We've spent the past two weeks covering the consumer-facing side of agentic commerce — Ulta Beauty letting Gemini shop for you, Google's Universal Checkout Protocol, the UCP Tech Council bringing Amazon, Meta, and Microsoft together around a shared standard. But there's been a quieter question lurking behind all of those announcements: how do merchants actually get their products into these AI systems?

On Sunday, Commerce.com (Nasdaq: CMRC) — parent company of BigCommerce and data feed platform Feedonomics — announced an answer. Agentic Catalog Exports (ACE) is a new enterprise service that lets merchants prepare and syndicate product catalogs across the emerging landscape of AI-powered shopping surfaces, including OpenAI/ChatGPT, Google Gemini, Microsoft Copilot, PayPal, Stripe, Perplexity, and Amazon.

It's infrastructure, not flash. But it might be the most consequential announcement in agentic commerce this month.

The Problem ACE Solves

Right now, getting a product catalog "agent-ready" is an engineering project. Each AI shopping surface has its own specifications, data formats, and ingestion protocols. For an enterprise retailer managing tens of thousands of SKUs, building and maintaining individual integrations for OpenAI, Google, Microsoft, and every other emerging channel is expensive and fragile.

"Agentic commerce is quickly shifting from experimentation to real-world application, and merchants need a reliable way to participate," Sharon Gee, senior vice president of product for AI at Commerce, said in the announcement. "We're making it easier for enterprises to prepare their product data for this new environment without having to build and maintain complex, one-off integrations for every destination."

ACE centralizes the data transformation, enrichment, and syndication — turning it from a custom engineering problem into a managed service.

Dell Is Already In

Dell is among the early enterprises using the service, having prepared approximately 7,000 products — laptops, desktops, servers, monitors, and accessories — for AI-driven discovery.

"As AI agents become a more common starting point for product discovery, the quality and structure of product data matter more than ever," Paul Mansour, Dell's global marketing director, said in the release. "Feedonomics helped us optimize and structure our catalog so Dell products are not only more discoverable, but also more accurately and completely represented within ChatGPT."

That last phrase — "more accurately and completely represented" — is the key. In a world where an AI agent is making purchase recommendations, the quality of the structured data behind each product becomes a competitive advantage. Poor product data doesn't just mean a bad listing. It means the AI skips you entirely.

The Shift from Search Feeds to Agent Feeds

Feedonomics has long been in the business of syndicating product data to traditional channels — Google Shopping, Amazon Marketplace, Meta Commerce. ACE represents a deliberate pivot toward what the company calls a shift "from traditional search feeds to structured agent-ready catalog exports."

The distinction matters. A Google Shopping feed is optimized for search algorithms and ad auctions. An agent-ready feed is optimized for conversational AI systems that need to understand product attributes, compare alternatives, and execute transactions — a fundamentally different data structure.

As SalesTechStar reported, early implementations include a mix of crawled and feed-based experiences, with a long-term trajectory toward fully structured, feed-driven commerce interactions across all agentic surfaces.

What This Means for Retailers

ACE is initially available as an enterprise service, with self-service capabilities planned for mid-market merchants later. That means the largest retailers will get first-mover advantage in agentic discoverability — at least through Feedonomics.

For retailers not on BigCommerce, the announcement is still significant as a signal. The infrastructure layer for agentic commerce is being built now, and merchants who wait too long to structure their product data for AI consumption risk becoming invisible in the channels where consumers increasingly start their shopping journeys.

The storefront used to be a physical location. Then it became a website. Then an app. Now it's becoming a structured data feed that an AI agent reads before a human ever sees a product. The retailers who understand that shift — and invest in the plumbing accordingly — will be the ones the algorithms find first.