Adobe has launched AI-powered product discovery for large language model (LLM)-based shopping through Adobe Commerce. The capability uses Adobe Catalog Agent to provide structured product information to AI-powered discovery systems.
For more than two decades, ecommerce product discovery has primarily relied on search engines, marketplaces, and on-site search. Conversational AI is adding another channel, with customers increasingly using platforms such as ChatGPT, Microsoft Copilot, Claude, Gemini, and other LLMs to recommend products, compare options, and answer buying-related questions before visiting an online store.
According to Adobe Digital Insights, traffic from AI sources to US retail sites grew 125% year-over-year from April through June 2026. During the November-December 2025 holiday shopping period, AI traffic to US retail sites was up 693% year-over-year.
Adobe Catalog Agent adds structured product information
Adobe Commerce now includes Adobe Catalog Agent, which enriches product detail pages with structured information drawn directly from the Commerce catalog. The information is provided through a machine-readable layer intended for AI crawlers and LLM-powered discovery systems.
The structured product information can include:
- Product names and descriptions
- Attributes and specifications
- Categories and variants
- Compatibility information
- Pricing and availability
- Product relationships
- Use-case information
This gives AI applications additional context to interpret products, connect customer queries with relevant products, and generate recommendations. The capability works behind the existing storefront, so product detail pages, imagery, and the existing buying journey do not need to be changed.
Consistent product information across channels
Catalog Agent can enrich product names, descriptions, and use-case phrases directly within the Commerce product catalog. Making these changes at the source allows the same product information to be used across:
- Storefronts
- Advertising pipelines
- Marketplaces
- AI-powered discovery experiences
As product information changes, the updated catalog information can be used across these different surfaces, giving LLM-powered systems more context about each product and helping maintain consistent product information across channels.
Product discovery for LLM surfaces
Traditional product discovery includes:
- Search engine optimization (SEO)
- Product feed optimization
- Marketplace visibility
- On-site merchandising
AI-powered commerce adds another discovery layer in which applications can interpret shopper intent and use structured product information rather than relying only on keyword matching.
Product discovery for LLM surfaces in Adobe Commerce exposes structured commerce data that AI applications can use when responding to conversational shopping queries. Examples include:
- “Show me lightweight trail running shoes suitable for marathon training.”
- “Compare these two laptops for video editing.”
- “Find accessories compatible with this camera.”
The approach supports product discovery through conversational AI alongside existing ecommerce discovery methods.
Agentic AI and Adobe Commerce
Adobe positions Catalog Agent as part of its agentic AI capabilities for commerce. The agent provides a product knowledge layer that allows AI applications to work with structured catalog information when products are surfaced, understood, and recommended through AI-powered shopping experiences.
Adobe said its AI commerce roadmap will continue to include catalog intelligence, enrichment, governance, and discovery capabilities as LLM-powered shopping develops.
Business and customer use cases
Adobe identifies several outcomes from making structured product information available to AI-powered shopping systems. These include making products easier for AI applications to find, understand, and recommend, while providing AI assistants with more product context when answering customer questions.
The structured information can also support conversational responses that help customers evaluate products and make purchasing decisions. It complements traditional SEO by providing product information for intent-based discovery through AI applications.
As AI agents become more involved in the buying journey, product catalogs can also serve as a source of structured information for those systems.
Support for technical teams
Adobe Commerce provides native capabilities for AI agents to work with structured catalog data, reducing the need for businesses to build custom integrations or manually expose product information to AI applications.
Catalog Agent can use existing Commerce services and data, including:
- Catalog information
- Product attributes
- Inventory
- Pricing
- Product relationships
This enables AI applications to retrieve current product information instead of relying on outdated or inferred data.
The capability is designed to work with existing Commerce investments without requiring organizations to redesign their catalog architecture or replace their existing commerce systems.
Availability
Adobe Catalog Agent for product discovery on LLM surfaces is available as a native capability in Adobe Commerce on Cloud (PaaS). It is not a separate third-party platform or integration.
Adobe says existing Adobe Commerce customers can use the capability without introducing a separate solution or duplicating product data.