AI Agents Are Already Shopping. Here's How Your Store Can Join Them.

Your customers are already asking AI assistants what to buy. They are typing questions like "best running shoes for flat feet" into ChatGPT, browsing product recommendations in Google's AI Mode, and letting Perplexity compare options before they ever visit a store.

When those AI agents respond, they pull from product catalogs, structured data, reviews, pricing, and availability. They build a recommendation. They link the shopper to a product page. And increasingly, they complete the purchase right inside the conversation.

The brands showing up in those recommendations are earning a new kind of customer: one who arrives informed, engaged, and ready to buy. The AI has already done the research. The shopper just needs to confirm.

Three open protocols are making this new channel possible. They define how AI agents discover products, read catalogs, and complete purchases on behalf of real customers. Here is what each one does, why the landscape around them is shifting faster than most vendors are telling you, and what it means for your store.

Three protocols, one connected commerce layer

MCP: The data layer

Model Context Protocol gives AI agents a direct line to your product catalog. Anthropic open-sourced MCP in November 2024, and in December 2025 donated it to the Agentic AI Foundation, a directed fund under the Linux Foundation co-founded by Anthropic, Block, and OpenAI, with AWS, Bloomberg, Cloudflare, Google, and Microsoft as additional platinum backers.

In practical terms: you connect once, and every MCP-compatible agent can access your product information in real time. No custom integrations per platform. No rebuilding every time a new AI shopping surface launches. Your catalog stays current across all of them.

ACP: The checkout layer OpenAI and Stripe built for ChatGPT

Agentic Commerce Protocol turns AI conversations into completed purchases. Co-created by OpenAI and Stripe and released as an open standard under Apache 2.0, ACP defines the checkout flow: product discovery, selection, checkout initiation, and order confirmation, secured by Stripe's Shared Payment Tokens.

Etsy is the clearest live example: its sellers have been processing orders that originate entirely inside ChatGPT. Stripe's Agentic Commerce Suite has also brought in a wave of early adopters onboarding to agentic checkout, including URBN (Anthropologie, Free People, Urban Outfitters), Coach, Kate Spade, Revolve, and Ashley Furniture, alongside platform partners like Wix, WooCommerce, BigCommerce, and Squarespace.

UCP: The discovery and checkout layer for Google

Universal Commerce Protocol makes your store visible to Google's AI shopping surfaces. It was co-developed by Google with five launch partners, Shopify, Etsy, Wayfair, Target, and Walmart, and endorsed by more than 20 additional partners including Visa, Mastercard, Stripe, American Express, and Best Buy. UCP lets any merchant publish a JSON profile at /.well-known/ucp declaring what they sell, how they accept payment, and what APIs they support.

When a shopper asks Google's AI Mode or Gemini for a product recommendation, UCP is how the agent finds you. Shopify has built native support directly into its platform through a feature called Agentic Storefronts, so most Shopify merchants inherit UCP compatibility rather than building it from scratch.

The landscape is consolidating, and that's the real story

Here's what most explainers on this topic are missing: this isn't three protocols peacefully dividing up the territory. It's a live standards contest, and it's already producing a clear signal.

In April 2026, Amazon, Meta, Microsoft, Salesforce, and Stripe all joined the UCP Tech Council, the governance body that steers UCP's technical direction. That's ten organizations now co-governing UCP, including Stripe, the company that co-created the rival ACP standard eight months earlier. When a protocol's own co-author starts hedging into a competing standard's governance, that tells you where the industry-wide bet is landing.

That doesn't mean ACP is finished. Etsy is live on both ACP and UCP today, and several major retailers have deliberately implemented both rather than picking a side, using ACP for ChatGPT and UCP for Google's surfaces. For most merchants, the practical takeaway isn't "pick the winner." It's "make sure your data and infrastructure are ready for whichever surface your customer happens to be using," because the two protocols are converging in adoption patterns even as they compete at the standards level.

