February 8, 2017 | Written by: David Kulczar
Categorized: Big Data | Cognitive
Share this post:
The world is consuming an amazing amount of streaming video content.
From silly internet videos to binge watching hours of the latest trending show, the average person’s appetite for streaming video content is tremendous. By 2021, the research firm MarketsandMarkets expects the market to reach $70 billion. To compete in this increasingly crowded space, providers must find a way to differentiate themselves to keep subscribers engaged and actively paying subscriber fees.
There’s a simple method for reducing churn while engaging and pleasing viewers: providers must understand, at a deep level, what individual subscribers want to watch and point them directly to that content. This highly personalized approach to serving up streaming video content becomes possible when providers add machine-learning technologies such as IBM Watson to their streaming services.
Stop subscriber attrition with in-platform cognitive capabilities
IBM Watson can find patterns in the way people interact with video content, from the selections they make, to how often they rewind or pause, to which videos they have abandoned midstream or watched repeatedly. By identifying and analyzing commonalities between the types of programming a viewer enjoys, Watson can suggest options the viewer might not have even considered.
Today, a viewer’s search for video content typically only involves basic metadata, such as the title, genre, and so on. However, Watson can amass advanced metadata about what happens inside streaming videos. It can index and catalog at a much deeper level. This includes the ability to index spoken word, visual imaging, tone and much more.
These capabilities enable subscribers to interact with content in entirely new ways. In the future, a viewer could say, “Show me a movie to help me sleep,” or “Show me a move where people overcome difficult challenges,” and the library could bring up videos that help a subscriber’s mood or outlook in a new way.
A smarter way to plan a video content strategy
Watson collects viewer data from video platforms over time and combines that data from other sources such as social channels, third-party reviews, global trends, and other content including geospatial and real-time weather information. As that happens, providers can compile data-rich profiles of individual subscribers and proactively predict targeted and highly relevant content recommendations.
Watson’s capabilities can also help providers identify users who are likely to drop off their platforms so they can take steps to prevent that from happening. For example, if customers who enjoy watching romantic comedies are particularly prone to churn, the provider can examine its current offerings and decide if it should license more movies. Or it might find that it has a wide range of similar content that would appeal to this segment of subscribers it could recommend instead.
The opportunities that this technology will uncover in improving customer experience and recommendation success on a video streaming platform are limitless. In a future blog, I will explore how cognitive systems can help with areas such as content acquisition and creation, marketing strategies, advertising intelligence, and general business decision making.
Learn more about IBM Cloud Video.