Taboola Uses AI, Deep Learning To Build Personalized Loyalty Segments For Publishers
Taboola Uses AI, Deep Learning To Build Personalized Loyalty Segments For Publishers
Taboola on Monday introduced a set of loyalty features for publishers as part of its personalization and segmentation suite.
The feature, which uses artificial intelligence and deep learning to identify loyal site visitors, is built on a loyalty score. The score is based on understanding the number of times readers visit a site, how long they stay, how many pages they view and the recency of their visits.
Based on the criteria, readers are segmented and receive a loyalty score. For example, if one set of site visitors has an average of 10 page views in one week earlier in the month versus another set that had 10 page views last week, the latter group would get a higher loyalty score based on their recency.
One large unnamed U.S. news publication has already achieved an 18% desktop and 10% mobile RPM increase by adjusting their site content shown to flyby readers.
The score also factors in whether the site visitor originated from a search engine, social site, or typed the publisher’s URL directly into the browser. Readers who come directly to publisher site by typing in their URL, rather than searching in—Google, Bing, or Facebook—tend to have a higher loyalty toward the publisher, compared with people coming from social media.
Publishers using Taboola to deliver the ideal reader experience for each reader segment aims to promote increased readership and unlock new opportunities.
This new way for publisher to segment readers based on their loyalty toward the site also lets publishers identify readers who likely won’t return.
Segments include flyby readers, groups that visit less frequently and consume the least amount of pages; loyal readers, groups that demonstrate site affinity and consume the most amount of page views; and casual readers, those who fall between.
Publishers can personalize the reader experience based on these loyalty levels, to optimize user engagement and monetization, based on what makes most sense for each segment.
This may include adjusting the ratio of publishers’ own content versus advertising content, enabling loyal readers to engage with more publisher content; or surfacing abridged articles that feature a “Read More” button, targeted only to flyby users.
The feature also provides the ability to provide special offers only to loyal readers to subscribe to content or sign up to newsletters. The same webpage real estate can then be utilized for other purposes for other less loyal reader segments.
(29)