Google says There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary.
Its list of basics that still matter includes Making sure your structured data matches the visible text on the page
and Checking that your Merchant Center and Business Profile information is up-to-date
. Google’s page on AI features and your website.
The same basics carry over to other AI tools, but each one gets your products by a different route. This post covers those routes for a Shopify store, what AI tools can read on your pages, and what you control.
How each AI tool gets your products
- Gemini and Google AI Mode. Shopify’s help center says products reach them through the Google & YouTube sales channel or Google Merchant Center, and that product discovery there is powered by Merchant Center. To sell there through Shopify, you need a valid Merchant Center account that meets all of Google’s requirements. A Merchant Center warning or suspension matters here too: see why Google Merchant Center suspends accounts. If you ship from a supplier, our post covers whether dropshipping stores can show up in Gemini and AI Mode.
- ChatGPT. Shopify says products reach ChatGPT through Shopify Catalog, which is on by default for eligible stores. ChatGPT’s search features also crawl the web with OpenAI’s OAI-SearchBot. See will my Shopify store appear in ChatGPT Shopping.
- Perplexity. Perplexity’s crawler page says PerplexityBot is designed to surface and link websites in Perplexity’s search results. Your web pages are what it reads.
Your product data in Shopify feeds the first two. Your storefront pages feed every tool that crawls the web.
What AI tools can read on your store
1. Product JSON-LD on product pages. This is structured data that states product facts. Google’s merchant listing guide requires name, image and offers (with price and priceCurrency), and recommends description, brand, availability, priceValidUntil, aggregateRating and review, among others. Shopify’s example output for themes that use its structured_data filter doesn’t include aggregateRating or review; those usually come from a reviews app. Missing brand is another common gap.
2. Identifiers and attributes. Google also recommends gtin, mpn, sku, material, pattern, color and size for merchant listings. Your theme’s built-in markup may not include them.
3. llms.txt at the site root. A Markdown map of your key pages, proposed on llmstxt.org in 2024 and still not a formal standard. Shopify serves one for every store. Google says Google Search ignores it; other AI tools may read it.
4. Organization and WebSite JSON-LD on the homepage. Your store’s name, logo and social profiles (through sameAs), and your site name.
If all four are in place and accurate, you’ve covered what’s in your control on the page.
Adding identifiers to your structured data
Shopify’s product markup may leave out identifiers like GTIN and MPN. Three ways to add them:
Manual (small catalogs): fill in each product’s metafields in Shopify Admin (gtin, mpn and so on), then update your theme’s product template to output them in the JSON-LD block.
Theme block (medium catalogs): a schema app from the Shopify App Store can add a block that reads your metafields and outputs the extra properties on every product page.
Feed (large catalogs): fill the fields your Google & YouTube sales channel sends to Merchant Center. This reaches Google, including Gemini and AI Mode for Shopify stores. Other AI tools read your storefront or, for ChatGPT, Shopify Catalog.
llms.txt
Shopify’s help center says every store serves /agents.md, /llms.txt and /llms-full.txt automatically, with your store name, store URL, sitemap link and store policies. A custom file might look like this:
# Cool Socks Co
> Merino wool socks, made in Vermont since 2019.
## Contact
- Email: hello@coolsocks.com
- Phone: +1 (555) 123-4567
## Policies
- [Refund policy](https://coolsocks.com/policies/refund-policy)
- [Shipping policy](https://coolsocks.com/policies/shipping-policy)
- [Privacy policy](https://coolsocks.com/policies/privacy-policy)
## Products
- [Product feed](https://coolsocks.com/products.json)
- [Sitemap](https://coolsocks.com/sitemap.xml)
To customize it, add an agents.md.liquid or llms.txt.liquid template to your theme. Google Search ignores the file, so treat it as an extra for other AI tools. More in what llms.txt is and how to add it to your Shopify store.
Which AI crawlers to allow
You don’t have to allow every AI crawler. Each company publishes the names its crawlers use in robots.txt:
- OpenAI (crawler page): GPTBot (content that may be used to train its models), OAI-SearchBot (ChatGPT’s search results), ChatGPT-User (pages a ChatGPT user asks about; OpenAI says robots.txt rules may not apply)
- Anthropic (help page): ClaudeBot (training), Claude-SearchBot (search results), Claude-User (pages a Claude user asks about)
- Perplexity (crawler page): PerplexityBot (search results; Perplexity says it isn’t used to crawl content for AI foundation models), Perplexity-User (user requests; Perplexity says it generally ignores robots.txt)
- Google (crawler list): Google-Extended, a robots.txt token rather than a separate crawler, controls whether Google may use your content for Gemini training and grounding. Google says
Google-Extended does not impact a site’s inclusion in Google Search nor is it used as a ranking signal in Google Search.
- Common Crawl (CCBot page): CCBot, which builds Common Crawl’s open repository of web crawl data
- Amazon (Amazonbot page): Amazonbot, which Amazon says may be used to train Amazon AI models
- Apple (Applebot page): Applebot-Extended, which Apple says doesn’t crawl pages; disallowing it opts your content out of training Apple’s generative foundation models
Each setting is separate. OpenAI says, for example, that you can allow OAI-SearchBot to appear in search results while disallowing GPTBot. Blocking training crawlers while allowing search crawlers is one option; block only the crawlers you have a reason to block.
To set this up, add a robots.txt.liquid template to your theme. Shopify’s developer docs say Shopify generates a robots.txt for every store and the template replaces it. Shopify’s help center adds that robots.txt rules are advisory, not every crawler follows them, and blocking crawlers doesn’t stop Shopify Catalog from sending your product data.
Organization and WebSite schema
The homepage JSON-LD that tells search engines and AI tools who you are. Two blocks:
Organization: name, url, logo, sameAs (links to your social profiles), contactPoint, address.
WebSite: name, url, and optionally potentialAction (a search URL template). Google still uses WebSite markup for site names, but it removed the sitelinks search box that used the search template. Its announcement said we’ll be removing this visual element starting on November 21, 2024
, and that site names also uses a variation of WebSite structured data, which continues to be supported.
Theme support varies. Shopify’s free Dawn theme outputs both (WebSite on the homepage only); other themes may not. If yours doesn’t, paste hand-written JSON-LD into a custom Liquid block on the homepage.
Putting it together
- Keep your Merchant Center account in good standing and your Google & YouTube product sync on. Shopify says to keep syncing on so the information in Merchant Center matches your online store.
- Check your Product JSON-LD includes brand, and reviews if you show them, and that it matches what the page shows.
- Add identifiers (gtin, mpn) and attributes (color, size, material), manually for small catalogs or with a theme block for larger ones.
- Add Organization and WebSite JSON-LD to your homepage if your theme doesn’t.
- Decide which AI crawlers to allow in robots.txt.
- Optionally, customize llms.txt for AI tools that read it. Google ignores it.
For the Merchant Center side, see the Google Merchant Center suspension checklist. The free store check looks at your policies, contact details, whether your storefront is publicly reachable and whether your product pages carry working product markup.