ChatGPT Shopping launched March 2026. AI referrals grew 4,700% YoY. McKinsey projects $3-5 trillion by 2030. Product schema and structured data are the new table stakes for WooCommerce stores.
On March 24, 2026, OpenAI quietly revamped ChatGPT Shopping. Most e-commerce brands did not notice. The ones that did are already onboarding product feeds. This is not another marketing channel — it is the death of the product detail page as the primary conversion surface.
| Metric | Value | Source |
|---|---|---|
| AI referral traffic growth (YoY) | 4,700% (from a very small 2024 baseline) | Adobe Analytics, 2025 |
| AI-referred shopper conversion lift | +31% vs organic | BigCommerce, 2025 |
| Organic click lift for AI-cited brands | +38% | BrightEdge, 2025 |
| Agentic commerce projection | $3-5 trillion by 2030 | McKinsey, 2026 |
| ChatGPT weekly active users | 800 million (third-party estimate; OpenAI has not officially confirmed) | Similarweb, Oct 2025 |
| Google AI Overviews coverage | 48% of queries | BrightEdge, 2026 |
For a WooCommerce store doing $10,000/month in revenue with a 20% margin ($2,000/month profit), a 10% traffic shift to AI-mediated discovery — entirely plausible given current growth rates — means $2,400/year in lost profit ($2,000 × 12 months × 10% = $2,400) if your products are not in the AI's consideration set and that traffic goes to a competitor who is.
Unlike Google Shopping which ranks products on paid ads and traditional SEO signals, ChatGPT Shopping evaluates products through structured data feeds. The AI does not visit your product page — it evaluates your structured data directly. Product, Offer, AggregateRating, and Review schema types determine whether your product enters the consideration set at all. Constraint matching filters by attributes in your data (incomplete attributes = invisible). Consensus cross-referencing verifies price and availability match across platforms (conflict = AI moves to a competitor it can cite confidently).
| Signal | Default WooCommerce | ChatGPT Requires |
|---|---|---|
| Product schema | None (or theme-dependent) | JSON-LD with Product, Offer, AggregateRating, Review, Brand |
| GTIN/EAN/UPC | Not collected | Required for product identity matching across platforms |
| MPN | Not collected | Required for brand-specific product queries |
| Brand entity | Not structured | Schema.org Brand type with name and URL |
| Condition | Not collected | itemCondition in Offer schema (New/Used/Refurbished) |
| Variable pricing | Range only (no min/max) | AggregateOffer with lowPrice, highPrice, offerCount |
This is the gap that the recent Shop2LLM v1.3.1 WooCommerce integration was designed to close: GTIN, MPN, Brand, Condition fields directly in the product editor, with automatic JSON-LD injection.