What’s in this article
- What the Agent Payments Protocol (AP2) is — The open standard that lets AI agents buy things for users.
- The AP2 Readiness Checklist — My 5-point checklist you can copy and use to make a Shopify store agent-buyable.
- How I make a store agent-buyable — The exact steps to configure Shopify and Google Merchant Center.
- What this changes for agencies — How agentic commerce creates a new, high-value service offering.
- FAQ — Answers on security, ROI, and how this differs from traditional e-commerce.
🚀 Plug this into Claude Code or Claude Desktop
This post explains the strategy. The downloadable spec gives you a Claude Code prompt to audit any Shopify store against my 5-point checklist, with step-by-step instructions for getting compliant.
Want hands-on help applying this to your client’s store? That’s what the Talk-to-Build community is for.
I’m going to give you my 5-point checklist to make any Shopify store “agent-buyable.” This means AI agents can browse, add to cart, and check out without a human touching a thing. You can copy it and use it on a client’s site today.
Google’s Universal Cart is rolling out in the US, letting AI like Gemini buy products directly from search results. But most stores are invisible to it. They don’t speak the right protocol. This checklist is the fix that gets you seen and, more importantly, gets you paid by AI.
What is the Agent Payments Protocol (AP2)?
The Agent Payments Protocol (AP2) is an open-source standard that allows AI agents to securely complete purchases on behalf of a user. It creates a verifiable chain of consent using three signed mandates—for user intent, the shopping cart’s contents, and the final payment. This protocol is what enables features like Google’s Universal Cart to transact with e-commerce stores programmatically and securely.
The AP2 Readiness Checklist I Use — Copy It Right Now
This is the exact 5-point checklist I use to audit and configure a Shopify store for agentic commerce readiness. It ensures the store is technically compliant with protocols like AP2 and discoverable by platforms like Google’s Universal Cart. You can use this as a new service offering for your e-commerce clients starting this week.
# 5-Point Agent-Buyable Checklist (for Shopify)
[ ] 1. Enable Shopify Agentic Storefronts:
In Shopify Admin, go to `Settings` > `Channels` and add the `Agentic Storefronts` channel. This is the master switch for AI visibility.
[ ] 2. Verify Product JSON-LD Schema:
Use Google's Rich Results Test on a product page. Ensure schema includes `name`, `price`, `availability`, `sku`, and `image`.
[ ] 3. Review AP2 Manifest Requirements:
Familiarize yourself with the spec at `github.com/google-agentic-commerce/AP2`. This informs agents what payment methods your store supports.
[ ] 4. Submit to Google Merchant Center AI Surfaces:
In Google Merchant Center, navigate to `Growth` > `Manage programs` and ensure you are opted into "AI Mode product surfaces."
[ ] 5. Add GA4 Custom Event for Tracking:
Create a custom event named `ai_assisted_purchase` to track revenue from this channel separately from your other organic traffic.
I run through these five steps for every e-commerce client. It’s the difference between a store that AI can only look at and a store that AI can actually buy from.
Your Store
│
├─ 1. Agentic Storefronts (Enabled) ──┐
├─ 2. Valid JSON-LD Schema ───────────┤
├─ 3. AP2 Manifest ───────────────────┤──> [AI AGENT] ---> Purchase
├─ 4. Google Merchant Center (AI On) ─┤
└─ 5. GA4 Tracking ───────────────────┘
Here’s Exactly How I Make a Shopify Store Agent-Buyable
My process involves five specific configuration steps inside Shopify and Google Merchant Center. I start by enabling the core agentic channel in Shopify, then I verify the product data is structured correctly for AI to read. Finally, I connect it to Google’s ecosystem and set up tracking to measure the results. It’s a technical but straightforward workflow.
- Enable Shopify Agentic Storefronts. In the Shopify Admin, I navigate to `Settings` > `Channels`. I search for and add the “Agentic Storefronts” channel. This is the foundational step that signals to platforms like Google that the store is open for AI business.
- Verify Product Schema. I take a few key product URLs and run them through Google’s Rich Results Test. I’m looking for valid Product schema with all the critical fields filled: `name`, `price`, `availability`, `sku`, and a high-quality `image`. If it’s missing, I use a Shopify app or edit the theme’s liquid files to add the correct JSON-LD.
- Opt-in to AI Surfaces in Google Merchant Center. I log into the client’s Google Merchant Center account, go to `Growth` > `Manage programs`, and find the program for AI-powered surfaces. I make sure it’s active. This explicitly tells Google to include the store’s products in Universal Cart and other agentic experiences.
- Set Up GA4 Conversion Tracking. To prove this works, I need data. In Google Analytics 4, I create a new custom event called `ai_assisted_purchase`. I then work to ensure that transactions coming through the AP2 protocol trigger this specific event, which allows me to isolate and report on agent-driven revenue.
- Review the AP2 Manifest Spec. This is less of an action and more of a preparation step. I review the official AP2 spec on GitHub. This helps me understand what’s coming and advise clients on which payment gateways (like Shopify Payments, PayPal) are most ready for agentic transactions.
