Harley Finkelstein

Shopify grew GMV 29% and Revenue 30% in 2025 ("full throttle"). They announced a $2 billion buyback and now power 14% of US e-commerce. The market fears AI will displace traditional e-commerce interfaces. Finkelstein argues the opposite: Shopify built the "Universal Commerce Protocol" to syndicate billions of products to AI agents. By creating "agentic storefronts," Shopify ensures that when an AI (like Gemini or Copilot) buys something, it buys through Shopify's rails. LONG. The company is successfully pivoting from a "website builder" to the "backend of AI commerce," turning a potential existential threat into a distribution multiplier. Failure of AI agents to gain consumer adoption for shopping; continued skepticism from the market regarding AI displacement of SaaS models.

Shopify developed an open standard "Universal Commerce Protocol" with Google and is syndicating product catalogs to "every major AI platform," specifically naming Google (Gemini) and Microsoft (Copilot). For AI agents to be useful in commerce, they need structured, real-time inventory data and a transaction layer. Shopify is providing this "feed" to Big Tech. This partnership validates Google and Microsoft's AI models as viable shopping interfaces, moving them from information retrieval to transactional utility. LONG. These hyperscalers benefit from the immediate integration of 14% of US e-commerce inventory into their AI models without having to build the merchant relationships themselves. Regulatory scrutiny on data sharing; low conversion rates on AI-driven purchases compared to traditional search.

Shopify developed an open standard "Universal Commerce Protocol" with Google and is syndicating product catalogs to "every major AI platform," specifically naming Google (Gemini) and Microsoft (Copilot). For AI agents to be useful in commerce, they need structured, real-time inventory data and a transaction layer. Shopify is providing this "feed" to Big Tech. This partnership validates Google and Microsoft's AI models as viable shopping interfaces, moving them from information retrieval to transactional utility. LONG. These hyperscalers benefit from the immediate integration of 14% of US e-commerce inventory into their AI models without having to build the merchant relationships themselves. Regulatory scrutiny on data sharing; low conversion rates on AI-driven purchases compared to traditional search.