The NineHertz collaborated with a music streaming platform from Canada, which serves the South Asian diaspora market in North America, Australia, and the UK. The site’s primary offering is libraries of songs from Bollywood, Telugu, Punjabi, Tamil, and South Asian musicians, and it’s a very niche culture.
The platform already had over 28,000 active listeners, thanks to word-of-mouth publicity and community-driven social media presence. However, the client noticed that despite love from their target audience, average session lengths were very short. The users were signing in but not staying. At the same time, a big fraction of premium subscribers have been churning for the past 90 days.
The company was facing a high churn rate among premium subscribers, with more than 14-17% churn in the first 90 days of signing up. The identified cause is the repeated content and the lack of new artists on the platform. Another challenge was low session depth as average listening sessions were only 18 minutes, where they should be not less than 35-40 minutes to be considered a healthy benchmark.
The platform was also unable to serve language preferences and niche regional content. The users have to manually search for their favorite artists or music, as there is no recommendation system. The personalization touch was also missing, which could have analyzed the user’s taste to suggest the content accordingly.
NineHertz built a core product architecture that offers personalization at the center of the listening experience by analyzing user data and enabling recommendation engines.
Our team built a real-time event tracking infrastructure that could capture data for each user, like skip point, play duration, listening frequency, search-to-play conversion, and playlist completion rate.
A smart recommendation system has been implemented in the new solution that analyzes the user’s listening taste using their historical search and music patterns to suggest content from similar genres or artists.
The recommendation triggers directly into the listening experience that enables the platform to surface matched suggestions after every three tracks, resembling in mood, language pattern, and tempo.
A dedicated discovery surface was built in the platform that surfaces underrepresented independent artists with a similar music genre and suggests them to the listeners, enabling meaningful distribution of content on the platform.
Within 6 months of the implementation of the new solution, the platform witnessed significant improvement in subscriber retention, engagement depth, and overall listening volume.
Weekly active users as a proportion of the registered base increased from 34% to 55% within 12 weeks across all three markets- Canada, Australia, and the UK.
The new recommendation engine now enables the user to easily find the music of their taste and genre without manual search that increased average session length from 18 to 31 minutes.
After the implementation, the 90-day premium churn rate decreased from 15.5% to 6.8% as personalized content discovery encouraged the user to explore the platform for a longer time.
Users now don’t listen to those limited popular artists but explore the streams from independent and emerging South Asian artists. The music listening from these artists grew by 120%.