What This Changes for E-commerce Agency Work
This shift changes an agency’s role from just building beautiful stores to building machine-readable, transactable endpoints for AI. The core deliverables now include technical schema, protocol compliance, and AI channel management. This creates a new, high-value service category focused on what I call “agentic readiness,” a key part of my agentic commerce playbook.
| Dimension | Old Way (Human-First E-commerce) | New Way (Agent-First E-commerce) |
|---|---|---|
| Customer Discovery | SEO keywords, social ads, human browsing. | AI agent recommendations, conversational search. |
| Checkout Process | Manual cart building and form-filling by user. | Autonomous, one-click purchase by an AI agent. |
| Technical Stack | Focus on visual design and UX for humans. | Focus on structured data (JSON-LD) and API compliance (AP2). |
| Agency Deliverable | A visually complete Shopify theme. | An “agent-buyable” store with verified data and protocol support. |
The job is no longer just to make it look good for a person. The new job is to make it work perfectly for a robot. Agencies that can spec and ship this work have a massive advantage in 2026.
My $0.02 — How I’d Roll This Out for a Client
I’d structure this as a three-day sprint for any existing client Shopify store. The goal is to move fast from audit to implementation to measurement. This isn’t a six-month project; it’s a focused technical upgrade that can be sold as a standalone package to prove immediate value and open the door to a larger retainer.
Day 1 — The “Agent-Ready” Audit. I’d take my 5-point checklist and run a paid audit on a client’s existing Shopify store. I’d deliver a simple report: red, yellow, or green for each of the five points, with a clear explanation of what’s missing and the risk of inaction (i.e., being invisible to AI-driven sales).
Day 2 — Implementation Sprint. Once approved, I’d execute the checklist. This means getting into their Shopify admin and Google Merchant Center to flip the right switches. If their schema is broken, I’d fix it. This is a day of focused, technical work that makes the store compliant.
Day 3 — Test, Measure, and Educate. I’d use test queries in Gemini’s latest version to see if the client’s products surface in agentic results. I’d walk the client through the new `ai_assisted_purchase` event in their GA4 dashboard and explain how we’ll track ROI from this new channel together. This final step turns a technical project into a business outcome.
FAQ
What is Google’s Universal Cart?
Google’s Universal Cart is an AI-powered feature in Search and Gemini that allows users to add items from any participating store to a single cart. An AI agent then handles the checkout process across these different stores on the user’s behalf, creating a single, unified shopping experience.
Do I need a developer to implement AP2?
For a full custom implementation, yes. However, platforms like Shopify are building no-code or low-code integrations via features like Agentic Storefronts. The checklist in this post is designed to be completed by a technically savvy agency owner or designer, without needing to write custom code.
Is AP2 only for Google?
No. While Google is a primary driver, AP2 is an open-source protocol with backing from major players like Mastercard, PayPal, Salesforce, and Shopify. It’s designed to be a universal standard for any company building AI shopping agents, making it a future-proof investment.
How is this different from Google Shopping?
Google Shopping is a marketplace for browsing and comparing products, which then sends you to the merchant’s site to buy. Universal Cart lets an AI agent *buy the product for you* directly from the search results, often without you ever needing to visit the store’s website.
Is this secure for customers?
Yes. AP2 is built on W3C Verifiable Credentials, which is a modern, secure standard. It creates a cryptographic chain of signed “mandates” for the user’s intent, the cart’s contents, and the final payment. This is far more secure than older methods like screen-scraping credentials.
What’s the biggest risk of ignoring this?
Your products become invisible to the fastest-growing channel for e-commerce discovery and purchase. As more users rely on AI assistants to shop, non-compliant stores will be completely bypassed. AI agents will simply buy from your competitors whose stores are configured correctly.
Which platforms support this today?
Shopify is the clear leader for independent merchants with its Agentic Storefronts initiative. Major enterprise retailers like Nike, Target, Sephora, and Walmart were launch partners, indicating that broad support across other major e-commerce platforms is either live or coming very soon.
How do I measure the ROI of becoming “agent-buyable”?
By setting up a dedicated conversion event in Google Analytics 4, like the `ai_assisted_purchase` event from the checklist. This allows you to build reports that directly compare revenue generated by AI agents against traffic from traditional organic search, paid ads, or social media.
Want help applying this?
Four ways to go deeper:
- Build with Builders. Join the Talk-to-Build community to learn to build AI-native websites, cinematic AI video, and agent-driven workflows you can sell.
- 1-on-1 working session. Book a screen-share with me — bring a real problem, leave with a working piece of it.
- Done-for-you. MK-Way builds AEO-ready websites, apps, and AI agent workflows.
- Quick question. DM me on Instagram or LinkedIn. I read every message.
Part of the AI Pulse series. If you commented “AGENT-COMMERCE” on one of my videos — this is the breakdown.
Last updated: 2026-07-13.