• Optimization metrics quantify how users engage within a session and act as proxy of satisfaction
  • save
  • long click
  • post
  • follow
  • click to stream
  • impression to click
  • dwell time
  • percentage completion
• Why several metrics
  • Different part/ type of content needs different metrics
  • Different user phases need different metrics
  • Different structure of the site and content need different engagement metrics
• Understanding intents
• Optimizing for the right metric
• Acting on segmentation
  • Genre diversity
  • Landing page quality
• Thinking about diversity
  • Satisfaction
  • Popularity
• Intra-session engagement
  • During the session
  • Easily mislead, especially for a short time
• Inter-session management
  • Across sessions
• We optimize (and monitor) intra-session metrics but those that move the inter-session metrics

Reference

• Jointly Leveraging Intent and Interaction Signals to Predict User Satisfaction with Slate RecommendationJointly Leveraging Intent and Interaction Signals to Predict User Satisfaction with Slate Recommendation
  **
  
  
  
  Reference
  
  R. Mehrotra, M. Lalmas, D. Kenney, T. Lim-Meng, and G. Hashemian, “Jointly Leveraging Intent and Interaction Signals to Predict User Satisfaction with Slate Recommendations,” in Th...
• You Must Have Clicked on this Ad by MistakeYou Must Have Clicked on this Ad by Mistake
  
  Reference
  
  G. Tolomei, M. Lalmas, A. Farahat, and A. Haines, “You Must Have Clicked on this Ad by Mistake! Data-Driven Identification of Accidental Clicks on Mobile Ads with Applications to Adve...
• Metrics, Engagement & Personalization at Spotify

#recommender-system #recommender-system/quality #metrics