ChatGPT Shopping, Perplexity Buy, and Google AI Overviews are rewriting the path-to-purchase. Product schema — not your product page — is becoming the primary conversion surface.
For 25 years, the product detail page (PDP) was the centerpiece of e-commerce. You optimized it for conversion rate, A/B tested button colors, and obsessed over image angles. Agentic commerce — AI assistants that research, compare, and purchase on behalf of users — is making the PDP less relevant, not more. Here is what is replacing it.
Agentic commerce describes transactions where an AI agent — ChatGPT Shopping, Perplexity, Google AI Overviews Shopping, Amazon Rufus — mediates the purchase decision. The user says "find me the best espresso machine under $500 with a stainless steel boiler." The AI queries structured data across multiple merchants, compares options, and presents a recommendation. Sometimes it completes the purchase. Sometimes it sends the user to a checkout link. Either way, the user never visits your product page.
McKinsey projects agentic commerce will reach $3-5 trillion by 2030. Adobe Analytics reported AI referral traffic grew 4,700% year-over-year in 2025 (from a very small 2024 baseline — absolute volumes remain modest compared to organic search, but the growth rate signals a clear trajectory). BigCommerce found AI-referred shoppers convert 31% better than organic search visitors. These are not marginal numbers.
| Era | Primary Surface | What You Optimize |
|---|---|---|
| Web 1.0 (1995-2010) | Product page | HTML, images, copy |
| Mobile (2010-2020) | Mobile product page | Page speed, mobile UX, app deep links |
| Marketplace (2015-2025) | Amazon/Shopify listing | Marketplace SEO, reviews, Buy Box |
| Agentic (2024-2030) | Structured data feed | Schema, GTIN, attributes, data consistency |
The product page still exists. Humans still visit it — especially for high-consideration purchases. But the first touchpoint is increasingly the AI assistant, not your website. And the AI assistant does not read your product page. It reads your structured data.
We analyzed the crawling behavior of ChatGPT Shopping, Perplexity, and Google AI Overviews across 50 WooCommerce stores over 90 days. The pattern is consistent:
| Signal | Default WooCommerce | What AI Assistants Need | Gap |
|---|---|---|---|
| Product schema | None or theme-dependent | JSON-LD: Product + Offer + AggregateRating + Brand | Critical |
| GTIN/EAN/UPC | Not collected | Required for cross-platform identity matching | Critical |
| MPN | Not collected | Required for brand-specific queries | High |
| Brand entity | Text field only | Schema.org Brand type with name + URL | High |
| Condition | Not collected | Offer.itemCondition (New/Used/Refurbished) | Medium |
| Variable pricing | Range only | AggregateOffer (lowPrice, highPrice, offerCount) | Medium |
| Reviews | Theme-dependent | Review schema with author, datePublished, rating | High |
| Data consistency | Manual | Price/stock must match across site + feeds + marketplaces | Critical |
The old funnel: Search → Click → PDP → Cart → Checkout. The agentic funnel: AI query → Schema parsing → Constraint matching → Consensus check → Recommendation (with or without PDP visit) → Purchase. Your product page is now step 5 of 6 — and only for high-consideration items. For commodity products (phone chargers, cleaning supplies, basic apparel), the AI may complete the purchase without the user ever seeing your brand.
This is not a threat. It is an opportunity. Stores with complete structured data get into the AI consideration set for free. Stores without it are invisible — not because they ranked poorly, but because the AI never knew they existed.