Live shopping’s explosive success in China highlights a transformative ecommerce channel that collapses discovery and purchase into seconds. Whatnot’s recent acquisition of AI startup Shaped aims to solve discovery challenges in North America by using machine learning to deliver personalized, real-time stream recommendations as inventory and buyer interests shift during broadcasts.
- China’s live shopping sales reached $900B in 2025, dwarfing other markets.
- US live shopping adoption lags; discovery is the key bottleneck.
- Whatnot acquires Shaped to boost real-time personalized recommendations.
Market signal
Live shopping has emerged as a powerful ecommerce format by integrating live video demonstrations, interactive comments, and direct purchase capabilities within a single platform. China’s market dominance with nearly $900 billion in sales for live shopping in 2025 indicates the enormous potential of this channel. However, adoption varies widely by region, with the US market still in early stages where most consumers have not engaged in social media shopping.
This disparity underscores a critical technology challenge: the need for sophisticated, real-time discovery mechanisms that can dynamically match buyers to relevant live streams. Unlike static ecommerce catalogs, live shopping inventories fluctuate rapidly, and user preferences may evolve within a session. These changing data flows require AI-powered systems capable of instant personalization to convert viewers into buyers.
Operator impact
Whatnot’s acquisition of AI startup Shaped exemplifies how live commerce platforms are evolving to address real-time recommendation and discovery challenges. Shaped’s machine learning infrastructure, combined with Whatnot’s extensive live video and user interaction data, aims to deliver faster, more personalized stream suggestions based on both historical purchases and real-time behavioral cues.
This integration enables Whatnot to manage the complexity of an expanding product set, which recently grew by 45 new categories within half a year, and to mitigate the risk of losing shopper engagement when inventory pulses or auctions close rapidly. Enhanced AI-driven discovery will help reduce churn and improve customer lifetime value by ensuring viewers find compelling products at just the right moment.
What to watch next
The broader US and global live shopping landscape will be shaped by platforms’ ability to harness AI and machine learning to overcome discovery friction. Industry observers should track how effectively Whatnot’s AI research group applies real-time recommendation technologies to scale engagement and transactions beyond niche audiences.
Additionally, growth in AI-mediated shopping traffic, which rose dramatically by 393% year-over-year in early 2026, indicates rising consumer receptivity to AI as a shopping aid. Companies that can quickly adapt to session-driven preference shifts and increase personalization will be well positioned as live shopping looks to transition from a frontier channel to mainstream retail.