Retailers running on Commerce's ecommerce platforms generated more than $538 billion in online sales in 2025, according to Digital Commerce 360's Top 2000 Database — and the company now wants that base of merchandise data made legible to AI agents, not just human shoppers.

Commerce, the parent company of BigCommerce, Feedonomics and Makeswift, has released two AI tools — Feedonomics Enrichment and BigCommerce Catalog Enrichment — designed to make merchant product catalogs ready for what the company calls agentic commerce: the preparation of product data for discovery and transactions by AI agents rather than conventional search.

The stakes are structural. Consumers and B2B buyers increasingly ask AI products to compare products, answer questions and, more recently, complete purchases. Merchants whose catalog data is incomplete or inconsistent risk invisibility in those channels. For distributors and hybrid B2B-B2C operators, the question is whether enriched, agent-readable product data becomes a prerequisite for maintaining channel position as answer engines displace traditional search.

"AI agents can only answer questions about your products as well as your data allows," said Sharon Gee, senior vice president of product for AI at Commerce. "Whether you're a marketer driving AEO or a product leader building shopping agents, better data is the foundation for better agentic experiences. Feedonomics Enrichment and BigCommerce Catalog Enrichment give brands on any platform a data pipeline that lifts performance everywhere agents meet shoppers from Google to OpenAI and on their own sites and agents."

What the tools actually produce

Commerce said the enrichment tools "turn the product information a brand already has into more robust, consistent data." Both generate product titles, descriptions, feature bullets, FAQs and SEO metadata. They also produce facts, snippets and question-and-answer fields that generative AI platforms use to interpret what a product is, who it targets and how relevant it is to a user query.

The company said the process creates complete records that give human shoppers and AI agents consistent details across storefronts, marketplaces, answer engines and other channels, and eliminates CSV exports and third-party tools. BigCommerce Catalog Enrichment allows self-serve enhancements; Feedonomics Enrichment offers both self-managed and managed-service models.

The distribution problem is one-to-many. Gee offered the example of a retailer translating a footwear catalog into 15 languages and then feeding those translations to Amazon, eBay, Google, Meta, TikTok and Walmart. Retailers now also must reach answer-engine channels: ChatGPT, Claude, Copilot, Gemini and Perplexity. "It's a one-to-many problem — scale," Gee said.

Commerce disclosed no pricing, adoption figures or transaction volumes for the new tools; the $538 billion figure measures the web sales of Top 2000 retailers using its platforms, not the enrichment products' traction. Both claims — that enriched data lifts performance across agent channels and that the tools remove third-party dependencies — are the company's assertions, not independently measured results.

The B2B angle

The customers coming to Commerce are typically "deep in B2B and in enterprise retail," Gee told Digital Commerce 360. Hybrid businesses selling both B2B and B2C often need flexibility and composability in their stacks, she noted.

"It's fairly trivial to sell things online anymore," Gee said. "That's a well-solved problem over the past two decades. There are many use cases that have not been solved well yet."

One unsolved case: manual purchase orders with thousands of line items. An AI agent can complete that work in a "very short amount of time" where a human would need hours, Gee said. Commerce's Purchase Order Agent lets users drag and drop a PDF, from which the agent automatically builds a cart. "Now those humans could get back to doing human work instead of rote, nasty data entry that is super manual that doesn't actually add more value," she said.

For B2B distributors, the near-term economics are labor arbitrage — hours of data entry compressed into automated cart-building — rather than a proven revenue lift. The longer-term bet is that as buyers delegate product comparison and purchasing to agents, catalog data quality becomes a determinant of which suppliers agents surface and which they skip.

Commerce has not yet reported adoption or performance data for the enrichment tools, so distributors evaluating them this year will be judging asserted capability against measured results.