How to Get Your Shopify Store Recommended by ChatGPT

Published 2026-11-03 · ipok.fyi

When someone asks ChatGPT, Perplexity or Claude "what's the best merino running shoe?", the answer comes from pages those systems could crawl, parse and trust. You can't buy your way into that answer. You can make sure nothing technical is keeping you out.

To see where real stores stand, we ran a basic AI-visibility check on 10 well-known Shopify/DTC brands on November 3, 2026. One returned HTTP 429 (rate-limited) to our checker, so it was excluded, which leaves 9 valid stores. The check covered the homepage plus one product page per store. Here's what we found, and how to fix each gap on your own store.

What the sample showed

The average score was about 86/100. Big brands get the basics right. The gaps are in structured product data, and those are exactly the gaps that matter for shopping-style answers.

Step 1: Make sure AI crawlers are allowed, explicitly

There are two kinds of AI crawler. Some collect training data (GPTBot, Google-Extended). Others fetch pages live to answer a user's question: OAI-SearchBot powers ChatGPT search, ChatGPT-User fetches pages on a user's request, and PerplexityBot and ClaudeBot serve their own products. If you want to be recommended, the search and user-fetch bots are the ones you can't block.

Shopify's default robots.txt allows all of them. That's why 9/9 stores passed. But none of the 9 named these bots explicitly. A theme change, an SEO app or a CDN bot-protection rule can block them without anyone noticing. Bot protection was the reason one store in our sample couldn't be checked at all.

Edit Online Store → Themes → Edit code → Templates → robots.txt.liquid. If the file doesn't exist, add it. Keep Shopify's default loop, then append:

{% for group in robots.default_groups %}
  {{- group.user_agent }}
  {%- for rule in group.rules %}
    {{ rule }}
  {%- endfor %}
  {%- if group.sitemap != blank %}
    {{ group.sitemap }}
  {%- endif %}
{% endfor %}

User-agent: OAI-SearchBot
Allow: /

User-agent: ChatGPT-User
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: ClaudeBot
Allow: /

If you use Cloudflare or another WAF, check that "bot fight" modes aren't returning 403 or 429 to verified AI crawlers.

Step 2: Put Product JSON-LD in the raw HTML, with GTIN and ratings

AI systems rarely run your JavaScript. If your price, stock status and reviews only show up after scripts load, the crawler may never see them. Open view-source: on a product page and search for "@type": "Product". If you can't find it there, it isn't reliably visible.

Most of our sample had the basic Product fields (name, offers, price, availability). The gaps were:

• GTIN missing on 6 of 8 product pages. Fill in the Barcode field on each variant in Shopify admin. Then make sure your theme's JSON-LD outputs it, for example "gtin13": {{ variant.barcode | json }}.
• aggregateRating missing on 4 of 8. Turn on your review app's JSON-LD or rich-snippet option. Then confirm aggregateRating (ratingValue and reviewCount) appears in the raw HTML, not just in a widget injected later.
• Brand. Add "brand": {"@type": "Brand", "name": "{{ product.vendor }}"} if your theme leaves it out.

Step 3: Add Organization and WebSite schema to your homepage

5 of our 9 stores were missing at least one of these. They tell an AI model who you are and which social profiles belong to you. Add this to theme.liquid, inside {% if template == 'index' %}:

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@graph": [
    {"@type": "Organization", "name": "{{ shop.name }}", "url": "{{ shop.url }}",
     "logo": "https://yourstore.com/logo.png",
     "sameAs": ["https://instagram.com/yourbrand", "https://tiktok.com/@yourbrand"]},
    {"@type": "WebSite", "name": "{{ shop.name }}", "url": "{{ shop.url }}"}
  ]
}
</script>

Step 4: Publish an llms.txt

/llms.txt is a proposed convention: a short markdown file that tells language models what your site is and where the important pages are. It isn't an official standard and no AI company guarantees they read it. It costs minutes to make, though, and 7 of 9 stores in our sample already have one. Shopify can't serve arbitrary files from the root, so use a URL redirect (/llms.txt → a file uploaded under Content → Files) or an app. Include:

• One paragraph on what you sell and who it's for
• Links to your top collections and bestsellers
• Shipping, returns and warranty policy URLs
• Your About page

Step 5: Answer the questions people actually ask

0 of 9 stores had FAQ content marked up. Shoppers ask AI comparison questions: "Is X machine washable?", "How does sizing run?", "What's the return window?" Add a short FAQ block to product or collection pages with real answers in plain HTML. Optionally mark it up with FAQPage schema. Plain, specific, quotable answers are what end up in AI responses.

Step 6: Don't neglect the basics

Titles, meta descriptions, canonicals and sitemaps were fine on nearly every store we checked, and Shopify handles most of them for you. Two things to check: meta descriptions under about 160 characters (one store's was 193), and an og:image on your homepage (3 of 9 had missing or incomplete Open Graph tags).

What this won't do

Passing these checks means AI systems can read and understand your store. Whether they recommend you still depends on reviews, press mentions, Reddit threads and how well your product fits the question. Technical readiness is the entry ticket.

Check your own store

You can run the same free check we used on your store at ipok.fyi. It shows which AI crawlers can reach you and what's missing from your product schema.

Methodology: automated check of 10 stores on Nov 3, 2026, with 1 excluded for rate-limiting (HTTP 429). Results cover the homepage and one auto-selected product page per store. In one case the selected page wasn't a real product, so product stats use n=8.

Run the free GEO / AI visibility check on your store →