Agentic commerce
Agentic commerce is online shopping that happens through AI agents — assistants like ChatGPT, Claude, and Perplexity that find, compare, and recommend products on a shopper's behalf.
Instead of a person browsing a storefront, an AI agent reads structured product data and answers the shopper's question directly ("find me a waterproof jacket under $150"). The store that publishes the clearest machine-readable data is the one the agent can understand and recommend.
For merchants, agentic commerce is a new discovery channel that sits alongside search and social — and it rewards clean structured data over clever marketing copy.
How AI shopping assistants find your store →
AI shopping agent
An AI shopping agent is an assistant that researches and recommends products for a shopper by reading the web's structured data rather than rendering pages like a human.
Examples include ChatGPT (with shopping), Perplexity, Claude, Google's AI Overviews, and Gemini. They prefer sources they can parse cleanly: Schema.org data, an llms.txt index, clear policies, and real ratings.
A store that's hard for an agent to parse simply doesn't get recommended, no matter how good the products are.
llms.txt
llms.txt is a plain-text file at a site's root that gives AI assistants a curated map of its most important pages, similar to how robots.txt and sitemap.xml help search crawlers.
It lists your key URLs with short descriptions so an assistant can find your products, collections, policies, and articles without crawling everything.
It's an emerging convention, not a guaranteed ranking factor — but it's cheap to publish and increasingly looked for by AI tools.
What is llms.txt, and should your store have one? →
Structured data
Structured data is machine-readable information about a page — its products, prices, reviews, and policies — added in a standard vocabulary (Schema.org) so search engines and AI agents can understand it precisely.
Without it, a machine has to guess what a page means from messy HTML. With it, the price, availability, rating, and return policy are stated explicitly.
On Shopify, structured data is what turns a product page into something an AI agent can confidently recommend.
The Shopify product schema guide →
JSON-LD
JSON-LD is the format Google and AI agents prefer for structured data — a small block of JSON, placed in the page's HTML, that describes the page using Schema.org types.
It sits in a `<script type="application/ld+json">` tag and doesn't change how the page looks. It's the practical way to publish Product, Organization, FAQPage, and other Schema.org data on a Shopify store.
Schema.org
Schema.org is the shared vocabulary — backed by Google, Microsoft, and others — that defines types like Product, Offer, Organization, and FAQPage used in structured data.
When a store publishes Schema.org JSON-LD, search engines and AI agents read the same agreed-upon fields (name, price, availability, aggregateRating) rather than parsing free text.
GEO (Generative Engine Optimization)
GEO is the practice of optimizing content so AI answer engines cite it in their generated responses — the AI-era counterpart to SEO.
Where SEO aims for a ranking on a results page, GEO aims to be part of the answer itself. It rewards clear, answer-first content, FAQ structure, factual sourcing, structured data, and freshness.
AEO (Answer Engine Optimization)
AEO is optimizing content to be the direct answer an engine returns to a question, rather than one of ten blue links.
It overlaps heavily with GEO; both prioritize concise, well-structured, citable answers and strong Schema.org markup (especially FAQPage).
AI Overviews
AI Overviews are Google's AI-generated summaries that appear above traditional search results, synthesized from pages Google already trusts.
Being cited in an AI Overview leans on classic strengths: ranking organically, strong E-E-A-T signals, and structured data. Good SEO still feeds AI visibility.
MCP (Model Context Protocol)
MCP is an open standard that lets AI assistants connect to external tools and data sources through a consistent interface.
For commerce, an MCP server can expose a store's catalog and capabilities directly to compatible AI clients, so an assistant can query products in a structured way rather than scraping a storefront.
Product schema
Product schema is the Schema.org structured data for a single product — its name, description, images, brand, price, availability, identifiers, and ratings.
It's the most important structured data for commerce: it's what enables rich results in Google and confident recommendations from AI shopping agents. Per-variant price and availability matter, not just a vague range.
Product descriptions that rank and parse →
FAQPage schema
FAQPage schema marks up a list of questions and answers so search engines and AI agents can lift a direct answer straight from your page.
It's one of the highest-impact structured-data types for AI visibility because its format matches exactly how people ask assistants questions.
FAQ schema for Shopify rich results →
Schema readiness score
A schema readiness score grades how completely and correctly a store publishes the structured data AI agents and search engines need.
It typically breaks down into coverage (is every product, page, and policy marked up?), structure quality (is the schema valid?), variant resolvability, and trust signals (real ratings, clear policies).
AI crawlers (GPTBot, ClaudeBot, PerplexityBot)
AI crawlers are the bots that retrieve web content for AI assistants — GPTBot (OpenAI), ClaudeBot (Anthropic), PerplexityBot (Perplexity), Google-Extended, and others.
Your robots.txt decides whether they can read your store. Blocking them keeps you out of AI answers; explicitly allowing them is how you opt into the channel.
Measuring AI-agent traffic to your store →
Theme app embed
A theme app embed is a Shopify feature that lets an app add functionality (like structured data) to a storefront through a single toggle, with no theme-code editing.
It's the Built-for-Shopify-friendly way to inject JSON-LD on every page, and it can be switched off cleanly so nothing stale is left behind.
