Is Recommendation Tech Doing Enough?

Recommendations are driving consumption like never before. Video content providers are beginning to embrace the opportunity of personalization to increase consumption and engagement of their services.

By Luke Gaydon, Head of Marketing, Solutions and Innovation, Accedo and Jerónimo Macanás, CEO, Jump-Data-Driven Video

No wonder when you consider that 75% of content consumed by Netflix users is accredited to service recommendations, and they lead to an increase of 2 hours in TV viewing per week. This is according to some great stats shared at the AWS summit last year.

However, despite this, there remain some challenges with existing recommendation technology. Currently, most recommendation systems operate using explicit information provided by the user about their preferences (for example, by scoring previously watched content) using a technique known as collaborative filtering. Most recommendation systems may also factor in a user’s profile when suggesting appropriate content, and in doing so may use the preference and feedback information provided by similar users. Such user profile information may contain demographic and geographic data, in addition to more dynamic data, such as the user’s web activity (e.g. pages visited, videos watched, activity on social networks).

However, what the majority of these are lacking is AI in-content analysis techniques.

Additionally and even this could seem obvious, if you are not tracking the effectiveness of recommendations, you cannot adapt and improve for maximum performance. All of this means that you will not get a good return on investment for implementing these technologies.

Deep Recommendations

Adding advanced machine learning content tagging  techniques to enhance content metadata to be used in recommendations makes it possible to target those recommendations much better as well as tracking and analysing their impact on your video service.

Advance AI in-content analysis  means you can enhance  content similarity based on the recognition, analysis, and tagging of video frames and audio samples, which identify factors such as:

·      Scene speed (fast, slow, loud, etc.)

·      Colours and luminosity

·      Scene locations (landscapes, city landmarks, indoor/outdoor scenes, etc

·      Predominance of a time of day (morning, afternoon, night

·      Character sentiment and emotion detection (joy, sadness, violence, etc.)

·      Predominant elements (sea, fire, sky, etc.)

Jump Deep Recommender

Jump’s Deep Recommender is a cutting-edge AI recommendation engine, which gives your users recommendations based on a wide range of parameters, including region, device, time of day, day of the week, etc. in order to maximize engagement and consumption. It can be simply and easily integrated with your existing video service through an API or SDK.

Combined with advice from Accedo’s leading video user experience experts, it can transform the video experience, enabling content providers to improve engagement from content recommendations and keep subscribers happy.

Find out more about the combination of Accedo’s expertise and Jump’s Deep Recommendation platform or schedule a time for a demo.