Etsy Recommender Model Rewire

Enhancing Discovery Through Intent and Emotion

Problem

Etsy’s recommender system currently centers on historical behavior and popularity—but falls short in key areas:

• Detecting evolving user intent (e.g., self vs. gift, browsing vs. goal-driven)
• Capturing emotional context (season, sentiment, mood, aesthetic)
• Delivering serendipitous but relevant discovery
• Avoiding repetition fatigue from overly narrow personalization

Why This Matters

This is about reimagining recommender systems to go beyond behavioral echo chambers. When we layer soft signals, context-aware UX, and adaptive intelligence, we can transform browsing into an emotionally intelligent, trust-building journey — one that resonates with fulfilling, real user intent and moment-to-moment context.

Targeted User

• Gift seekers shopping with emotional or recipient-driven intent
• Casual browsers unsure of what they’re looking for
• Returning users whose intent and interest shift over time

Key Friction Points

• Feels static and repetitive over time—same items shown repeatedly