By Steve Merrill, Founder of WRKNG Digital | June 26, 2026
Most Shopify stores fail at least 7 of these 10 checks. If AI shopping assistants aren't surfacing your products, something on this Shopify agentic commerce readiness checklist is missing — and it's almost never the thing you'd guess first.
1. Shopify Catalog Is Enabled and Synced
ChatGPT Shopping pulls product data from the Google Shopping Graph, which your store feeds through the Google & YouTube channel in Shopify admin. If that sync is off, paused, or erroring, you're gone before the conversation starts. Log into your Shopify admin, open the Google & YouTube app, and confirm your catalog sync status shows active with no feed errors.
2. All Products Have Complete Product Schema with Offer, Price, and Availability
Shopify adds basic Product schema by default. Basic is not enough. AI shopping agents parse Offer schema to verify price and real-time availability before recommending anything. According to Google's structured data documentation, missing or incomplete Offer markup is one of the most common reasons products get excluded from rich results. Check your schema output at search.google.com/test/rich-results.
3. FAQ Schema Added to Product Pages
Nine out of 10 Shopify stores skip this entirely. FAQ schema on product pages puts the questions shoppers ask AI assistants directly into structured data. "Does this run large?" "Can I return it after 30 days?" "Is it available in wide width?" Those answers need to be machine-readable, not buried in body copy. Add three to five question-and-answer pairs per product using FAQ schema and you give AI agents a direct citation source.
4. llms.txt File Exists at /llms.txt
The llms.txt standard tells AI crawlers what your store is, what you sell, and how you want to be described. It takes about 20 minutes to create. Almost no Shopify stores have one. AI agents use it to build context about your brand when forming a recommendation, so if you're not providing that context, they're filling in the blanks themselves.
5. Return Policy Is Machine-Readable
Most stores have a return policy page. It's written as prose. That's not enough for an AI agent evaluating whether to recommend your product. Return policy data — returnPolicyCountry, returnFees, returnMethod — needs to live inside your Offer or Product schema. Without it, the agent either ignores you or makes assumptions about your policy. Neither is good for a sale.
6. Product Descriptions Answer Questions, Not Just Describe Features
"Crafted from full-grain leather" is a feature. "Fits a standard 15-inch laptop with room for a charger" is an answer. AI recommends products that answer shopper questions, not products that read like a spec sheet. Rewrite your opening paragraph to answer what this is, who it's for, and when to use it. Features can stay in bullet points. Answers belong at the top.
7. Google Merchant Center Is Connected and Feed Is Approved
A rejected or suspended Merchant Center feed blocks your products from Google Shopping, Google AI Overviews, and the broader Shopping Graph. ChatGPT Shopping and Perplexity both trace their product data back to infrastructure that runs through Google's ecosystem. One feed error upstream cascades into missing visibility everywhere downstream. Check your Merchant Center dashboard for any active disapprovals or account-level warnings.
8. Product Images Meet AI Visual Search Standards
Google's product image guidelines require a clean background, the full product in frame, and no watermarks. For AI visual search, multiple angles matter more than you think. A single front-facing image leaves AI agents with incomplete data when a shopper asks what the back looks like or runs a visual search from a screenshot. Minimum three angles per product: front, back, and lifestyle or detail shot.
9. Breadcrumb Schema Is Implemented Sitewide
Breadcrumb schema maps how your catalog is organized: products to collections to categories. Without it, AI agents have a harder time surfacing the right product in the right context. Many Shopify themes, especially third-party ones, drop breadcrumb markup entirely. Run a schema test on your collection and product pages to confirm breadcrumb structured data is present and correctly nested.
10. Brand Consistency in Product Titles Across All AI Platforms
Your product title in Shopify, in your Merchant Center feed, and in your meta tags should be identical. AI systems cross-reference titles from multiple data sources to verify product identity. "Blue Leather Wallet" in Shopify and "Men's Leather Wallet - Blue" in your feed creates ambiguity. Products flagged as ambiguous get excluded from results or surfaced with low confidence scores. Pick one title format and apply it everywhere.
How We Built This List
This checklist comes from auditing live Shopify stores with our AI readiness scoring tool, which evaluates 47 data points across schema, feeds, content, and crawlability. The 10 items above are the ones that appear as failures in more than 70% of stores we've assessed.
FAQ
How do I check if my Shopify catalog is synced to Google?
Open your Shopify admin, go to Sales Channels, and click on Google & YouTube. The Overview tab shows your catalog sync status. Any errors appear there with a description of the issue.
Do I need a developer to add FAQ schema to product pages?
Not always. Several Shopify apps add FAQ schema without code changes. If your theme supports metafields, you can also add structured FAQ data manually with a JSON-LD snippet in your product template.
What is llms.txt and is it required for AI shopping visibility?
It's a plain-text file at yourdomain.com/llms.txt that describes your site to AI crawlers, similar to robots.txt but for language models. It's not required by any platform, but it gives AI agents accurate context about your brand and product catalog when no other structured signal is present.
Will fixing these items guarantee my products appear in ChatGPT Shopping?
No. ChatGPT Shopping and other AI platforms make their own recommendation decisions. But these items are the structural requirements they check first. Missing any of them is a hard disqualifier. Getting them right puts you in the pool — and most of your competitors aren't in the pool yet.
How long does it take to fix all 10 items?
For a store with clean data and a developer available, most items are 1-2 days of work. The harder ones are product description rewrites (item 6) and return policy schema (item 5), which require content and schema work together. Start with items 1, 2, and 7 first — those have the biggest immediate impact on feed eligibility.
Sources
- Shopify — Structured Data and JSON-LD
- Google — Product Structured Data Guidelines
- Google — Merchant Center Product Feed Specifications
- OpenAI — Introducing ChatGPT Shopping
- llmstxt.org — The llms.txt Standard
We audit Shopify stores against all 10 of these items (and 37 more) using our AI readiness scoring tool. If you want to know exactly where your store stands, get your free AI commerce audit at WRKNG Digital.
