AgentReady
By Dylan HuntJune 22nd, 2026Updated November 12th, 2026AIAgentic commerceCatalog

We Rebuilt AgentReady Around Shopify's Global Catalog: Here's Why

We Rebuilt AgentReady Around Shopify's Global Catalog: Here's Why

A year ago, the most important thing about your product page was whether a shopper could find it. Today, the shopper might never see it.

When someone asks ChatGPT for "a quiet space heater under $80" or tells Shop to "reorder the coffee I bought in March," the assistant doesn't open twelve tabs and read your storefront. For Shopify products, Shopify says they reach AI channels through Shopify Catalog, a single cross-merchant index of products that Shopify builds, normalizes, and partly infers from every store's data. The shelf moved. It moved off your domain and into a catalog you don't render and can't see.

That shift is the single biggest thing to happen to Shopify merchants this year, and most stores haven't noticed because nothing on their site looks different. So we did something deliberate: we rebuilt a large part of AgentReady around the Global Catalog. Not as a checkbox feature. As the spine.

Here's the thinking, and what we built.

The catalog is the new shelf, and the data is the rent

The catch is in how the catalog gets built. It isn't your storefront copied over. Shopify takes your product data, normalizes it into a shared structure, and infers the fields you left blank, the category and attributes an agent needs to match a query. Shopify's own documentation is candid that inference accuracy "depends on available product data." Translation: if your data is thin, the catalog guesses, and the guess is what an agent reads about you. Guess wrong and you're miscategorized, mispriced, or simply absent when the buyer asks.

So the work is concrete and a little boring: clean, complete, identifier-rich product data. That's the rent you pay to be on the new shelf. We built AgentReady to help you pay it, and to check that Shopify picked it up.

Measure: see where you actually stand

You can't fix what you can't see, and the catalog is, by design, something you can't see from your admin. So the first thing we built is a window into it.

Run our free Catalog Readiness check and it grades the data on your storefront that the catalog is built from: structured product data, identifiers, the product feed, discovery signals. It reads your store, not the catalog. Inside the app, AgentReady goes a step further for a sample of your products and compares what Shopify's catalog returns with what's in your admin. Wrong category. Stale price. A product the catalog never picked up at all. The specific, unglamorous gaps that make an agent misread you.

Scoring is free, with no signup and no card. It's the honest part of the funnel: we'll tell you exactly where you stand before you pay us a cent. For the full picture of how the catalog works under the hood, the Global Catalog guide is the long version.

Fix: one click, in your admin, never behind your back

Knowing your category is wrong doesn't help if fixing it means a spreadsheet and an afternoon. So the fix has to be one motion.

When the comparison flags a gap, AgentReady drafts the fix and applies it to Shopify when you confirm. A clearer title that names what the thing actually is. The right Shopify taxonomy category. Your brand in the Vendor field. A description with the specifics an agent can match on. A nudge to add the barcode that lets the catalog cluster you onto the shared listing instead of stranding you on your own. One field at a time, with one-click undo.

That "when you confirm" is load-bearing. We don't do blind bulk writes, and we don't run a timer that edits your catalog while you sleep. Every change is propose, confirm, apply. You always pull the trigger. It's slower than "fix everything," and we think that's correct: your catalog is your storefront's source of truth, and nobody should rewrite it without you watching.

Check: did the catalog absorb your fix

This is the part I'm proudest of.

Applying a fix is a hypothesis: this should help. After you apply a fix, AgentReady re-queries Shopify's Global Catalog over the UCP MCP endpoint and checks whether the change was actually absorbed, whether the catalog now reads your product the way you fixed it. Absorbed is a fact about the catalog's copy of your product. It is not a promise that you will rank higher or sell more, and we don't present it as one. You can see the absorption check in action in how to tell if the catalog absorbed your products.

We keep that proof honest, too: per Shopify's catalog usage rules we never store catalog results or images, only derived metrics, the booleans and scores and timestamps that make the chart. The evidence, not the contraband.

Why we made it the spine, not a feature

We could have shipped "catalog support" as a tab and moved on. We didn't, because the loop, measure then fix then check, is the whole product thesis in miniature. Structured data, agents.md, llms.txt: those make you readable on your own domain. The catalog work makes sure the copy Shopify hands to agents is readable too.

It also lines up the incentives honestly. Checking is free because you deserve to know where you stand. Fixing, with the absorption check, is on every paid plan, from Core at $9 a month. Rank and buy-box tracking are not for sale: we ran them as a research preview and wrote up what we learned at /labs. If you want the full picture of where this fits, the catalog for AI-driven commerce piece maps it to the rest of agentic readiness.

Where this goes

Shopify is clearly building toward a world where agents don't just discover products in the catalog, they transact through it, and referrals and placements flow to the merchants with the cleanest data and the best agent experience. Nobody should model revenue off an invite-only preview, and we don't. Our bet is that the qualification for whatever comes next is the same thing that matters today: data the catalog can read correctly, and a record that you keep it that way.

That's the bet, and it's why we rebuilt around it now instead of waiting for the rest of the market to notice the shelf moved.

If you want to see where your store stands, it takes about thirty seconds and no signup: check your Catalog Readiness. AgentReady does not sell buy-box tracking today; what the app does do is check that the product facts on your store are complete and published correctly.

The shelf moved. The good news is the data that gets you onto it is yours to fix.

Quick answers

Frequently asked questions

What is Shopify's Global Catalog?
The Global Catalog is Shopify's cross-merchant index of products, built from each store's data, normalized and partly AI-inferred. Shopify says products reach AI channels such as ChatGPT, Copilot and Gemini through Shopify Catalog, so for a Shopify store, being in it with accurate data is the precondition for showing up in those channels.
Why does catalog data decide whether AI agents recommend me?
The catalog is only as good as the product data it's built from, and Shopify's own docs note inference accuracy depends on the source. Shopify does not publish how agents pick one product over another, but clean, complete, identifier-rich data is what lets an agent read and match your product correctly in the first place.
How does AgentReady help me win in the Global Catalog?
AgentReady checks how complete your product data is (free), fixes the gaps in your Shopify admin one field at a time with your confirmation and an undo (paid plans, from Core at $9 a month), then re-queries the catalog to check whether Shopify absorbed each fix. It does not sell rank or buy-box tracking; we ran that as a research preview and wrote it up at /labs.
Will AgentReady change my products automatically?
No. Every fix is propose, confirm, apply, one field at a time, with one-click undo. There are no blind or scheduled writes, and per Shopify's catalog rules we never cache or store catalog results; the absorption checks keep derived metrics only.

See where your store stands

Get found and recommended by AI shopping assistants.

Run the free AI-Readiness Checker to see, in about ten seconds, how ChatGPT, Perplexity, and Google read your store today and exactly what is holding it back. Then AgentReady prepares the fixes, starting with Schema.org structured data, and publishes them after you approve. Install free. Paid plans start at $9/mo.

Comments

Every comment here comes from a verified email. Write yours, confirm from your inbox, and it's live.

Loading comments…

Leave a comment

ShareXLinkedInFacebook

Written by Dylan Hunt, Founder, AgentReady. AgentReady measures what AI says about Shopify stores and helps teams improve the answer. See plans.