Agentic commerce
Agentic commerce is shopping in which an AI agent searches, compares and can buy products for a person, using structured product data and open protocols.
What is Agentic commerce?
Agentic commerce is ecommerce in which an AI agent does part of the shopping for a person or a business. The agent searches catalogs, compares products on their attributes, answers questions and, when the user allows it, completes the purchase. The Agentic Commerce Protocol, released by OpenAI and Stripe in September 2025, describes this as programmatic commerce between buyers, AI agents and businesses. The merchant stays the seller of record: the agent reads the merchant's product feed, starts a checkout through the merchant's systems and pays with a payment token the user has approved.
Why Agentic commerce matters for e-commerce
Agents choose products from data, not from page design. An assistant comparing three blenders reads price, availability, wattage, capacity, GTIN, return policy and reviews from a feed or structured markup, so products with missing or inconsistent attributes are hard to recommend. The rails are being built now: OpenAI publishes a product feed specification for ChatGPT, Google introduced the Universal Commerce Protocol in January 2026 with Shopify, Etsy, Wayfair, Target and Walmart, Google's Agent Payments Protocol covers agent-led payments, and Visa and Mastercard run their own agent payment programs. For merchants, complete and accurate product data quality becomes a condition for being visible in AI shopping.
Free tool
Generate Product schema for one product
Fill in the name, price, brand and identifiers, then copy a schema.org Product JSON-LD block for your product page.
Examples of Agentic commerce
- 1A shopper asks ChatGPT for a quiet dishwasher under $800. The assistant reads merchants' product feeds, compares noise levels, prices and availability, and links to the products that match.
- 2Google announced native checkout in AI Mode in Search and in Gemini, built on the Universal Commerce Protocol, so a buyer can complete an order with the merchant inside the conversation.
- 3An agent calls Shopify's Catalog MCP servers to search products across merchants or in one store and gets back variants with availability and a checkout link.
- 4A procurement agent reorders consumables for a business when stock runs low, within a spending limit the company has set for it.
How WISEPIM Helps
- WISEPIM fills missing attributes and specifications with AI enrichment, so agents can compare your products on facts instead of marketing copy.
- Quality Guard checks GTINs and required fields before a product is published, the identifiers agent feeds ask for.
- WISEPIM exports the same structured product data to your webshop, marketplaces and feeds, so every channel describes a product the same way.
- See how WISEPIM makes a catalog ready for agentic commerce.
Common mistakes with Agentic commerce
- Treating agentic commerce as a copywriting job. Agents compare typed attributes, so a long description without dimensions, materials or compatibility does not help them.
- Inventing GTINs or MPNs to complete a feed. OpenAI's feed specification asks merchants to leave the GTIN out when none is assigned and not to make up an MPN.
- Refreshing price and availability once a week. OpenAI's feed specification asks for updates whenever price or availability changes.
- Leaving return and shipping details out of the feed. They are fields in OpenAI's specification, and Google requires return policy details for UCP checkout.
Tips for Agentic commerce
- Map your catalog to the fields agents require. In OpenAI's feed format the required fields are item ID, title, description, URL, brand, seller name, image URL, availability and price with currency.
- Add a GTIN wherever one is assigned and check that the check digit is valid.
- Group variants under a shared group ID so an agent can show sizes and colors as one product.
- Publish return policy, shipping costs and a support contact in machine-readable form, not only on a web page.
- Update price and availability as soon as they change, and check feed errors in each platform's merchant tools.
Trends around Agentic commerce
- Open protocols define the flow: ACP from OpenAI and Stripe (Apache 2.0), UCP from Google and retail partners, and AP2 for agent payments, announced in September 2025.
- Card networks prepare for agent payments: Visa Intelligent Commerce and Mastercard Agent Pay, both announced in April 2025, use tokenized credentials and limits set by the shopper.
- The Model Context Protocol (MCP) is becoming the interface agents use to reach catalogs, as with Shopify's Catalog MCP servers.
- Checkout models are still moving: Digital Commerce 360 reported in March 2026 that OpenAI is moving Instant Checkout into apps and focusing ChatGPT on product discovery.
Tools for Agentic commerce
- WISEPIM: AI enrichment, translation and quality checks that make product data complete enough for AI agents to compare.
- OpenAI commerce documentation: the product feed specification merchants use to appear in ChatGPT.
- Google Merchant Center: product data and UCP checkout eligibility for AI Mode in Search and Gemini.
- Shopify Catalog: MCP servers that let agents search Shopify products.
- Stripe: the first payment provider in ACP, with its Shared Payment Token.
Related Terms
Also Known As
Frequently Asked Questions
Still have questions?
Can't find the answer you're looking for? Please get in touch with our team.
Contact Support