Knowledge Graph–Driven Feedback Loops to Optimize TikTok Shop Video Marketing
summary
Claims a closed-loop system that links creative video features to real sales using Neo4j knowledge graph built from TikTok Shop seller analytics, with frame-by-frame LLM+Vision extraction of hooks, shots, scripts and voiceovers. Tests run at scale on real brand data with paired t-tests and time-adjusted regressions; examples include filtering videos with 5,000+ GMV and a category case showing ~$25k GMV for egg-prep content. Resulting dashboard runs flexible pivots far faster than heavy SQL joins and yields actionable audience–hook–GMV recommendations.
AI-written summary · about
chapters (9)
- 0:00Introduction and speakers
- 1:06TikTok Shop growth and challenges
- 6:05Measuring real success metrics
- 11:47Relational database limitations
- 14:16Why graph databases excel
- 18:31Flexible schema for marketing
- 22:03Graph database query examples
- 25:56Reinforcement learning feedback loop
- 27:22Impact and future automation
related talks
19mKnowledge Graphs & Deep Learning at YouTube
25mUsing feedback loops to tune predictive models in a video ad marketplace
40mKnowledge graphs --enter- the Hype Cycle
38mKnowledge Graphs : Adding a New Dimension to Vector Search and Generative AI
27mData storytelling in the era of generative AI
30mHow Graph Data Science Can Turbocharge Your Knowledge Graph
01h 40mKnowledge Graphs & GraphRAG: Techniques for Building Effective GenAI Applications
24mKeynote: AI Trends in Performance Marketing