How AI assistants see a Shopify store

Some shoppers now ask an assistant before they search. For a merchant, the first question isn’t how to be recommended; it is whether the assistant can reach the store at all, and what decides that. The answer is less mysterious than the marketing around it suggests.

Four ways an assistant reaches your store

1. Its own search crawler. OpenAI, Anthropic and Perplexity run crawlers that build the index their assistants search: OAI-SearchBot, Claude-SearchBot and PerplexityBot.
2. A live fetch for one user. When someone asks a question, the assistant may visit a page there and then, as ChatGPT-User, Claude-User or Perplexity-User. Because a person started the request, OpenAI says robots.txt rules may not apply to ChatGPT-User, and Perplexity says Perplexity-User generally ignores them. Anthropic says Claude-User honors robots.txt.
3. Google’s index. AI Overviews and AI Mode draw on Google Search. Google says there are no additional technical requirements: a page must be indexed and eligible to be shown with a snippet.
4. Shopify Catalog. If you sell through agentic storefronts such as ChatGPT or Microsoft Copilot, Shopify shares your product data through Shopify Catalog, independently of robots.txt. That sharing is managed in your agentic storefront settings, not by crawler rules.

Separate from all four are training crawlers, such as GPTBot and ClaudeBot. They collect content that may be used to train models. Blocking them is a decision about training, not about whether you appear in search answers.

The crawlers, and what blocking each one does

What controls this on Shopify

Every Shopify store serves a default robots.txt. It allows public content and blocks pages such as the admin, cart, checkout and filtered or sorted collection URLs that duplicate content. Bot management at the network level is handled by Shopify, with no action needed from the merchant.

You can change it by adding a robots.txt.liquid template to the theme, to allow or disallow URLs, add sitemap URLs or block specific crawlers. Shopify is blunt about the risk: it is an unsupported customization, Shopify Support can’t help with it, and incorrect use “can result in loss of all traffic”. Shopify also recommends editing with Liquid rather than pasting plain text, so the store keeps receiving Shopify’s own updates to the default rules.

Two limits are worth knowing. Robots.txt rules are advisory: Shopify notes they aren’t guaranteed to be honored by every crawler. And blocking AI crawlers doesn’t stop Shopify Catalog from sending product data to agentic storefronts you have activated.

Shopify’s default template loops over robots.default_groups, the default rules, which Shopify updates regularly; keeping that loop is how Shopify recommends preserving those updates. A rule that isn’t part of a default group goes outside that loop. This example keeps every default rule and opts out of OpenAI’s training crawler only, leaving OAI-SearchBot, and so ChatGPT search, untouched:

{% 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: GPTBot
Disallow: /

Whether to opt out of training at all is a business decision, not a technical one. The point of the example is the separation: a training opt-out and a search opt-out are different lines.

Before changing anything, open yourstore.com/robots.txt and read what it says today. If a theme or developer added rules in the past, check that none of them block a search crawler you want to keep. Shopify also notes that a store with a robots.txt.liquid template can still serve rules the current default file has dropped, such as Disallow: /search; deleting the template restores the current default.

What they can read once they arrive

Access is only the first step. Google’s guidance for its AI features is the same as for Search: allow crawling in robots.txt and at the CDN or hosting layer, link pages internally, offer a good page experience, and make important content available as text. That last point matters on stores, where key facts often live only in images, sliders or app widgets.

• Product facts in text. Materials, sizes, compatibility and care instructions written on the product page, not only in images.
• Answers where the question is asked. Shipping, returns and sizing questions answered on the pages where shoppers have them, in plain sentences.
• Consistent names. The same product and brand names across product pages, collections and policies.
• No snippet blocks on pages you want used. nosnippet, max-snippet and noindex limit how Google can show a page, including in AI features.
• Structured data that matches the page. Google lists structured data that matches the visible text among its best practices, and says no special schema is needed for AI features. It is ordinary SEO, not a separate AI tactic.

Checking where you stand

1. Read your live robots.txt and list every rule that names a crawler.
2. In Search Console, AI Overviews and AI Mode traffic is included in the Performance report under the “Web” search type.
3. Ask the assistants your customers use the questions they would ask, and note whether and how the store appears. It is a manual spot check, not a measurement.
4. Fix what you control: crawl access, text content, consistent facts. Then check again later.

What nobody can promise

Each assistant decides for itself which sources to use, and those systems change often. No one can guarantee that a store will be cited or recommended. What a merchant can do is make sure the store isn’t shut out by accident and that its pages state the facts clearly enough to be quoted correctly.

That is the scope of Lintel’s generative engine optimization work, and it rests on the same foundations as performance and technical SEO.