Why video streaming platforms are obsessed with knowing you better

Platform executives say people who visit a service even once a year or for marquee titles leave clues about themselves, such as IP address and language preferences that can be used to train the algorithm to surface personalised content.

“Driving engagement and improving watch time on the platform is every platform’s goal and priority," said Phanimohan Kalagara, chief technology officer, Gracenote, the content data business unit of market research firm Nielsen. 

“In an era where content abundance is skyrocketing—with over 582,000 shows and movies now available across leading SVoD (subscription video-on-demand) platforms—personalization has become a critical differentiator for viewer retention and subscription renewal," he said.

The difficulty in navigating a vast content landscape is directly impacting user satisfaction and churn. Studies show that 48% of subscribers have cancelled a service because they couldn’t find something to watch, Kalagara said.

While streaming platforms have attempted user personalization and customised recommendations for a while now, these initiatives were mostly driven by foreign players. The trend is catching up with homegrown platforms, albeit with limited budgets and regional language offerings, according to entertainment industry experts. 

Kalagara said metadata solutions such as those provided by Gracenote, including video descriptors and personalized images, help streaming platforms serve their users better with recommendations that stand out for each user persona. 

In a six-month pilot with a streaming service covering a set of 1,000 titles, Gracenote’s Personalized Imagery solution delivered a 11.2% increase in hours watched per visitor and a 7.7% increase in titles watched per visitor, Kalagara said. 

“This proves how visually personalized recommendations drive tangible lifts in engagement, boosting both platform stickiness and subscription value, ultimately helping streaming services retain viewers in a highly competitive landscape," he said.

Also read | Why aggregation could be a game changer for niche OTTs

The tools of personalization

Personalization is not solely about which programs to show to a user but also where the program or carousel should appear on the page for a user as well as what metadata or imagery to use for each program. 

The popularity of a program is another signal that platforms often use to showcase and stitch together the home page for a user. Personalization often also includes how a program is merchandized and presented to users so they are compelled to hit play on the screen.

“OTT (over-the-top or streaming) platforms in India use viewing history, watch time, device type, and regional language preferences to personalize recommendations," said Russhabh R. Thakkar, founder and chief executive of Frodoh, an ad tech company. “Personalization is directly linked to user retention. If viewers don’t find relevant content within the first few scrolls, drop-off rates increase sharply."

However, heavy personalization can suppress content discovery, especially for smaller or regional titles, Thakkar added. 

Platforms counter this by algorithmically inserting promoted or editorial picks. On the backend, personalization data helps decide what genres to invest in, which languages to prioritize, and how to structure the user interface for different user cohorts.

To be sure, streaming services balance personalization with big-ticket content launches, combining personalized recommendations with trending or popular content. 

“Platforms also suggest content based on user behaviour, like genre or theme. They create dedicated sections for new releases or trending content, blending personalized recommendations with editorial curation," said Rajat Agrawal, chief operating officer and director of Ultra Media & Entertainment Group, which owns OTT platform Ultra Jhakaas. 

Also read | Real-time revolution: Live entertainment takes centre stage on streaming

In the big league

At the inaugural World Audio Visual & Entertainment Summit (WAVES) in Mumbai this month, Bharath Ram, chief product officer, JioStar, said the objective of personalization is to retain customers who come to a streaming service for a big event, such as the Indian Premier League cricket tournament, for the rest of the year. 

JioStar, a joint venture between Reliance Industries Ltd, Disney, and Viacom18, is doubling down on personalization using information such as quality and the premiumness of devices. Real-time language tools add another layer of personalization, Ram had said. 

At the same event, Monika Shergill, vice-president of content at Netflix India, referred to the increasing stress on interactivity via different shows. For example, an episode of Bear Grylls’ Adventure in the Wild that had actor Ranveer Singh as a guest allowed viewers to choose the plot and character arcs at several points, thus getting involved in shaping the narrative.

Entertainment industry experts say content creators, OTT platforms, and marketers now have greater access to data, enabling them to deliver precise search results and personalized playlists to drive up consumption, content discovery, and subscriber retention. 

With third-party data integration and advanced audience profiling, personalization can be elevated by recommending new content based on a user’s online behaviour, such as shopping and social media activities.

“Personalization is crucial to enhancing user experience. When an OTT platform understands and caters to an individual, they (viewers) are more satisfied and likely to spend more time on the platform or renew their subscription," said Devyani Ozarde, managing director and lead–media and entertainment, Accenture in India. “Advanced AI and generative AI further enhance content discovery, as each user views content selection through a unique lens."

Also read | OTT platforms pivot: Short, sweet shows slowly replacing long, dark dramas