In early 2026, Google introduced the Universal Commerce Protocol (UCP) -- a standardized interface through which AI agents can retrieve product information, compare options, and execute purchases on behalf of users. What sounds like a technical footnote is actually the beginning of a fundamental shift in e-commerce. Going forward, it will not just be humans shopping online, but increasingly AI agents acting on their behalf.
For online retailers and businesses with digital products, this raises an urgent question: is your shop ready for a world where machines make the buying decisions?
What Is Agentic Commerce?
Agentic Commerce describes a model where AI agents independently execute purchasing processes. The user defines criteria ("Find me a running shoe under 150 euros suitable for overpronation, delivered by Friday"), and the agent handles research, comparison, selection, and ordering.
This differs fundamentally from traditional online shopping:
- Traditional: Human searches, compares, selects, orders
- Assisted (today): Human asks AI for recommendations, decides and orders themselves
- Agentic (2026+): AI agent researches, compares, decides, and orders autonomously within defined parameters
The difference lies in the degree of autonomy. An Agentic Commerce system does not require human confirmation for every individual step -- it acts independently within a defined framework.
Google Universal Commerce Protocol: The Technical Foundation
UCP is Google's answer to how AI agents should communicate with online shops. It defines a standardized data format through which agents can query product data, check availability, compare prices, and initiate orders.
Core components of UCP:
- Product Data Feed 2.0: Extended product data in Google Merchant Center with new attributes
- Agent Query Interface: Standardized API for AI agents to query product information
- Trust Verification: Mechanisms for verifying merchants and products
- Transaction Protocol: Secure ordering process between agent and shop
New Merchant Center Attributes
Google has extended Merchant Center with attributes specifically relevant for AI agents:
- FAQ Answers: Structured answers to common product questions. Agents use these to directly answer user queries
- Accessories: Accessories and compatible products. Enables cross-selling through agents
- Alternatives: Comparable products from your own range. Gives the agent fallback options
- Use Cases: Concrete usage scenarios for the product. Helps agents match user needs to products
- Comparison Attributes: Standardized comparison features for price comparisons across shops
Practical Example: An AI agent is looking for a coffee machine for an office with 20 employees. Via UCP, it queries Merchant Center and receives not just price and availability, but also FAQ answers ("How many cups per hour?"), accessories (water filter, milk frother), and alternatives from the catalog.
Structured Data: The Language of AI Agents
When AI agents make purchasing decisions, they need machine-readable information. Structured data (Schema.org markup) is the bridge between your shop and the agents.
Essential schema types for Agentic Commerce:
- Product: Price, availability, ratings, specifications
- Offer: Current offers, shipping options, return policies
- AggregateRating: Aggregated customer reviews
- FAQ: Frequently asked questions and answers about the product
- HowTo: Instructions for use or installation
- BreadcrumbList: Category structure for navigation
Important: Agents process structured data more systematically than human visitors. Incomplete or inconsistent data causes your product to drop out of the agent comparison -- even if it is objectively the best offer.
First-Mover Advantage: Why Act Now?
Most online shops are not yet prepared for Agentic Commerce. This is simultaneously a threat and an opportunity:
Threat: Those who do not provide machine-readable product data will be ignored by AI agents. Within two to three years, 15-25% of all e-commerce transactions could run through agents.
Opportunity: Early adopters now have the chance to establish their product data as a reference. Agents that have positive experiences with a shop (fast data delivery, correct information, smooth transactions) prioritize that shop for future queries.
Concrete Steps for First-Mover Advantage
- Clean up Merchant Center feed: Populate all new attributes (FAQ Answers, Accessories, Alternatives)
- Expand structured data: Implement Product, Offer, and FAQ schema on all product pages
- Prepare API interface: REST API for real-time queries of availability and price
- Improve product data quality: Unique product descriptions, complete specifications, high-resolution images with alt tags
- Set up monitoring: Track which requests come from agents (User-Agent analysis)
Impact Across Different Industries
Agentic Commerce will not affect all industries simultaneously or uniformly:
High Impact (2026-2027)
- Electronics and hardware: Standardized products, easily comparable
- Office supplies and consumables: Recurring orders, clear specifications
- Grocery delivery: Shopping-list-based ordering
Medium Impact (2027-2028)
- Fashion and clothing: Agents need sizing recommendations and style preferences
- Furniture and decor: Space compatibility and design preferences are complex
- Cosmetics and personal care: Skin type matching and ingredient preferences
Later Impact (2028+)
- Luxury goods: Emotional purchase decision, consultation is central
- Custom manufacturing: Individual configuration requires human interaction
- Consultation-intensive products: Complex B2B solutions
Optimizing Product Data for AI Agents
The quality of your product data determines whether agents recommend your products. Here are the most important optimizations:
Product Descriptions
- Fact-based: "Battery life: 14 hours under normal use" instead of "incredibly long battery life"
- Comparable: Use standardized units and formats
- Complete: Cover all relevant specifications, not just highlights
- Current: Maintain prices and availability in real time
FAQ Structure
Agents use FAQ data as their primary information source. Structure your FAQs according to this schema:
- Pre-purchase: Compatibility, comparison with alternatives, use cases
- Purchase process: Delivery times, payment options, warranty
- Post-purchase: Installation, maintenance, support contact
Media Content
- Images: Multiple perspectives, isolated on white background, with descriptive alt tags
- Videos: Product demos with subtitles (for AI transcription)
- 3D Models: Where possible, for immersive agent experiences
Conclusion: Agentic Commerce as a Strategic Priority
Agentic Commerce is no longer a futuristic concept but a development that has already begun. With Google's Universal Commerce Protocol, a concrete technical standard exists that businesses can prepare for now.
The three most important steps for 2026:
- Audit product data for Agentic Commerce readiness: Is your data machine-readable, complete, and current?
- Populate new Merchant Center attributes: FAQ Answers, Accessories, and Alternatives are the low-hanging fruit
- Expand structured data: Product, Offer, and FAQ schema as a minimum
Companies that position themselves now secure an advantage that grows exponentially as AI agent adoption increases. The question is not whether Agentic Commerce is coming, but whether you will be ready when it does.
