By Steve Merrill, Founder of WRKNG Digital — June 8, 2026
The five new ChatGPT shopping features that matter most in 2026 are image search, browse-in-chat, conversational refinement, side-by-side comparison, and visual product cards. Each one changes which parts of your product data ChatGPT actually uses — and which stores get recommended.
ChatGPT is now handling50 million shopping queries per day. According toAdobe Analytics, traffic from ChatGPT to retail sites is up 354. 7% year-over-year as of May 2026. That's not a trend to watch. That's a live channel most Shopify stores aren't ready for.
The reason they aren't ready: their product data was built for Google. ChatGPT needs something different — and those differences just got bigger.
1. Image Search
ChatGPT now lets users upload a photo and find visually similar products in real time. The model pattern-matches against your images — your title or keyword field. Your store needs high-resolution photos from multiple angles, accurate alt text that describes what's actually in the image, and any visual metadata your platform supports. No good images means no match.
2. Browse-in-Chat
Users can now browse products inside a ChatGPT conversation without opening a new tab. ChatGPT crawls your product pages directly to pull the details it displays, so clean URL structure, fast page loads, and anllms. txt fileare no longer optional. If your content is locked behind JavaScript or your pages aren't crawlable, you're invisible to this feature entirely.
3. Conversational Refinement
When someone says "show me a cheaper option" or "do you have this in blue," ChatGPT needs to find that specific variant in your data. If your color fields are empty, your size options aren't structured, or your variants aren't mapped to individual prices, ChatGPT can't refine — it just moves to a store with complete data.This feature rewards the stores that did the boring work of filling in every attribute field.
4. Side-by-Side Comparison
ChatGPT now presents products head-to-head on price, reviews, and specs. If your features live in a block of unstructured text, they won't populate a comparison card. You need review count, star rating, all-in pricing including shipping fees, and product attributes in clean, parseable fields — becauseOpenAI's ChatGPT Shoppingpulls structured data directly when it builds those cards.
5. Visual Product Cards in Responses
Every ChatGPT shopping response now surfaces a product card with your hero image, title, brand name, and availability status. Your title needs to be under 70 characters — anything longer gets cut in the card. Your hero image needs to be high resolution and your availability needs to reflect real-time stock, because a card showing an out-of-stock product loses clicks fast.
The Data Layer Underneath All Five Features
All five features connect throughACP — Agentic Commerce Protocol— the structured data layer ChatGPT uses to access product information. According toShopify's Agentic Storefronts announcement, over 5. 6 million Shopify stores are now activated for AI shopping. Being activated and being optimized are two different things. Most stores are the first. Very few are the second.
How We Chose This List
These five features were announced or significantly updated by OpenAI between May and June 2026. Each one was selected based on a single criterion: where incomplete product data most often causes a store to get skipped over in ChatGPT shopping responses.
Frequently Asked Questions
What product data does ChatGPT shopping look for?
ChatGPT looks for structured product attributes: title under 70 characters, description. Price with all fees included, real-time availability, review count and star rating, high-resolution images from multiple angles, variant data (color, size, material), and crawlable product pages with clean URLs. The more complete your data, the more features ChatGPT can use to surface your products.
Do I need an llms. txt file for ChatGPT shopping?
Yes. An llms. txt file tells AI crawlers what's on your site and how to navigate it. Without it, ChatGPT's browse-in-chat feature may miss or misread your product pages. It's a simple text file and one of the highest- fixes for Shopify stores right now.
How does ChatGPT's image search work for shopping?
A user uploads a photo and ChatGPT finds visually similar products by pattern-matching image embeddings and visual metadata. Your product images need to be high resolution, shot from multiple angles, with alt text that accurately describes what the product looks like — the product name.
What's the difference between ChatGPT shopping and Google Shopping?
Google Shopping ranks products based on bids, feed quality, and historical performance data. ChatGPT shopping ranks based on data completeness, conversational relevance, and how well your product answers the specific question being asked. The optimization strategies are fundamentally different.
How much ChatGPT shopping traffic is there in 2026?
OpenAI reports 50 million shopping queries per day on ChatGPT. Adobe Analytics puts year-over-year growth in ChatGPT-driven retail traffic at 354. 7% as of May 2026. Both numbers are still rising. This is a current, active channel — not a future one.
Want to know exactly where your Shopify store stands on these five features? We built a tool that audits your product data against what ChatGPT actually needs to recommend you.Get your AI Commerce Readiness score at WRKNG Digital.
