Following an eight-month testing period, The New York Times has implemented Taboola's EngageRank video recommendation technology on its online video section, NYTimes.com/video.
Founded in 2007 by members of Israel's National Security Agency and the country's intelligence corps, New York-based Taboola's video search offering for Web publishers offers users a personalized list of videos based on past viewing patterns.
Taboola video recommendations are based on mathematical algorithms that combine behavioral and contextual data to analyze users' engagement with the videos they watch and generate real-time predictions for additional videos of interest.
The recommendations appear as a list of thumbnail images in a section titled "Other Videos You May Like," below the video player and at the end of the video.
According to Adam Singolda, founder and CEO of Taboola, the company's recommendation engine enhances user experience, increases time spent on site, boosts loyalty, and drives profitability.
"We have A/B testing built into our culture and product innovations, constantly quantifying the lift our technology provides against different alternative types of recommendations."
Presently, Taboola says it makes recommendations to over 51 million unique users monthly across a network of publishers, including CNN, Kiplinger, Slate, Demand Media, and Revision3.
During the 8-month trial period with The New York Times, Taboola offered video recommendations below the video player and at the end of the video labeled "Other Videos You May Like."
The likelihood for users to watch another video was found to be higher than baseline recommendations, according to internal data cited by Taboola.
Video publishers using Taboola's ViDiscovery technology also report seeing significant increases in video views, and in the percentage of viewers who watched an entire video.
The privately held company has recieved roughly $6 million in funding from Evergreen Ventures, and a small group of Israeli and American individual investors.