Challenge: Declining user retention
Since launching a year and a half ago, the Dating Platform has expanded its user base thanks to investment in user acquisition. Yet, they faced a challenge: user retention began to dip as the number of profiles climbed up.
Discovery: Misguided algorithm
Our analysis revealed a core issue—the existing user feed algorithm. Initially made for a smaller number of users, it failed to scale effectively, favouring a small group of users and neglecting the majority. This reduced match rates and user engagement.
Solution: Real-time recommender system
To address this, we spent two months developing a real-time, tailor-made recommender system. This new system consisted of advanced machine-learning models built on the Platform's data. It was designed to adapt dynamically to user preferences, ensuring an engaging experience for all.
Results: Improved UX and revenue growth
The new system transformed the platform. Users started seeing profiles that match their interests, increasing satisfaction and engagement. The system helped retain both existing and new users, ultimately driving up revenue by 30% and improving the Platform's position in the market.