Is Your Store Ready for ChatGPT Shopping?

"Find me a lightweight waterproof jacket under £150 with good reviews."

Two years ago, that query went into Google. Today, a growing number of shoppers type it into ChatGPT, Perplexity, or Google Gemini - and expect a direct answer with product recommendations, prices, and buy links.

This shift is happening faster than most store owners realise. OpenAI's ChatGPT now has shopping features built in. Perplexity has launched a shopping hub. Google's AI Overviews surface product recommendations directly in search. Microsoft Copilot weaves product suggestions into conversations.

The question isn't whether AI shopping will matter. It's whether your store is ready for it right now.

How AI Shopping Actually Works

When someone asks an AI assistant to recommend a product, the AI doesn't browse shops like a human. It follows a fundamentally different process:

Step 1: Understanding the Query

The LLM parses the user's intent. "Lightweight waterproof jacket under £150 with good reviews" becomes a set of requirements: category (jacket), attributes (lightweight, waterproof), price constraint (<£150), quality signal (good reviews).

Step 2: Finding Products

Depending on the platform, the AI either:

Retrieves from training data - products and stores it learned about during training
Searches the web in real time - using browsing tools to fetch current product pages
Queries a product index - some platforms maintain dedicated shopping databases

ChatGPT's browsing mode and Perplexity actively crawl the web. They visit your product pages, read the content, and extract information. Google Gemini pulls from Google's existing Shopping index plus web content.

Step 3: Evaluating and Ranking

The AI evaluates found products against the user's criteria. This is where structured data becomes critical - the AI needs to confidently extract price, availability, specifications, and reviews to make a comparison.

Products with incomplete or ambiguous information get skipped. The AI won't recommend a product if it can't verify the price is under £150, or if there's no review data to assess quality.

Step 4: Presenting Recommendations

The AI generates a response with product recommendations, typically including the product name, key specs, price, a brief justification for why it fits, and a link to buy.

Your store either appears in that recommendation - or it doesn't.

What Makes a Store AI-Shopping Ready

Based on analysis of thousands of product pages through PageDiag, the factors that determine whether AI assistants recommend your products fall into five categories.

1. Complete Product Schema (Critical)

This is the single most important factor. JSON-LD Product schema tells AI crawlers exactly what your product is, what it costs, whether it's in stock, and what customers think of it.

The minimum viable Product schema for AI shopping:

{
  "@context": "https://schema.org",
  "@type": "Product",
  "name": "Trailblazer Waterproof Jacket",
  "brand": { "@type": "Brand", "name": "PeakGear" },
  "description": "Lightweight 3-layer waterproof jacket. 280g, fully taped seams, adjustable hood. Suitable for hiking and trail running.",
  "sku": "PG-TWJ-BK-M",
  "gtin13": "5060012345678",
  "image": ["https://example.com/images/trailblazer-front.webp"],
  "material": "3-layer Gore-Tex Active",
  "weight": { "@type": "QuantitativeValue", "value": "280", "unitCode": "GRM" },
  "offers": {
    "@type": "Offer",
    "price": "139.99",
    "priceCurrency": "GBP",
    "availability": "https://schema.org/InStock",
    "url": "https://example.com/products/trailblazer-waterproof-jacket"
  },
  "aggregateRating": {
    "@type": "AggregateRating",
    "ratingValue": "4.7",
    "reviewCount": "892"
  }
}

Missing fields we see constantly:

No brand - the AI can't match brand queries
No aggregateRating - reviews exist on the page but aren't in schema
No availability - the AI doesn't know if it's in stock
No material or weight - can't match specification queries
description is a marketing tagline - not a factual product summary

2. Detailed, Factual Product Descriptions

AI assistants extract product descriptions to evaluate and compare. "Our best-selling jacket - you'll love it!" tells an AI nothing. Compare that with:

Lightweight 3-layer waterproof jacket weighing 280g. Features fully taped seams, a helmet-compatible adjustable hood, two zippered hand pockets, and one chest pocket. 20,000mm waterproof rating with 25,000g/m² breathability. Packs into its own chest pocket. Available in 5 colours, sizes XS-XXL.

The second description gives the AI concrete data points to match against user queries. Weight, waterproof rating, breathability, features, sizes - every detail is a potential match.

3. Review Data in Structured Markup

Reviews are a trust signal for AI recommendations. An AI assistant answering "what's the best budget waterproof jacket?" needs to assess quality, and review data is the primary input.

