ShemarooMe Boosts Video Watch Time by 80% Using Amazon Personalize

Shemaroo Entertainment Ltd. (Shemaroo) is a media company in India specializing in creating, aggregating, and distributing a wide range of content, including Bollywood blockbusters, classic movies, regional films, and originals. The company covers various genres such as fiction, non-fiction, comedy, devotional content, and children's entertainment. Shemaroo’s over-the-top (OTT) platform, ShemarooMe, streams content to millions of users and provides access to more than 4,300 movies, television shows, music videos, and channels.

In early 2023, Shemaroo aimed to enhance user experience and engagement on its ShemarooMe mobile app through personalization. The business introduced Clips, a dedicated section on the app showcasing 30-second video clips contributed by the community from movies, TV shows, and live TV.  

Abhishek Joshi, head of ShemarooMe, says, “The digital video ecosystem has always struggled with the 'problem of plenty'. Content discoverability is one of the major hurdles in engagement and retention of viewers. This made us go back to the drawing board to find a solution to this problem. That is how Clips was incubated. It provides an additional entry point for our viewers to discover content recommendations based on their viewing preferences, which are a function of a host of data points that are accumulated across the spectrum.”

Kinnari Dave, deputy vice president for product and innovation at Shemaroo, says, “To meet our goals, we needed to add layers of personalization and recommendations to keep users engaged on the platform.

Since its inception, the ShemarooMe OTT platform has been running on Amazon Web Services (AWS), utilizing AWS services like Amazon Simple Storage Service (Amazon S3), Amazon CloudFront, AWS Lambda, and Amazon Elastic Compute Cloud (Amazon EC2). For its Clips platform, ShemarooMe opted for Amazon Personalize, integrating advanced artificial intelligence (AI) capabilities into the short video platform to achieve a higher level of personalization and enhance business value.

Amazon Personalize helps developers build and deploy curated recommendations and intelligent user segmentation using machine learning (ML) technology. The Amazon Personalize algorithm is continuously being refined and complemented with rich data sets for the business, and the flexibility of the solution allows ShemarooMe to incorporate its own user-level consumption into the recommendation engine.

As a result, ShemarooMe can now provide personalized recommendations to its diverse user base that consumes content across various languages and regions. When users view popular short videos on ShemarooMe, these interactions are ingested into Amazon Personalize in real-time, to further fine tune the recommendations based on data points collected. The recommendation engine suggests genres like action, comedy, or romance to align with individual user preferences. Moreover, the engine identifies patterns after a user views a certain number of videos, and suggests content based on those patterns.  

Since implementing Amazon Personalize, ShemarooMe experienced a significant boost in video viewer uplift, which measures overall minutes of watch time. “By using Amazon Personalize, we increased video watch time by more than 80 percent, and we eventually increased the number of viewers who watch our short videos by 88 percent,” says Kinnari.

As a result of providing personalized content, ShemarooMe anticipates revenue growth through an increase in ShemarooMe subscriptions and increased retention rates. The company is now creating its own promotional short videos for movies and shows. “Traditional movie trailers adhere to industry standards that can't be customized. However, we can personalize our own trailers based on user preferences to boost subscriptions, offering viewers something they might not find in a standard trailer, such as exclusive highlights from the movies that triggers a user’s interest,” explains Abhishek.

Additionally, ShemarooMe is exploring broader utilization and increased adoption of Amazon Personalize to tailor its homepage experience and enhance personalized search functionality on the platform. "Amazon Personalize has proven to be highly effective,” comments Kinnari. “The algorithm and its implementation have contributed to the development of a precise recommendation engine, which has resulted in increased engagement and a better experience for our users.”