Agent Readiness in Shopify

Where agent readiness breaks on Shopify

A Shopify page is built on Shopify's servers by your theme's Liquid templates, then delivered through Shopify's CDN. That gives you complete server-rendered HTML for the theme's own content, without any configuration. What arrives in that HTML, and whether a crawler is allowed to ask for it, is still up to you. The usual trouble spots are:

Server speed is rarely the issue on Shopify, because you do not run the servers. If the audit reports slow rendering, it is almost always the amount of app script and the content that waits for it.

Store access and robots.txt

The password page

New stores are password protected until they are launched, and some owners switch it back on during a redesign. While it is on, every crawler receives the password page instead of your content. Check under Online Store, Preferences, and fetch your home page with curl to make sure the response is your store rather than a login form.

Editing robots.txt the Shopify way

You cannot upload a robots.txt file to a Shopify store. Shopify generates a default one with sensible rules, such as keeping crawlers out of the basket, checkout, customer accounts and internal search results. To change it, you add a robots.txt.liquid template to your theme, which then controls the output. The template Shopify documents as the starting point reproduces the default rules:

{% for group in robots.default_groups %}
  {{- group.user_agent }}

  {%- for rule in group.rules -%}
    {{ rule }}
  {%- endfor -%}

  {%- if group.sitemap != blank -%}
    {{ group.sitemap }}
  {%- endif -%}
{% endfor %}

Anything you add goes after that loop. For example, if you decide to opt out of a particular training crawler while leaving AI search crawlers alone, you add a group for that crawler's documented user agent name:

User-agent: ExampleTrainingBot
Disallow: /

If the audit says a crawler is blocked, open /robots.txt on your live domain and check whether a robots.txt.liquid template exists in the theme. Themes bought from third parties and edits made by past developers are the usual source of surprise rules. Which crawlers to allow is a business choice; the crawler access guide explains the options, and the Shopify robots.txt and sitemap guide covers the template in more depth.

App content that crawlers never see

Shopify apps usually add content to your storefront in one of two ways. Some use theme app blocks or Liquid snippets that render on Shopify's servers, so the output is in the HTML. Many others add a script that runs in the browser, calls the app's own servers and inserts the content afterwards. For a crawler that does not run scripts, that content does not exist.

Content commonly added this way includes:

• Review text and star ratings
• Product questions and answers
• Size guides and fit advice
• Delivery estimates and stock messages
• FAQ accordions and specification tabs added by page-builder apps
• Bundles, "frequently bought together" and upsell blocks

To check a product page, fetch it with curl or view the source (not the element inspector) and search for a phrase from a review, a size-guide row and an FAQ answer. Whatever is missing is script-loaded. Then check the app's settings and documentation: some review apps can write a summary to product metafields or render a server-side block, which puts at least the rating and a few reviews in the HTML. Where an app has no such option, and the content matters, consider whether the same information can live in the product itself.

Product pages a crawler can use

An AI assistant answering "which waterproof walking boots have a wide fit under £150" needs facts in text: materials, sizes, fit, weight, care, price. On many stores those facts are in a supplier image, a PDF or a tab filled by an app. Put them in the page as text instead.

Specifications in metafields

Create product metafield definitions in the admin under Settings, Custom data, for the specifications you sell on: material, dimensions, weight, compatibility, care. Fill them for each product, then connect them to text or table blocks in the theme editor using dynamic sources. The values are rendered by Liquid, so they arrive in the HTML, stay consistent across products and can also feed your structured data.

Descriptions that answer questions

Descriptions copied from a supplier appear word for word on every store that sells the product, and they rarely say who the product suits. Rewrite the first paragraph for your best sellers so it states plainly what the product is, who it is for and what distinguishes it, then follow with the detail. For a fictional store, Fernside Outdoor, that might read: "A waterproof leather walking boot in wide and standard fits, for day walks on wet ground, with a grippy rubber sole and a 12-month guarantee."

Collection and blog pages

Collection pages with only a product grid give a crawler little to work with. A short introduction explaining what is in the collection and how to choose helps. On the blog, Shopify records the staff member who wrote each article, but most themes show little about them; the author and trust signals guide covers what to add.

The fix, step by step

The rest of the audit on Shopify

Several findings work the same on Shopify as anywhere else. Use the general guides for these, with the Shopify notes below:

• Structured data. Many current themes output Product JSON-LD automatically, and SEO apps often add their own. Keep one source and make sure it includes price, availability and brand. See fixing missing or incomplete JSON-LD and the Shopify JSON-LD guide.
• Duplicate URLs. Collection-path product URLs, filtered collections and market subfolders are covered in fixing canonical and URL conflicts.
• Page structure. Section files with headings chosen for size are the usual Shopify problem; see fixing non-semantic HTML.
• Extractability. Content in tabs, sliders and images; see fixing poor content extractability.
• llms.txt. Because Shopify does not let you upload files to the root of the domain, adding one needs a workaround; the Shopify llms.txt guide covers the options.

Staying ready through theme and app changes

1. Before installing an app that adds storefront content, ask whether it renders in the HTML or loads by script, and test it on a duplicate theme.
2. After each theme update, check that robots.txt.liquid, the canonical tag and your structured data survived, since updating a theme can mean moving to a fresh copy of its code.
3. Keep specifications in metafields so a theme change keeps the data, and you only need to reconnect the blocks.
4. Re-run the Agent Readiness Audit on a product, a collection and an article every few months and after any major store change.

If you also want AI agents to be able to add to basket or check out, rather than read your pages, that is a separate question covered in the Shopify agent compatibility guide.