By Steve Merrill, Founder of WRKNG Digital | June 11, 2026
AI shopping agents evaluate nine core signals before completing a purchase on your Shopify store: product schema, return and shipping policies, inventory status, merchant identity, payment trust markers, review data, buy link structure, and privacy transparency. If any of these are missing or unreadable, the agent stops — and your sale doesn't happen.
Most stores have none of these configured correctly. That's not speculation. We've audited hundreds of Shopify stores, and the gap between what AI agents need and what stores actually publish is significant. Here's what you're up against.
1. Schema.org Product Markup with Offers
This is the foundation. Google's Product structured data spec defines the fields AI agents look for first: name, description, SKU, brand, price, currency, and availability. Without a valid Offer block inside your Product schema, an AI agent can't confirm the item is purchasable. No Offer block means no transaction. Shopify's default themes don't publish complete Product schema out of the box — you need to add it.
2. Machine-Readable Return Policy
AI agents acting on behalf of buyers won't complete a checkout unless they can verify the return policy. The Schema.org MerchantReturnPolicy type lets you publish your policy in a format agents can parse — return window, conditions, fees, and method. A policy page written in plain HTML prose isn't enough. The agent needs structured data it can read, not a paragraph it has to interpret.
3. Shipping Policy with Delivery Estimates
Delivery timeline is a critical purchase decision for any buyer, human or AI. The OfferShippingDetails schema lets you attach shipping regions, carrier, estimated delivery days, and shipping rate directly to your product Offer. Agents shopping for time-sensitive purchases will deprioritize, or skip entirely, stores where delivery data isn't machine-readable. Publish it explicitly.
4. Real-Time Inventory and Availability Status
The availability property in your Product schema tells an agent whether an item is in stock. The valid values are defined by Schema.org: InStock, OutOfStock, PreOrder, and others. An agent won't add a product to cart if it can't confirm availability from structured data. Stale or missing availability data is one of the most common blockers we see in store audits.
5. Verified Merchant Identity Data
AI agents want to know who they're buying from. Google's merchant identity guidelines specify that your Organization or LocalBusiness schema should include a legal name, address, phone, and contact URL. Stores without verified merchant identity data are flagged as higher risk. The agent either skips you or routes the purchase to a more identifiable competitor.
6. Trusted Payment Method Signals
Agents completing purchases need to know what payment methods you accept. Shopify's checkout supports this natively through Shopify Payments and the Storefront API, but your schema needs to reflect it. Publishing acceptedPaymentMethod in your Offer block, with recognized payment types like Visa, Mastercard, PayPal, or Shop Pay, tells the agent the checkout is compatible with the buyer's stored credentials. Missing this field adds friction.
7. Aggregate Review and Rating Schema
Social proof isn't just for humans. AI agents use aggregate rating data to rank products when multiple options are available. The AggregateRating type in your Product schema publishes your average rating and review count in a format agents can compare. A product with a 4.7-star rating from 312 reviews will consistently outrank a product with no rating data in AI-generated recommendations.
8. Clear Add-to-Cart and Buy Link Structure
When an AI agent is ready to execute a purchase, it needs a clean, predictable URL structure to initiate checkout. Shopify's native cart permalink format, /cart/add?id=[variant_id]&quantity=1, is the standard agents expect. If your store uses custom JavaScript-only cart interactions with no fallback URL, the agent can't complete the action. Keep your buy links crawlable and direct.
9. Privacy Policy and Data Handling Transparency
This one surprises most store owners. AI agents operating as proxies for buyers are increasingly subject to policy-level checks before completing transactions. A published, linked privacy policy, ideally marked up with your Organization schema, signals that your store handles customer data responsibly. The Google Merchant Center policy requirements provide a baseline. No privacy policy is a trust blocker, full stop.
How We Chose This List
This list is based on direct analysis of what AI shopping agents, including ChatGPT Shopping, Google's AI Overviews, and early agentic commerce deployments, require to progress through a purchase flow. We cross-referenced Schema.org's ecommerce spec, Google's merchant documentation, and Shopify's Storefront API documentation against real audit data from live stores. If a signal wasn't required to complete a transaction, it didn't make the list.
FAQ
Q: Does Shopify automatically add the Product schema AI agents need?
No. Shopify's default themes include basic Product schema, but the Offer block is often incomplete, missing availability, shipping details, or return policy. You need to extend it manually or use an app that publishes complete structured data.
Q: What happens if an AI agent can't find my return policy in structured data?
It skips your store. AI agents completing purchases on behalf of buyers treat missing return policy data as a red flag. They route to a competitor with a readable policy rather than interpret a plain-text page.
Q: How do I check if my Shopify store has the right Product schema?
Run your product URL through Google's Rich Results Test or the Rich Results Test tool. It shows exactly which fields are present, which are missing, and which have errors. Most stores fail on Offer completeness.
Q: Do AI agents actually complete purchases right now, or is this future-proofing?
Both. ChatGPT Shopping is live, Perplexity is running purchase flows in beta, and Google's agentic commerce is rolling out through Gemini. The stores getting optimized now are the ones that will compound while everyone else scrambles to catch up. I've seen this movie before.
Q: Is structured data enough, or does the checkout experience itself matter?
Both matter. Structured data gets you discovered and evaluated. A clean, reliable checkout, fast load, no redirect loops, predictable cart URLs, gets you the transaction. Schema without a working checkout is just good marketing for a broken store.
Want to know exactly which of these signals your store is missing? Get your free AI commerce audit →
