The State of AI Commerce Readiness — Feb 2025

AI shopping agents are no longer a future concept — they're live. Google AI Mode and Gemini are shopping for users. ChatGPT has a full shopping experience. Perplexity surfaces products with buy buttons. But are merchants ready?

We set out to answer that question by auditing 30+ major brands across three protocols: Google's Universal Commerce Protocol (UCP), ChatGPT's product feed spec (ACP), and Schema.org structured data.

Key Findings

The results are sobering. The average AI readiness score across all brands is 47 out of 100. Even the top performers — Stance (70), Bombas (69), Glossier (69) — barely crack 70. Household names like Nike, Amazon, and Target score surprisingly low.

Schema.org Is the Strongest Signal Today

Brands built on Shopify tend to score well here, thanks to native JSON-LD support for Product markup. But almost everyone is missing the fields that AI agents weight most heavily when deciding which products to recommend:

• GTINs (Global Trade Item Numbers) — Without these, ChatGPT can't confidently match your product to a user's query. Only ~30% of the brands we scanned include them.
• Aggregate ratings and review counts — AI agents use these as a trust signal. Missing them means your products look unvetted.
• Q&A markup — Almost nobody has this. It's a massive missed opportunity for long-tail queries like "is this shoe good for flat feet?"
• Multiple high-quality images — AI agents prefer products with rich visual data. Many brands serve only a single image in their structured data.

Google UCP Is Virtually Non-Existent

Launched in late 2024, the Universal Commerce Protocol is how Google AI Mode and Gemini will handle product discovery and checkout. It works through a /.well-known/ucp.json manifest that declares what services a merchant supports — product search, cart management, checkout, returns.

Out of 30+ sites we scanned, only a handful even have a UCP manifest — and none are fully configured with checkout capabilities. This is the biggest untapped opportunity in e-commerce right now. The first merchants to implement UCP properly will have a massive advantage when Google AI Mode goes mainstream.

ChatGPT Shopping Readiness Is Mixed

Most brands have basic product data, but feed completeness is low. OpenAI's shopping experience pulls from product feeds and needs specific fields to confidently recommend products:

• Product titles and descriptions — Must be clear, specific, and include key attributes (size, color, material)
• GTINs — Required for product matching and deduplication
• Star ratings — ChatGPT uses these to rank competing products
• Availability and pricing — Must be current and accurate
• Product images — Multiple angles, high resolution

The Scoring Methodology

Our overall readiness score weights three protocols:

• Schema.org (45%) — The most universal signal across all AI agents today
• ACP / ChatGPT Feed (40%) — Live and high-impact for shopping queries
• Google UCP (15%) — Weighted lower because the protocol just launched and adoption is nascent

Each protocol is scored individually by checking for specific capabilities: manifest presence, field completeness, data quality, and protocol compliance.

Brand Benchmarks

Here's how some well-known brands performed:

Brand	Overall	UCP	Schema	ACP
Stance	70	25	84	72
Bombas	69	0	85	76
Glossier	69	25	81	72
Allbirds	62	25	82	53
On Running	61	25	81	53
Brooklinen	61	25	63	72
Warby Parker	58	0	81	53
Gymshark	44	25	69	23

What This Means for Merchants

The brands that invest in AI agent readiness now will have a massive first-mover advantage. The infrastructure is being built — Google, OpenAI, and Perplexity are actively developing commerce capabilities — but almost nobody is ready for it yet.

The good news? Most of these fixes are straightforward:

1. Add GTINs to your product data — This single change dramatically improves your ChatGPT and Perplexity visibility
2. Implement complete Schema.org Product markup — Include offers, ratings, images, and identifiers
3. Create a UCP manifest — Even a basic one puts you ahead of 95% of merchants
4. Enrich your product descriptions — AI agents match queries to products based on natural language, not keywords

The window of opportunity is narrow. As AI shopping becomes mainstream — and it's happening faster than anyone expected — the brands that are ready will capture disproportionate market share.

Want to know where your store stands? Run a free scan at AgentCart and get your AI readiness score in 30 seconds.