• How Recommendation Technology Is Evolving to Meet Customer Expectations

    In the media business, content is king, so content traditionally (and understandably) takes priority over user experience. But priorities are shifting as streaming evolves into a more complex, competitive space where differentiated products can make a big difference to the bottom line.

    To truly personalize discovery, Comcast is investing heavily in improving how its customers search and browse content. And by valuing its personalization tech at $1 billion a year, Netflix firmly established that a truly personalized entertainment platform presents large opportunities for companies trying to hit the moving target of user expectations.

    Such companies are driving those expectations higher, while pay-TV providers play catch-up. Once users see a truly personalized experience, anything less will disappoint them. These pressures put the entertainment world on an exciting path to a predictive and individualized experience for every consumer.

    To stay a step ahead of that curve and meet rising consumer expectations, video discovery products need to do four things: provide context, delight the user, stay flexible, and inspire interaction.

    People like context. They want to know why something is recommended. Introducing a new product is about earning trust. Great recommendations expose people to new titles, which means the recommendations’ quality can be confirmed only after users actually enjoy those titles.

    Before that happens, they have to take a leap of faith. To establish that your recommendations are in the ballpark, it helps if users can see some familiar titles they already like. This kind of transparency helps them understand how your system works and increases their trust in your recommendations.

    Surprises delight users. You want to show users something they didn’t expect to see by exposing titles perfectly positioned on the edge of the category, concept, or list they’re considering. Delivering these surprises is both an art and a science, requiring advanced technology and experimentation to refine a recommendation system so that it delivers quality and variety.

    Researchers in the recommendation space often refer to this surprise concept as “serendipity,” and they’ll adjust algorithms to introduce it into their results. Users are bored with the obvious. When you help expand their horizons, they will spread the word.

    Flexibility is key. Users expect recommendations to stay as fresh as their catalogs, and sometimes this means providing unique or personalized discovery features to channel those recommendations.

    For example, so much great content exists, but users grow frustrated when content comes and goes, seemingly at random. Video On Demand services can’t expect users to track catalog changes. However, to avoid frustration, they could provide a way to receive helpful alerts and updates from apps or services that users like or a “New for You” category tailored to their tastes.

    This is just one scenario in an ever-evolving market. Video services need a consistent recommendation system that is ready for anything and flexible enough to power cutting-edge features their users have grown accustomed to experiencing elsewhere in their apps and devices.

    Interactive discovery inspires. User taste profiles and behavior can be as chaotic and diverse as the content universe they are trying to navigate. People like to discover in different ways, and all are valid. Yet product teams often overlook interactive discovery because they assume users are lazy. Emboldening these power-users can lead to better social curation and turn a consumption platform into a primary discovery portal.

    It’s all about balance. A single user might pursue multiple methods of discovery depending on his or her mood, so it’s important to allow for those multiple paths without overwhelming the user. Recommendation features also exist on a spectrum — from purely content-based modules like related titles to purely user-based recommendations that focus on the user’s unique taste profile. So the best personalization platforms will combine content-based recommendations with collaborative filtering that includes the wisdom of the crowd, spanning multiple points on that spectrum at once.

    Video platforms set themselves up for success by finding that balance of providing context, delighting the user, staying flexible, and inspiring interaction. Product road maps may diverge down different paths, but the one constant is that consumer expectations will continue to rise, and the companies without true product competency or a good partner won’t be in position to meet them.