Amazon has made a big business in part because of its recommendations. Customers who bought this also bought that has proven tremendously helpful for book readers, as well as consumers searching out cameras, pet toys and other products on the world’s biggest shopping site. Netflix, similarly, has grown its business because of suggested movies and shows.
That’s why online video technology providers have staked a claim on personalized video recommendations, with companies such as Ooyala, Taboola and Digitalsmiths having skin in the game.
Ooyala said that its recently added content discovery features had boosted both viewing time and revenue for customers. Prior to the product launch last month, the company had tested the tools with existing customers, which includes broadcasters, media companies and TV networks. Ooyala found that the first video abandonment rate after one play - i.e. leaving to go to another site - had dropped to 9% with its tools, from a 24% rate with the technology customers had previously been using to encourage viewers to watch more videos. “The more viewers are engaged with personalized videos tailored to their tastes, the longer they watch, and the more money publishers earn,” Ooyala said in a blog post.
I asked other companies in the business about results they’ve seen with recommendations. Taboola focuses on recommending videos for its customers across their sites including on home pages, article pages, or next to the video player. The click-through on recommendations can range from single digits to as high as 35% depending on the placement, with article pages generating strong lift in terms of the number of clicks for recommended videos on those pages, said Adam Singolda, CEO of Taboola, adding that when viewers convert from a non-video page to watching video they tend to watch at least two videos. “The opportunity to increase video views is primarily by showing video recommendations on the entire site to convert readers to become video viewers,” he said. Video views can often be monetized at higher rates than page views.
Digitalsmiths said with its content discovery tools it can predict with 95% accuracy what a user is going to watch next. The company’s Seamless Discovery tool, launched in November and used by service providers, content providers and consumer electronics makers, leverages the metadata in its database to make better recommendations. Customers have seen a three times increase in video consumption when using it.
“On the user side, we create a profile for each viewer that constantly evolves based on the viewer's explicit ratings of content and implicit ratings derived from viewing behavior,” said Ben Weinberger CEO of Digitalsmiths.