Under the Hood

Traditional recommendation technologies that rely solely on relevance work poorly because it’s not just about getting site visitors to click on a link. It’s about what happens after they click. It’s about making people stick around longer, viewing more content, sharing links, and responding to calls-to-action.

  • Laser-Accurate Recommendations

    The Taboola algorithm-based recommendation engine surfaces the content that is most likely to keep visitors engaged. It excels where others fail by going beyond relevance to look at the following factors:

    • Similarity

      what content other visitors also found interesting

    • Personal

      what content visitors have already viewed

    • Contextual

      what content is relevant to the content on the page being visited

  • Video and Text

    Taboola started from recommending video, a format far more difficult to analyze than text, then added recommendations of textual content, offering the best-performing content distribution platform available.

  • Optimized Widget. Scalable Infrastructure.

    Taboola’s personalized content is served within a customizable on-page widget that is highly optimized to deliver unmatched clickthrough rates and post-click engagement metrics. The widget leverages Taboola’s highly scalable back-end infrastructure that delivers billions of recommendations to hundreds of millions visitors monthly.