NoticeNotice
Guide Agentic commerce

What is agentic commerce?

Agentic commerce is a model of online shopping in which an AI agent — inside an assistant like ChatGPT, Gemini, or Perplexity, or a platform like Shopify — discovers, compares, and buys products on a shopper's behalf, instead of the shopper browsing a store themselves.

The shopper stops browsing and starts delegating. They tell an assistant what they want, and the agent does the searching, comparing, and — increasingly — the buying:

ChatGPTshopper prompt

What's a healthy soda I should try? Something low-sugar that actually tastes good.

Agent surfaced three resellers before the brand's own store. The brand never saw the query.

For a brand, the customer relationship now runs through a layer you don't own and can't see by default. McKinsey puts the agentic-commerce opportunity at $3–5 trillion by 2030; Shopify reported AI-driven orders growing 13× year over year in Q1 2026, at order values about 30% higher than non-agent orders.

The store stops being where the sale happens. The agent is.

How agentic commerce works

Every agentic purchase moves through four steps — and a brand can win or lose the sale at any of them, usually invisibly:

  1. 01

    Intent

    A shopper asks an assistant a buying question in natural language, not a keyword.

  2. 02

    Retrieval

    The agent queries one or more product catalogs — Shopify's global catalog, merchant feeds, or the open web — for candidates that match the intent.

  3. 03

    Ranking

    It orders those candidates on relevance, structured data, reviews, price and availability, and authority — not on ad spend. Results are organic and unsponsored today.

  4. 04

    Routing

    It presents a short list and, increasingly, completes checkout in the conversation — sending the order to whichever seller holds the winning slot.

Agentic commerce on Shopify

In March 2026, Shopify activated agent commerce across US merchants. Instead of optimizing a storefront for human visitors, you list a product once in Shopify's global catalog and it syndicates to AI surfaces — ChatGPT, Perplexity, Microsoft Copilot, and others — through what Shopify calls Agentic Storefronts. An agent can then recommend and buy your product without a shopper ever loading your website.

The protocols behind agentic commerce

Agentic commerce runs on a new layer of open protocols — the plumbing that lets an assistant read a catalog and complete a purchase. You don't implement them yourself (your platform does), but knowing the landscape helps:

Google

UCP

Universal Commerce Protocol — Google's open standard, co-developed with Shopify, Etsy, Target, and Walmart and endorsed by Mastercard, Visa, and Stripe. Powers agentic buying in Google's AI Mode and Gemini; keeps the brand as merchant of record.

OpenAI + Stripe

ACP

Agentic Commerce Protocol — the standard behind instant checkout in ChatGPT.

Payments

AP2

Agent Payments Protocol — a shared standard for authorizing agent-initiated payments, backed by the major card networks. UCP is built to interoperate with it.

Context

MCP

Model Context Protocol — how an agent reads your live data (inventory, specs, shipping) so it recommends accurately instead of guessing. Shopify's AI toolkit ships native MCP support.

The protocols will keep multiplying and converging. What doesn't change: under every one of them, a reseller or competitor can still win the slot for your product, and the order still lands in your store looking like direct traffic.

Why DTC brands are losing control

Agentic commerce breaks the three things a brand normally relies on — and the data shows how wide the gap already is:

Visibility blind spot84%of brands don't systematically track AI search performance. — McKinsey CMO Survey, Sept 2025
Attribution leak70.6%of AI referrals are misclassified as "direct" traffic in GA4. — Averi, 2026
Channel velocity13× YoYgrowth in Shopify AI-driven orders in Q1 2026, at +30% AOV. — Shopify, Q1 2026

The blind spot and the leak compound: agent-driven revenue lands as an undifferentiated lump misread as direct traffic, so the channel never becomes measurable, optimizable, or budget-eligible — and the slots you lose to resellers stay lost because no one is watching them.

How to take control of the channel

There are three levers, and they only work in order:

  1. 01

    Make your catalog machine-legible

    Complete Product, FAQ, and Organization structured data; an llms.txt file; rich product attributes; and review volume. This is answer engine optimization applied to commerce.

  2. 02

    Measure where you stand

    Track where your SKUs rank across ChatGPT, Gemini, and Perplexity — and how you get recommended in ChatGPT in the first place. You can't improve a channel you can't see.

  3. 03

    Audit every slot, attribute every order

    Find out who holds each slot in Shopify's catalog, the one structural fix that takes it back, and map agent-driven orders to the surface and prompt that caused them.

Notice is built for the second and third levers — per-SKU rank tracking across the AI surfaces, a slot-ownership audit of Shopify's global catalog, and agent-order attribution — turning a channel most DTC brands can't see into one they can measure and control.

See who's winning your slots — and where your SKUs rank across ChatGPT, Gemini, and Perplexity — in a free instant audit.

Frequently asked questions

What is agentic commerce?
Agentic commerce is a model of online shopping in which an AI agent — inside an assistant like ChatGPT, Gemini, or Perplexity, or a platform like Shopify — discovers, compares, and buys products on a shopper's behalf. The shopper states an intent in natural language; the agent reads product catalogs, ranks the options it finds, and routes the order to whichever seller wins the slot.
What is the difference between agentic commerce and traditional ecommerce?
In traditional ecommerce a shopper finds your storefront through search or ads and browses it directly. In agentic commerce an AI agent sits between the shopper and the store: it reads catalog data, decides which products to surface, and often completes checkout in the conversation. Discovery shifts from your site to the agent's ranking, and the order can be fulfilled by you, a retailer, or a reseller depending on who the agent picks.
What is Shopify's agent catalog?
In March 2026 Shopify activated agent commerce across US merchants. Products are listed once in Shopify's global catalog and syndicated to AI surfaces through Agentic Storefronts, so an agent searching for a product can recommend and buy it without visiting the merchant's website. Any seller carrying a product can appear, which is why a reseller or competitor can win the agent slot for a product you make.
How do I optimize my products for agentic commerce?
Make your catalog machine-legible (complete Product, FAQ, and Organization structured data, an llms.txt file, and rich product attributes), earn reviews and brand mentions across the open web, then monitor where your SKUs rank across ChatGPT, Gemini, and Perplexity and who holds each slot in Shopify's catalog. You cannot improve a channel you cannot see, so measurement comes first.
What is the Universal Commerce Protocol (UCP)?
The Universal Commerce Protocol (UCP) is Google's open standard for agentic commerce, co-developed with Shopify, Etsy, Target, and Walmart and endorsed by payment networks including Mastercard, Visa, and Stripe. It lets AI agents in Google Search's AI Mode and Gemini discover and buy products through one standard, across the full journey from discovery to checkout to post-purchase. The merchant stays the merchant of record and keeps ownership of the customer relationship and data.
What is the difference between UCP and the Agentic Commerce Protocol (ACP)?
They are two emerging standards for the same job — letting AI agents transact — backed by different camps. UCP is Google's, used by AI Mode and Gemini; ACP is OpenAI and Stripe's, used by ChatGPT's instant checkout. Both are built for interoperability (UCP works alongside the Agent Payments Protocol, Agent2Agent, and the Model Context Protocol). For a brand the practical point is the same regardless of protocol: your product still has to win the agent's recommendation, and you still need to see which surface and prompt drove each order.