AI Can't Find What It's Looking For on Most Product Pages
Source: Beet
Mirakl's research reveals a gap in e-commerce infrastructure: most product pages lack the structured data and metadata that LLMs need to process information reliably. This creates friction for retailers betting on AI-driven shopping agents—whether through their own platforms or third-party marketplaces—since agents that can't read product pages accurately can't complete transactions. The constraint is not AI capability but data standardization: retailers who invest in LLM-ready product pages now will capture early wins in agent commerce, while others risk invisibility to the next wave of shopping behavior.