AI shopping visibility. What’s already automatic, and what actually decides whether you’re recommended.
Your products are probably already in ChatGPT and Google’s AI surfaces. Being listed isn’t the same as being understood — and understood is what decides whether you get recommended.
I make the catalogue readable enough to be compared properly.
I can’t make an assistant recommend you. Nobody can.
What you are already getting for free
Some of this you already have, and I’m not going to sell it to you.
Already on
Your products are likely already in ChatGPT.
Shopify feeds its product catalogue through, so for most eligible stores the listing happens without anyone doing anything. If someone is charging you to “get listed in AI,” check whether you already are.
Already on
Google’s AI shopping surfaces run on your Merchant Center data.
The same feed you already maintain. There is no separate AI feed to buy.
Near-free
The
llms.txtlayer is close to free.It’s a small file, and there are low-cost apps that write it. It is not a service.
So the work isn’t getting you in. The work is what happens after you’re in — whether an assistant can tell what your product is, who it’s for, and when to pick it over the other three it’s comparing.
What’s in it.
What’s in it
Attribute completeness, product by product.
Brand, category, product type, materials, size, dimensions, variants, identifiers, availability. The boring fields. An assistant comparing four products drops the one it can’t describe.
What’s in it
“Who this is for” written into the product data, not just the marketing copy.
AI shopping questions arrive as constraints — a budget, a use case, a body type, a restriction. A product that doesn’t state its own use case can’t be matched to one.
What’s in it
Comparison material.
What this product is better at, what it isn’t, what it doesn’t do. Stated plainly enough to be quoted.
What’s in it
Reviews treated as product knowledge.
“Runs small”, “battery lasts about eight hours”, “good for sensitive skin” — those are attributes, and most stores bury them.
What’s in it
One source of truth.
Your feed, your product page, your structured data and your checkout agreeing on price and availability. They disagree more often than you’d think, and a disagreement is a dropped product.
When this is the right page.
We want to be recommended when someone asks an AI assistant.
The usual starting point — and the first thing to check is what you already have.
Our products are listed but never come up in comparisons.
Usually attribute completeness rather than placement.
How it goes.
Step 1
Same seven steps as everything else I do. What’s different here is what step two inspects: your catalogue, your feed and your structured data, read the way a shopping system reads them rather than the way a customer does.Step 2
Then I fix what’s missing — in the theme and in the product data. Same person. A list of missing attributes handed to someone else is a list of missing attributes next month.Step 3
Then I re-check it against the same validators and feeds. You can run them yourself afterwards; that’s the point of choosing checks a third party owns.
What I won’t be doing.
I can’t promise an assistant recommends your product. Nobody can, and anyone who does is selling you something they don’t control. What gets recommended depends on the model, the question, the competition and a dozen things that change without notice.
No ads, no off-site SEO, no link building, no content farm. And no “AI share of voice” reporting yet. Google says it’s coming to Merchant Center. When there’s a real instrument I’ll use it. Until then I’d be charging you for a number I made up.
Where to go next
Otherwise: send me your store and I’ll tell you what an assistant can and can’t currently work out about your products.
Send your store URL- If your products are being disapproved or your markup doesn’t validateStart with product schema and Merchant Center
- If the real problem is that traffic arrives and doesn’t buyStart with Shopify CRO