Advanced Television

IBM Watson Media, IRIS.TV AI video recommendations

September 13, 2018

By Colin Mann

IBM Watson Media and IRIS.TV, a cloud-based personalised video programming system have launched Video Recommendations, a new AI content personalisation engine designed to help publishers boost viewer engagement through highly relevant video programming recommendations. By combining viewer data with a deep understanding of specific elements within video, publishers are better able to predict and deliver captivating content— increasing consumption growth, building audience loyalty, and driving advertising revenue.

The pair note that video content is being produced rapidly, and although digital advertising spend is projected to reach $130 billion by 2021, publishers still struggle to monetise their content. With the AI-powered Video Recommendations engine, IBM Watson Media surfaces contextually relevant content to viewers, which helps enhance engagement and profitability. The new offering extracts rich metadata from video content and combines it with consumer viewing patterns to provide better suggestions that boost session duration, decrease bounce rates and drive advertising revenue. IBM Watson Media clients remain the sole owner of their data, empowering them to maintain a competitive edge in the industry and earn viewer loyalty.

“Media companies today are going through an unprecedented period of disruption and consolidation. TV subscribers are declining, consumers are moving their viewing to digital outlets, and both technology and social media companies are capturing the lion’s share of traditional advertising dollars through new channels,” said Richie Hyden, COO & Co-Founder of IRIS.TV. “To cut through the noise, content providers must be able engage audiences with exceptional content. Our new combined offering with IBM Watson Media extracts value from video to enhance suggestions and drive the bottom line for our clients.”

IBM Watson Media runs audio and visual analysis on a client’s video library to increase and improve metadata. IRIS.TV then utilises the metadata to predict viewing patterns and create a continuous learning system that can understand which videos have the highest probability of being viewed to completion. This means that content owners can better match video programming and brand advertisements to specific viewers, creating a highly personalised experience that will retain audiences across platforms.

“Consumers are constantly inundated with news, challenging publishers to promote targeted content that will keep viewers engaged with their media properties,” said David Mowrey, Head of Product and Business Development for IBM Watson Media. “Our Video Recommendations offering is a vital asset for publishers, that can help them deliver personalised suggestions to viewers, driving engagement and maximising advertising revenue.”

Categories: AI, Articles, Middleware, OTT, Search/Recommendation