographics and interests of users of the system and include at least one user of the selected set of users and at least one other user of the system, wherein the clusters are determined based on a machine learning model of the computing system trained based on training data that includes the demographics and the interests of the users of the system;
generating one or more suggested targetings based at least in part on a selection of a cluster of the determined clusters that satisfies a threshold size value;
providing the one or more suggested targetings;
receiving a refined targeting for the advertisement based at least in part on a selection of at least one of the one or more suggested targetings;
generating ranks for the interaction event associated with the advertisement with respect to the users of the refined targeting based at least in part on indications of willingness of the users to join groups, wherein the indications of willingness include a quantity of group join requests sent by the users within a selected period of time, wherein the generating comprises:
providing, to the machine learning model, user feature data, including at least an indication of willingness to join groups, for the users of the refined targeting;
providing, to the machine learning model, group feature data, including at least one of: a name of the group, a cover photo of the group, a description of the group, tags of the group, locations associated with the group, topics discussed in the group, and feature data of users who have joined the group, for the group of the system;
receiving, from the machine learning model, a prediction of likelihood of satisfying an interaction event with respect to the group for each user of the users of the refined targeting; and
ranking the users of the refined targeting based on the prediction of likelihood of satisfying the interaction event associated with the advertisement for each of the users of the refined targeting; and
delivering the advertisement based at least in part on the refined targeting and the ranks for the interaction event associated with the advertisement, wherein the advertisement is delivered to the users of the refined targeting that satisfy a threshold rank.
17. The non-transitory computer-readable storage medium of claim 16 , wherein the delivering the advertisement comprises:
placing the advertisement in news feeds of the users of the refined targeting based at least in part on the ranks for the advertisement with respect to the users.
18. The non-transitory computer-readable storage medium of claim 17 , wherein the
interaction event includes at least one of j oining the group, creating a post for the group, or creating a comment for the group.
19. The non-transitory computer-readable storage medium of claim 16 , wherein the generating the refined targeting comprises:
changing the initial targeting for the advertisement to include some or all of a targeting which is characteristic of a reference set of users of the system.
20. The non-transitory computer-readable storage medium of claim 16 , wherein the instructions, when executed by the at least one processor of the computing system, further cause the computing system to perform:
receiving an initial budget for the advertisement; and
generating a refined budget for the advertisement, wherein the delivering of the advertisement is further based at least in part on the refined budget.