CONTENT MARKETING & DISCOVERY

What If you Knew What Interests Your Customers Have in Common?

|Lior Charka

New Feature: lookalike audiences

If marketing is an art and a science, the art comes in crafting the right message, and the science is in finding the right audience to hear it.

But content consumption patterns are notoriously complex. That’s why content marketing requires so much effort—and continues to be a long-game. That is… until now.

Outbrain is gradually rolling out a powerful new solution called Lookalike Audiences that helps marketers find and engage with new people who “look like” existing customers and converters. The solution is available globally to managed accounts and will soon be accessible in our self-serve dashboard.

Now we are able to model similar audiences based on demonstrated common interests and what they are most likely to read and watch. We do this by marrying our user interest graph data with 1st party data to model audiences that resemble your existing customers and converters.

And it works—early results from over 30 brands showed that on average Outbrain Lookalike Audiences drove 45% more conversions with a 30% lower CPA.

Outbrain Lookalike’s Unique Value

Outbrain’s Lookalike Audiences gives you access to new and powerful data about your top-converters based on their category interests and what they read and watch. Moreover, our Lookalikes is also unique in its ability to allow you to model and scale audiences based on common interests that are most likely to lead to a desired action. For example, a purchase, a lead, or a sign-up.

Maybe you’ve used Facebook’s Lookalike Audiences tool before or Google Similar Audiences. If so, you know how powerful both can be. Outbrain’s Lookalikes adds a new layer to your efforts by breaking audience targeting out of the walled garden of social media and search, extending it across our global reach of one billion  users. And we do it based on users’ real interests, not the ones which may be gathered from their social actions when they are in a social mindset.

Lookalike Audiences in Action

To create a lookalike audience, the modeling tool can either use first-party data like DMP or CRM or use site cookie pools. The closer the seed audience is to the brand’s top customers, the better performance a lookalike audience can drive. Lookalike audiences can be optimized for either more precision or more scale, depending on the goals of the campaign.

For example, bed sheet and linen e-commerce brand, Brooklinen, modeled its lookalike audience from information it collected from its confirmation page. It then targeted this Lookalike audience with its top performing headlines and images and positive earned media.

Its efforts ultimately paid off. Desktop Lookalike audiences ultimately outperformed all run of network and publisher vertical prospecting campaigns. The lookalike audience drove 2.2x higher CVR vs. the control run of the network campaign, and delivered a 50% lower CPA. (Read the full Brooklinen case study here.)

Conclusion

Outbrain is the only discovery platform to offer lookalike audience targeting. Our Lookalikes offering uniquely drives performance and efficiency by building conversion audience pools off of first party data segments and continually optimizing to learn audiences interests in real time. While Lookalike Audiences is a powerful standalone tool, it can also work in harmony with search and social targeting, scaling audiences outside of the confines of a single social network.

As you can see, we’re very excited about the possibilities of this new feature. Contact Outbrain to get started, learn more about Lookalikes here, and read more about how Brooklinen used Outbrain Lookalike Audiences in AdExchanger, MediaPost, and Marketing Tech.

Start your Outbrain campaign!

Lior Charka

Lior Charka

Lior is a Product Manager at Outbrain Amplify. He's just a city boy, born and raised in South Detroit. He... Read more

Add A Comment

* Your email address will not be published. All fields are required.

Loading

Want to promote your content?