Skip to main content

​
Three Things Every Store Must Do

Your store might look stunning, but if AI agents cannot find you, none of it matters. This chapter covers three things every merchant must have in place. You do not need to write code, but you do need to understand what they are and why they matter. Step-by-step implementation guides are in SEO for AI.

​
3.1 Label Your Products for Machines — Schema.org

​
Analogy: Barcodes in a Grocery Store

Every item in a grocery store has a barcode. Shoppers read the product name and price tag; the checkout scanner reads the barcode. Your website works the same way. Shoppers see images and text descriptions; AI agents read a “barcode” — a type of structured markup called Schema.org.

​
What Is Schema.org?

Schema.org is a standard jointly created by Google, Microsoft, Yahoo, and Yandex. It defines a set of “machine labels” that let search engines and AI agents understand web page content. For example, your product page says “Nike Air Max 270, $129.” A human understands that instantly. But to an AI agent, it is just a string of text. With Schema.org markup added: 
{
  "@type": "Product",
  "name": "Nike Air Max 270",
  "brand": "Nike",
  "offers": {
    "price": 129,
    "priceCurrency": "USD",
    "availability": "InStock"
  }
}
Now the AI agent understands precisely: this is a product, named Nike Air Max 270, the brand is Nike, the price is $129 USD, and it is in stock.

​
What Happens Without Schema.org?

The AI agent might misread your page — confusing shipping costs for the product price, or not recognizing the page as a purchasable product at all. The result: it will not recommend you, because it is not confident it understood correctly. -> Hands-on guide: Chapter 3: Schema.org in Practice

​
3.2 Write a “Company Brief” for AI — llms.txt

​
Analogy: The Brochure in a Company Lobby

You visit a company’s office. In the lobby, there is a brochure on the reception desk. In 30 seconds, you know what the company does, what its strengths are, and how to get in touch. llms.txt is your website’s “lobby brochure,” written specifically for AI agents.

​
What Is llms.txt?

A plain-text file placed at your website’s root directory (yourdomain.com/llms.txt). In concise language, it tells AI agents:
  • Who you are (company name, brand)
  • What you do (products or services you sell)
  • Your advantages (why someone should choose you)
  • Where to learn more (links to key pages)

​
Example

# ORBEXA
> AI Agent Commerce Infrastructure

ORBEXA powers agentic commerce. We provide trust scoring (OTR),
protocol translation (UCP/ACP/MCP), and merchant tools to help
AI agents discover, trust, and transact with online stores.

## Key Pages
- /verify — Check any domain's trust score
- /api/otr/verify/{domain} — Trust verification API
- /docs — Documentation
That is all it takes. Five minutes to write.

​
What Happens Without llms.txt?

The first time an AI agent visits your site, it has no idea who you are. It would need to crawl many pages to piece together a vague picture. In most cases, it will simply skip you and visit a competitor that has an llms.txt file. -> Hands-on guide: Chapter 6: Writing an llms.txt File

​
3.3 Give AI a “Store Directory” — Sitemap

​
Analogy: A Shopping Mall Floor Plan

You walk into a large shopping mall looking for a specific brand. Without a floor directory, you have to wander through every level — slow and inefficient. A Sitemap is your website’s “floor directory.” It tells AI agents: what pages your site has, what each page contains, and when it was last updated.

​
What Is a Sitemap?

An XML-format file (yourdomain.com/sitemap.xml) that lists the URLs of every important page on your site.

​
Why a Product Sitemap Is Especially Important

A standard Sitemap lists all pages. But for an e-commerce site, you also need a Product Sitemap that specifically lists all product pages. This lets AI agents quickly locate all your products instead of getting lost among blog posts and “About Us” pages.

​
What Happens Without a Sitemap?

The AI agent can only discover pages by following links one at a time. If a product page is buried three clicks deep, the AI agent may never find it. -> Hands-on guide: Chapter 8: Sitemap Optimization

​
How the Three Pieces Fit Together

Schema.org = Label each product (AI can read individual products)
    |
llms.txt = Give AI an overview (AI knows who you are and what you sell)
    |
Sitemap = Give AI a map (AI knows where to find your products)
All three are essential. Labels without a map means AI cannot find the products. A map without labels means AI finds them but cannot read them. Labels and a map without an introduction means AI does not know whether to trust you.

​
Self-Check Checklist

  • Visit yourdomain.com/llms.txt — if you get a 404, you need to create it
  • Visit yourdomain.com/sitemap.xml — if you get a 404, you need to generate it
  • On a product page, press F12 and search for application/ld+json — if nothing comes up, you need to add Schema.org markup
  • All three in place? Visit orbexa.io/verify and check your D dimension (Data Quality) score
Next chapter: Letting AI Place Orders for Customers — Commerce Protocols Overview — UCP, ACP, and MCP explained with everyday analogies