One more data point worth knowing: Amazon hasn't joined either open protocol, and is currently in litigation over whether AI agents can even access its site to shop on a customer's behalf. Every day that standoff continues is a day non-Amazon brands with clean, agent-readable data have a structural opening that didn't exist eighteen months ago.

Quality data is the real competitive advantage

AI agents recommend products based on the richness and quality of data available to them. Shipping speed, return policies, product reviews, inventory status, user preferences: all of it factors into which products get surfaced and recommended.

This is good news for brands that invest in their catalog and customer experience. Unlike traditional search, where ranking often comes down to ad spend or keyword optimization, agentic commerce rewards substance: clean product data, strong reviews, transparent policies, and fast fulfillment. Brands that have already built a reputation for quality have a natural head start.

A parallel trust infrastructure is reinforcing this. Visa's Trusted Agent Protocol establishes an authentication framework for legitimate agent purchases. Stripe's Shared Payment Tokens provide tokenized credentials scoped to specific sellers. Together, they help ensure agents can transact securely on behalf of customers, which builds confidence across the entire ecosystem.

Not sure where your store stands on any of this? A free AI Commerce Readiness scan checks your site against the checkout, discovery, and structured data requirements behind all three protocols, in under a minute. Scan your store now →

How to get your store ready

The good news is that getting started is straightforward. Here are the steps that matter most.

Scan your store. Before you invest in implementation, see where you stand. A readiness scan can map your store's current UCP compliance, ACP readiness, structured data quality, bot accessibility, and product page machine readability in under a minute.

Strengthen your product data. Ensure every product has complete schema markup, accurate inventory status, clear pricing, and machine-readable attributes. This is the foundation that every protocol builds on.

Publish your UCP profile. A JSON file at /.well-known/ucp declares your store's capabilities to Google's AI surfaces. Shopify merchants get this through the platform's native Agentic Storefronts feature. WooCommerce, Magento, and BigCommerce stores can implement it with a plugin or a few lines of configuration.

Activate ACP. Connect Stripe's Shared Payment Tokens to enable purchases through ChatGPT's checkout experience. Shopify and Etsy merchants have built-in integrations that make this especially smooth; if you're on a different processor, Stripe's Delegated Payments spec is the fallback path.

Confirm bot accessibility. Make sure your product name, price, and stock status render in the initial HTML response, and that your robots.txt isn't inadvertently blocking GPTBot, Google-Extended, ClaudeBot, or the other agent crawlers. AI agents that can read your content can recommend your products.

Keep iterating. These protocols added new capabilities twice in their first five months, and the governance landscape shifted again in April. Treat AI readiness as an ongoing practice, similar to SEO, and it will continue to pay dividends.

See where your store stands

We built a free scanner that checks your store across 80+ AI readiness signals covering both Google and OpenAI ecosystems. Results come back in under 60 seconds, and the scoring model is refined continuously as the underlying protocols evolve, because a static checklist doesn't hold up in a market that changes this fast.

Our Solvers have been building ecommerce platforms for 16 years across Shopify, Magento, WooCommerce, BigCommerce, and custom stacks. We can help you turn your scan results into a clear implementation plan.

Two ways to start:

1. Commerce Stack Assessment: A full diagnostic of your underlying commerce platform: architecture, integrations, migration health, and technical debt, across Shopify, BigCommerce, Magento, WooCommerce, or custom builds. Best if you're not sure your foundation can support agentic commerce readiness yet, or if you've been through a platform migration and want confirmation nothing broke along the way. [Request your Commerce Stack Assessment →]

2. Agent Readiness Assessment: A focused scan against the specific checkout, discovery, and structured data requirements behind MCP, ACP, and UCP: where you stand today, what's blocking AI agents from recommending and transacting with your store, and a prioritized fix list. Best if your platform foundation is solid and you want to know exactly where you stand on agentic commerce specifically. [Run your Agent Readiness Assessment →]

Ready to Solve Forward®? Let's talk.