MERCHANT GUIDE
What AI shopping agents
mean for your brand.
A plain-language guide to agentic commerce. No code, no jargon β just what you need to know as a brand owner, marketer, or e-commerce leader.
01
What is agentic commerce?
Instead of customers browsing your website themselves, their AI assistant does the shopping for them. The customer tells the agent what they need, and the agent finds products, compares prices, and completes the purchase β all without a human touching a browser.
Your Customer
βFind me new running shoes under $150β
AI Shopping Agent
Searches, compares, adds to cart, checks out
Your Store
Order placed, payment processed, item shipped
02
Why does this matter for your brand?
THE OLD WAY
- βCustomers Google your products
- βSEO determines who gets found
- βHumans browse, compare, and buy
THE NEW WAY
- +AI agents search for products on behalf of customers
- +Agent-readiness determines who gets found
- +Agents browse, compare, and buy β automatically
If AI agents can't find or understand your products, they'll shop somewhere else. This is like not showing up in Google search results β except it's happening right now with AI assistants like ChatGPT, Claude, and Gemini.
03
What is an ASX Score?
Your ASX Score is a number from 0 to 100 that tells you how easy it is for AI agents to shop at your store. Higher score means more AI-driven sales. It's measured across three areas:
CLARITY
Can agents read your catalog?
Do your product pages have structured data that machines can parse? Is your sitemap complete?
DISCOVERABILITY
Can agents find products?
Does your site search return relevant results? Do pages load fast enough for automated browsing?
RELIABILITY
Can agents complete a purchase?
Is guest checkout available? Can agents select variants, manage a cart, and enter payment details?
04
How do I check my score?
Enter your store's domain in the scanner. An AI agent will visit your site, evaluate all 11 signals, and generate your ASX Score with a detailed breakdown of what's working and what needs attention.
05
How do I improve my score?
These are the most impactful things you can do. Share this list with your developer or agency β they'll know how to implement each one.
Add structured product data
Make sure each product page has machine-readable information β name, price, availability, images. Your developer can add this as JSON-LD markup.
Make your site search work well
AI agents use your site's search bar to find products. If it returns irrelevant results or doesn't work reliably, agents will leave.
Keep your sitemap updated
A sitemap tells agents where all your products are. Make sure it includes every product URL and stays up to date.
Support guest checkout
Agents can't create accounts or verify phone numbers. If your checkout requires registration, agents can't buy from you.
Don't block bots aggressively
CAPTCHAs and bot-detection that blocks all automated traffic will also block AI shopping agents. Consider allowing agent traffic.
Make your pages load fast
Slow pages mean slow agents. Heavy JavaScript that takes seconds to render makes it harder for agents to read your product data.
06
What happens next?
Once your store is agent-ready, the process works automatically:
Your store gets scanned and scored
A skill file is generated β machine-readable instructions for agents
The skill is published to the catalog
AI agents discover your store through the catalog
Agents shop at your store and complete purchases
Agent feedback improves your rating over time
Are you a developer?
Read the full technical specification β frontmatter fields, body structure, validation rules, and the complete scoring methodology.
Read the agentic commerce standardIs your store ready?
Check your ASX Score and find out what AI agents see when they visit your store.
Check your ASX Score