AUDIENCES - MARKETPLACE
AUDIENCES - ACCESS
- Access MediaMath Segments
- Access Twitter Segments
- Access Facebook Segments
- Access Amobee (Turn) DSP Segments
- Snap Integration Setup
INSIGHTS
- Welcome to Insights!
- Insights - Starter Guide
- Insights - Report Guide
- Insights - FAQs
- Insights - Release Notes
MEASUREMENT
Create Lookalike Models
Lookalike audiences are created by analyzing the app graph distribution of users in the original audience, and finding users who have statistically similar app graphs. You may recognize this as being similar to “users who bought x also bought y” feature most online retailers use.
Let's create some lookalike segments now.
Looking for Scale AND Relevance?
We suggest you stay within the green (high relevance) and the yellow (medium relevance) to gain scale while keeping your segment of users relevant.
Hey, my lookalike segment is larger than my target!
This sometimes happens because we create clusters of users who are similar to each other, and we add one cluster at a time into your expansion audience. As a result, we may overshoot the target because the next cluster to be added exceeds the target.
Learn how to activate your lookalike segment here.