How to get your Shopify products recommended by ChatGPT

More shoppers now ask ChatGPT what to buy before they search Google.

They describe what they need in a full sentence, compare a few options and click through when they're ready. That traffic is small today, yet it's growing fast and it buys.

Adobe found that generative AI traffic to US retail sites grew 693% over the 2025 holiday season, and AI referrals converted 31% better than other traffic.

A year earlier, AI visitors were far less likely to buy than anyone else.

AI shoppers went from browsing to buying

How likely visitors from AI tools were to buy on US retail sites, compared with other visitors

Source: Adobe Analytics, via Adobe and Marketing Dive.

I lead Shopify development at Blu Mint, and I'm currently working on the Shopify store for JALG, a maker of award-winning TV stands that sells to customers across the USA, Europe and Japan.

It's too early to share results on AI recommendations. This guide covers what we're putting in place and why.

How ChatGPT finds Shopify products

There are two routes into ChatGPT for a Shopify store. One is free, and one is paid.

Organic recommendationsChatGPT ads
How it worksShopify Catalog shares your products with ChatGPT and other AI tools through Agentic StorefrontsOpenAI's Shopify app syncs your catalogue to ChatGPT Ads Manager for paid placements
CostNo extra transaction fees beyond standard processingYou pay for ads, with CPMs reported from around $25
What decides itProduct data, reviews and your reputation across the webBudget, bidding and the quality of your product feed
Best forEvery Shopify store, as a foundationStores ready to test a new channel with a set budget

Organic recommendations are where I'd start. Paid ads only work well when the product data underneath is already strong.

Check that Agentic Storefronts is switched on

In March 2026, Shopify launched Agentic Storefronts, which brings merchants' products into ChatGPT, Microsoft Copilot, Google's AI Mode and the Gemini app from one place in the Shopify admin.

Shopify says it activates Agentic Storefronts automatically for eligible stores, and every sale still goes through Shopify Checkout. In ChatGPT, mobile shoppers complete their purchase in an in-app browser, while desktop shoppers are sent to your store.

Open the Agentic Storefronts section of your Shopify admin and check which channels are active. For Google's AI Mode and Gemini, you also need to connect your Google Merchant Center account through the Google & YouTube app.

Being switched on doesn't mean being recommended, however. That depends on what AI tools can learn about your products.

Give AI tools complete product data

AI shopping tools work from structured data. When a shopper asks for "a solid oak TV stand under $1,000 that fits a 75-inch TV", the tool needs your material, price, dimensions and stock level in a form it can read without guessing.

Fill in every product field

Complete every field in Shopify, including product type, vendor, variants and metafields for details such as material, dimensions and care. Add GTINs or manufacturer part numbers where your products have them.

Make your schema complete

Check that your theme outputs product schema with brand, price, availability, variants and review ratings. Test your product pages in Google's Rich Results Test and fix any missing fields.

Max out your merchant feeds

Sync your store with Google Merchant Center and Microsoft Merchant Center, and don't leave optional attributes blank. Colour, material, size and age group all help AI tools match your products to specific requests.

Write product pages that answer questions

People search Google in a few words. They ask ChatGPT in full sentences, with their situation included.

Write descriptions around how people use the product, not only what it is. "Waterproof boots" tells an AI tool very little. "Built for wet winter hikes and sub-zero mornings" matches the way people actually ask.

Add a short FAQ to each product page covering sizing, compatibility, materials and care. For JALG, we split care advice by material, because oak and birch need different treatment. Only someone who knows the products could write that.

That kind of detail is what AI tools look for and what AI-only content can't fake.

Collect reviews that say something

When a shopper asks which product is most durable or best for a small room, AI tools lean on what reviewers have said.

"Great product!" doesn't help. A review that explains how someone used the product, and what they liked or didn't, gives AI tools something to quote.

Use a review app that adds review data to your product schema. In your post-purchase emails, ask customers specific questions about how they used the product, not only for a star rating.

Build a reputation beyond your store

AI tools don't rely on your store alone. They read review sites, comparison articles, forums and press coverage, then summarise what people say about you.

If nobody talks about your brand outside your own website, AI tools have little reason to recommend it. Digital PR, reviews from credible publications and genuine discussion in communities such as Reddit all feed into how AI describes your products.

It's the same principle behind brand visibility in AI tools generally, and why I see GEO as an extension of good SEO, not a replacement for it.

Should you try ChatGPT ads?

ChatGPT ads launched in 2026, and in September OpenAI released a ChatGPT Ads app for Shopify that syncs your product catalogue to ChatGPT Ads Manager and sets up conversion tracking.

The app is free to install, and you pay for the ads themselves. Industry trackers report starting CPMs of around $25, shopping carousels for brands with product feeds and Sponsored Agents that answer shoppers' questions before they click.

It's still early.

As of now, the app has a 2.2/5 rating from 21 reviews, and several merchants report that their products don't sync.

I'd treat ChatGPT ads as a test with a fixed budget, not a main sales channel yet. Ecommerce brands are a better fit than B2B, since ChatGPT ads don't reach most B2B buyers yet. Our AI search ads team can run a test alongside your Google and social campaigns.

Blu Mint – eCommerce website

What's hype and what works

A lot of what's being sold as AI optimisation is repackaged SEO. Some of it is a real change in how people find products.

Hype to ignoreWhat actually works
ContentHundreds of AI-written blog posts to build "semantic depth"Fewer, expert articles that link to the products they discuss
TechnicalSpecial code or hidden tricks aimed at chatbotsComplete schema and merchant feeds with every attribute filled in
Product pagesStuffing exact-match keywordsUse cases and FAQs written the way people ask questions
ReputationOptimising your own pages and nothing elseReviews, press coverage and real discussion of your brand elsewhere
AdsTreating ChatGPT ads as a guaranteed sales channelA small, measured test once your product data is strong

AI tools compress what the web says about you into a short answer. You can't trick them. You can make sure your data is complete and that real people say good things about your products.

How to track AI recommendations

ChatGPT doesn't publish search volumes or click data, so tracking needs different tools. David Bailey, who leads our SEO and AI search work, uses Ahrefs Brand Radar to track how often AI tools mention a brand compared with its competitors.

Brand Radar separates mentions, where AI tools name your brand, from citations, where they link to your site. Both matter, because visibility and traffic aren't the same metric in AI search.

On a small budget, you can track manually. Write 20 to 30 questions your buyers would ask, run them in a logged-out ChatGPT session and record whether your brand appears, whether it's linked and which competitors show up.

David's advice is to look for trends, not single answers:

“Don’t stress over a single bad answer. Track the data weekly or monthly to spot directional movements.”
— David Bailey, Blu Mint Senior SEO Expert

You can also see visits from ChatGPT and other AI tools in Google Analytics 4 by creating a custom channel for AI sources.

Our Shopify development and ecommerce SEO services prepare stores for AI shopping from the ground up, and our AI search optimisation work tracks the results. We work alongside our SEO, content writing, PPC, social media and web design teams. If you're building a new store, the same foundations apply when you start an online store, and they'll matter even more as agentic commerce grows.

Previous
Previous

The science of writing: Flesch reading ease scores

Next
Next

How to start an online store and create an e-commerce website