Entertainment and Media - Content Personalization

Entertainment platforms face the challenge of keeping users engaged in order to maximize their time on the platform and retain them as customers. Platforms such as Netflix and Spotify have shown how great content recommendations lead to a great user experience on their platforms. Netflix reported that 75% of the content watched on the platform comes from some sort of recommendation, while Spotify’s recommendation algorithm facilitates 16 million artist discoveries per month.

 

Enabling such an experience requires a large amount of insight into user profiles, user behavior, and catalog items. Acquiring those insights can be challenging for new platforms with limited data. This is commonly known as “the cold start problem”.

Generating significant lift in conversions and an improved user experience by delivering personalized recommendations and messaging to users of an online platform.

Recommender systems have been around for a long time. Popular implementations normally come in one of, or a combination of, two approaches:

 

1.                 Collaborative filtering, where similar aspects of users and their behavior are used to recommend items; and

 

2.                 Content filtering, where item characteristics are used to suggest similar items.

 

Both approaches involve building a prediction model using machine learning techniques.

 

Jumbo Interactive is an Australian reseller of lottery games. By implementing Amplitude Recommend, Jumbo Interactive were able to create personalized recommendations and messages based on users’ on-site behaviors encouraging them to take further actions. Using this approach, Jumbo Interactive achieved a 158% lift in conversions on one checkout page in two months. Not only were key business metrics improved, but the overall experience was judged to be more fun and engaging by stakeholders.