#364 How to Enable Agentic Commerce with Nell Thomas, VP of Data at Shopify

Agentic commerce, the practice of using AI agents to conduct product discovery and purchase, is moving from theory to reality. Nell Thomas, Shopify’s VP of Data, explains that shoppers can now ask an agent in conversational language—“find a shirt for my three‑year‑old with a large head”—and receive curated options within seconds. This shift addresses the long‑standing pain point of vague searches and complex filter menus, especially for spec‑heavy categories like electronics. By embedding AI into the buying journey, retailers gain a competitive edge, while consumers enjoy faster, more intuitive experiences that bridge the gap between intent and product.

Shopify is building the technical foundation that makes agentic commerce scalable. The company’s agentic storefronts let merchants plug directly into discovery services such as ChatGPT, Microsoft Copilot, and Perplexity without deep engineering effort. Central to this is the Shopify Catalog API, which normalizes millions of SKUs, enriches attributes, and presents a single source of truth that AI platforms can query instantly. Data quality is monitored through continuous measurement of attribute mapping, duplicate‑product identification, latency, and reliability, creating a feedback loop that self‑improves. The Universal Commerce Protocol (UCP) further standardizes how any AI agent communicates checkout rules, discounts, and terms across all participating merchants, simplifying integration for both platforms and sellers.

For merchants, the impact is a reduction in operational friction and a focus on core business activities. Automated metadata extraction from product images and AI‑driven sidekick assistants allow even small‑scale sellers—like jam makers—to maintain rich catalogs without manual JSON edits. Data practitioners must prioritize trustworthy, well‑documented datasets, as AI agents rely on accurate information to generate recommendations. While discovery and pre‑purchase assistance are mature, full‑transaction automation still requires human oversight, keeping brand interaction authentic. As UCP adoption grows, retailers can expect broader agent reach, higher conversion rates, and a smoother path to scaling AI‑enhanced commerce.