Agentic Commerce Is Now Real: The 5-Point Checklist Every Shopify Merchant Needs Right Now

By Steve Merrill | March 29, 2026

March 24, 2026. That's the date agentic commerce stopped being something to watch and became something to act on.

On that day, Shopify activated agentic storefronts for all eligible merchants. Products from millions of stores became discoverable inside ChatGPT, Microsoft Copilot, and Google AI Mode simultaneously. Not a pilot. Not a phased rollout. A full launch.

Most merchants didn't notice. That's actually useful information, it means there's still a window to get positioned before everyone catches up.

Here's the five-point checklist. Run through it today.

What Exactly Changed on March 24, 2026?

Shopify confirmed three things at once. First, all eligible merchants can now sell inside ChatGPT via the Shopify Catalog. Second, Shop Pay is coming to Microsoft Copilot. Third, select brands are live in Google AI Mode through the Universal Commerce Protocol (UCP), an open standard Shopify co-developed with Google.

According to Shopify's official announcement, the company's Agentic Plan is now publicly available, meaning any brand can sign up to start participating in AI-powered commerce channels today. The announcement also confirmed that "commerce is no longer confined to websites or apps", a statement that's been building for months but is now backed by live infrastructure.

This is the open gate. What you do in the next few weeks determines whether you're in or out when the first wave of AI-driven traffic compounds.

Check 1: Is Your Shopify Catalog Connected and Active?

The Shopify Catalog is the central feed. Everything flows through it.

Go to your Shopify admin right now and verify the Catalog is active. If you're not sure where to find it, search "Catalog" in your admin search bar. This feed is what ChatGPT reads when someone searches for your product category. It's what Copilot will use when Shop Pay goes live. It's the foundation everything else is built on.

If the Catalog isn't connected, nothing else on this list matters. Fix that first.

Worth noting: the Catalog is only as useful as the data in it. A connected Catalog with incomplete product data is like having a storefront with empty shelves. The AI can see you exist, it just can't recommend anything.

Check 2: Is Your Product Data Complete Enough to Be Recommended?

This is where most stores fail. Not on the technical connection. On the data quality.

Run through your top 20 products. For each one, verify: complete title (brand + product name + key attribute), detailed description (200+ words, not a bulleted spec dump), multiple high-resolution images (minimum 1000x1000px, white background for product shots), accurate pricing and inventory status, and correct category tagging.

Products that fail two or more of those criteria don't get recommended. The AI doesn't have enough signal to confidently surface them when a shopper asks a specific question.

According to Digital Commerce 360, the Shopify-ChatGPT integration surfaces "the most relevant options from the brands that people love", which means relevance filtering is real. Products with incomplete data get filtered out before they ever appear in a recommendation.

Not great for most stores. Fixable in a few days of focused work.

Check 3: Is Your Schema Markup Live and Error-Free?

Schema markup is how AI agents understand the structure of your product data beyond what's in the Catalog feed.

Three types matter most right now:

Product schema, Confirms product name, price, availability, image, brand, and description are correctly structured.

Review schema, Shows AI agents that your products have social proof, what the aggregate rating is, and how many reviews you have. This is a trust signal.

Organization schema, Confirms your brand identity, website, contact information, and social profiles. It tells AI systems you're a legitimate business, not a drop-ship operation with zero brand identity.

Use Google's Rich Results Test (search.google.com/test/rich-results) to check your pages. If you're getting errors, fix them. If you're missing schema types, add them. This is a one-time task with ongoing compounding benefit.

Check 4: Do You Have an llms.txt File?

An llms.txt file is a plain-text document that lives at yourdomain.com/llms.txt. It tells AI agents what your store is about, what you sell, and how to understand your catalog. Think of it as a cover letter for your store, written for an AI reader.

It should include: a brief description of your brand and what you sell, your main product categories, key differentiators (what makes your store different), policies (shipping times, return policy in plain language), and any important context that helps an AI recommend your products accurately.

Creating one takes about 20 minutes. The effect on AI discoverability is disproportionate to the effort. I've seen stores add an llms.txt file and show up in ChatGPT product recommendations within two weeks when they weren't showing up before.

If you don't have one, create it today. It's 300 words in a text file. Worth doing.

Check 5: Have You Actually Tested Your AI Visibility?

This is the one step most merchants skip. Don't.

Open ChatGPT and search for your product category the way a customer would. "Best [product type] for [use case]." "Top-rated [product] under [price]." "Where can I buy [specific product]?" See what comes up. See what doesn't.

Do the same in Perplexity's shopping mode. Then check Google AI Mode if you have access.

If your products appear: great. Note what's working and double down on it.

If your products don't appear: you've just identified the most important thing to fix in your business right now. Go back through checks 1-4 and figure out which one is failing. Usually it's product data quality (Check 2) or missing schema (Check 3).

The eMarketer Shoptalk 2026 report noted that most AI assistant shoppers, after getting a recommendation, go do their own research before buying. That means showing up in the recommendation is step one. Your site's job is to close the sale after the AI makes the introduction. Both steps have to work.

How Long Will It Take to Complete This Checklist?

If your Catalog is already connected and you have decent product data: maybe 4-6 hours to verify everything, fix small gaps, and add an llms.txt file.

If your product data is incomplete across your catalog: a few days of focused work, prioritizing your top-revenue products first.

If you're starting from scratch with no schema and a thin Catalog: plan for a week of work, but start today. The merchants who moved early on Facebook ads in 2013 are still miles ahead. This window has the same dynamics. I watched this exact pattern play out before. The ones who wait for certainty miss the compounding period entirely.

The checklist is five items. None of them are technically complex. The only reason to delay is not knowing this mattered.

Now you know.