Audience Expansion leverages a user-based look-alike model that uses machine-learning to assign a “similarity score” to people who have similar characteristics to the targeted audience. These people are then bucketed into groups who closely, somewhat closely, or loosely resemble the targeted audience.
Advertisers are able to extend their campaign’s reach to include these similar people by enabling Audience Expansion and choosing from 3 levels of expansion:
- Defined - reach people that most closely match your targeted audience
- Expanded - reach people that closely or somewhat match your targeted audience.
- Includes people in the “Defined” bucket
- Broad - reach people that closely, somewhat, or loosely match your targeted audience.
- Includes people in the “Defined” & “Expanded” buckets
Based on the level of expansion chosen, the campaign’s targeting will automatically be extended to include people who meet the similarity criteria.