Outbrain Rolls Out 'Lookalike Audiences'

Content discovery and personalization platform Outbrain has launched a tool called Lookalike Audiences, to help marketers reach and engage with new people who are likely to be interested in a business because they 'look similar' to existing customers.

New York-based Outbrain delivers personalized online, mobile and video content to audiences, while using data to help publishers understand them. Lookalike Audiences helps find similar audiences, based on demonstrated common interests and behavior, then targets these audiences at scale across the firm's publisher network. By combining the tool with Outbrain's proprietary interest graph, brands can connect with the audiences with the highest propensity to become customers based on the content they are engaging with.

Moran Gavish (pictured), the data scientist leading the research and development on the lookalike framework, explains how the algorithm works: 'A marketer gives Outbrain a list of valuable users, for example - those who have made a purchase. We then use machine learning models, such as logistic regression, decision trees and matrix factorization to characterise these valuable users' content interests. Such interests may include the main content categories they read and are not likely to read, publishers they visit and are not likely to visit, personas and companies they're interested in etc. Using these models, we identify in real-time a user who is not included in the marketer's list, but like those users, and recommend them with campaigns by that marketer'.

Web site: www.outbrain.com .

All articles 2006-23 written and edited by Mel Crowther and/or Nick Thomas, 2024- by Nick Thomas, unless otherwise stated.