How Content Discovery Network Algorithms Actually Work?

CDN’s first entered the ad tech landscape 10 years ago, with Outbrain being the first to market. Their core value proposition was being able to provide users with a seamless and intuitive content consumption journey whilst providing publishers with the new monetisation channel.

At that time user content consumption journeys were convoluted as users navigated from one publisher homepage environment to another to find the content they were interested in.

This user journey was siloed and inefficient and often involved reverting to search engines in order to their content consumption journey.  

Users were spending too much of their time online trying to discover interesting content, and so therein lay the very problem that CDN’s came about to solve.


The secret sauce, and key differentiator for both Outbrain and Taboola are their algorithms. They draw data from 5 key pillars:

  1. Contextual – “this content is related to that content”
  2. Personal – based on users’ cookie and tracked across the web to serve content based on a profile (what content they like to consume and when).
  3. Behavioral – User groups. Look-a-like audience.
  4. Social – Trending content via the major social channels
  5. Popularity – What pieces of content are being clicked on most every day and promote it.

They combine over 50 algorithms that sit across these 5 pillars to serve up the most relevant and interesting piece of content to user in a given session. The weighting of this formula changes depending on the user, and the supply and demand balance.


Add a Comment

Your email address will not be published. Required fields are marked *