Influenster Feed Personalization & Retention Platform

Impact

• Redesigned Influenster’s home feed from a manually curated experience into a personalized, behavior-driven feed focused on discovery and engagement.
• Defined a modular feed system supporting reviews, photos, Q&A, products, galleries, and articles, with a controlled content hierarchy to balance freshness and relevance.
• Designed and validated a recommendation framework based on recent user interactions, category affinity, and distinct author filtering to prevent repetition.
• Partnered with backend engineering to plan a scalable architecture using category-based caching and featured content buckets, ensuring meaningful feeds even for low-activity users.
• Built and demoed a proof-of-concept using real production data to secure leadership buy-in before full infrastructure investment.
• Launched the feed through controlled A/B testing, tuning content weights (photo-heavy, follower-heavy, influencer-heavy) to optimize engagement.
• Used the new feed as a re-engagement surface for lapsed users, injecting older community content and pairing it with lifecycle messaging.
• Contributed to ~6% retained MAU lift in the first month post-launch and ~20% total MAU growth over ~4 months, with continued gains after rollout.