It's not enough to display reviews on the page. They need to be in your structured data. The aggregateRating property (average score + review count) is the minimum. Individual Review schema items with review text give the AI even more context.

4. Correct robots.txt Configuration

Here's a mistake that's surprisingly common: stores that block AI crawlers in their robots.txt.

Some store owners, concerned about AI training on their content, have added directives like:

User-agent: GPTBot
Disallow: /

User-agent: ChatGPT-User
Disallow: /

User-agent: PerplexityBot
Disallow: /

This blocks ChatGPT and Perplexity from accessing your product pages entirely. If they can't read your pages, they can't recommend your products. You've effectively made yourself invisible to AI shopping.

The key distinction: GPTBot is used for training data collection. ChatGPT-User is used for real-time browsing when a user asks ChatGPT to find something. Blocking GPTBot may be reasonable. Blocking ChatGPT-User means ChatGPT can't see your products when a user is actively trying to buy.

Think carefully about which bots you block. The major AI shopping crawlers include:

ChatGPT-User - ChatGPT browsing (active shopping)
GPTBot - OpenAI training
PerplexityBot - Perplexity search
Google-Extended - Gemini training
Amazonbot - Amazon/Alexa
ClaudeBot - Anthropic

At minimum, allow ChatGPT-User and PerplexityBot if you want AI shopping visibility.

5. Clean, Crawlable HTML

AI crawlers read your page's HTML, not the rendered visual output. If your product information is locked inside JavaScript-rendered components that require browser execution, many AI crawlers won't see it.

Test this yourself: view your product page source (Ctrl+U in Chrome). If the product name, price, description, and specs are in the HTML source, you're fine. If the page source is mostly empty <div> tags that get filled by JavaScript, you have a problem.

Server-side rendering (SSR) or static generation solves this. Most major ecommerce platforms handle this correctly, but custom headless builds sometimes don't.

The AI Shopping Score

PageDiag's AI Shopping Score quantifies all of this into a single number from 0-100. It evaluates:

Schema completeness - How many relevant Product schema properties are present
Content quality - Factual density of product descriptions
Review integration - Are reviews present and in structured data
Crawlability - Can AI bots access your content
robots.txt - Are you blocking AI crawlers

A score of 80+ means your products are well-positioned for AI shopping recommendations. Below 50 means you're likely invisible to AI assistants when shoppers ask for product recommendations.

How to Check Your Store

Quick Manual Check (5 minutes)

Open ChatGPT and ask it to recommend a product you sell. Include your brand name. Does it know about you?
Check your robots.txt - visit yourstore.com/robots.txt and look for AI bot blocks
View page source on a product page - is the product info in the HTML?
Test your schema - paste your product URL into Google's Rich Results Test

Comprehensive Check (1 minute)

Scan your store at pagediag.com. PageDiag analyses your product pages across all the factors above and gives you a specific AI Shopping Score with actionable recommendations.

Platform-Specific Tips

Shopify

Shopify's default Product schema is basic - often just name and price. Install a structured data app like JSON-LD for SEO or Smart SEO to add complete schema with reviews, brand, materials, and availability.

Check your robots.txt at yourstore.myshopify.com/robots.txt. Shopify manages this file automatically, but you can add custom rules via the robots.txt.liquid template.

WooCommerce

Use Yoast SEO or Rank Math for Product schema generation. Both support aggregateRating when connected to a compatible reviews plugin. Ensure your product descriptions are detailed - WooCommerce makes it easy to have a short description and a long description; use both.

BigCommerce

BigCommerce includes basic Product schema by default. Enhance it with the Schema App or by editing your Stencil theme templates directly. BigCommerce doesn't block AI crawlers by default, which is a good starting position.

This Is the Future of Product Discovery

AI shopping isn't replacing Google - it's adding a new channel. Just as stores had to optimise for Google Shopping, then Instagram Shopping, then TikTok Shop, they now need to optimise for AI shopping.

The difference is that AI shopping optimisation aligns perfectly with good SEO and good UX. Complete structured data, detailed product descriptions, genuine reviews, and fast-loading pages - these are things you should be doing anyway. AI shopping just raises the bar on data quality and completeness.

The stores that invest in this now will capture early traffic from a channel that's growing rapidly. The stores that wait will play catch-up when their competitors are already established in AI recommendations.

Check your AI Shopping Score today at pagediag.com - it's free, it takes 30 seconds, and it'll tell you exactly where you stand.