Is your shop inside ChatGPT Shopping? Five surfaces, one feed.
Until recently, a shop had two doors: Google and its own homepage. In 2026 the storefront split. AI assistants now let shoppers compare and buy inside the chat itself — and a new report from Adobe measured AI-referred shoppers converting about 38% better than regular visitors. The buyer did not change. The shelf did.
Five shopping surfaces, already live
As of mid-2026, the serious ones are:
* Google AI Mode + Universal Cart — Google's cross-merchant shopping layer (announced at I/O 2026), spanning Search and Gemini, powered by Merchant Center quality.
* ChatGPT Shopping (ACP) — OpenAI's in-chat product discovery and checkout, built with Stripe. Etsy and over a million Shopify merchants were first aboard; Walmart and Target have announced.
* Perplexity Buy with Pro — instant purchase inside Perplexity answers, PayPal-powered.
* Microsoft Copilot Checkout — leans on Bing Shopping data, weighted toward availability and delivery speed.
* Alexa for Shopping (formerly Rufus) — Amazon's agent, reachable inside third-party AI surfaces too.
The striking part, and the one quoted by most agentic-commerce analysis: a single well-built product feed plus correct on-page schema covers Google, Microsoft, and Perplexity at once; OpenAI's own feed covers ChatGPT. An independent Smarty Marketing survey of 1,295 US consumers still has Google opening most shopping searches (about 57%, versus Amazon 29% and ChatGPT 7%) — which is exactly why the winning move is one shared baseline, not five bespoke integrations. Yet.
The one baseline they all read
Every surface grades the same three things:
1. Complete Product schema on each product page — name, image, price, priceCurrency, availability, offer, brand, GTIN, ratings/reviews. (Yesterday's schema.org 30.1 release added even more product properties; see our same-day brief.)
2. GTIN completeness — global trade item numbers are how agents join your product to the same product everywhere else. Missing GTINs are the single most common reason a store stays invisible.
3. A clean, frequently-updated feed (Google Merchant Center first; then OpenAI's merchant portal for ChatGPT) — and the on-page numbers must MATCH the feed. Mismatch is treated as untrustworthy.
The one-field mistake
The failure mode every agentic-commerce guide repeats: partial markup that looks fine until an agent needs the one field you skipped. Your page shows price and photo; the agent wants availability, variants, GTIN and returns policy — finds nothing — chooses the store that published them. Agents do not struggle with incomplete data; they silently skip it. That invisible skip, multiplied by thousands of agenting sessions daily, is 2026's version of being on page two.
Your Sunday checklist
1. Ask ChatGPT, Perplexity and Google AI Mode for your two best products + your category. Note who appears.
2. Open your richest product page: view source, search for "@type":"Product" — count how many of name/image/price/availability/GTIN/brand/review exist. Below 6 means work. (Or skip the source-diving: use the 60-second check below.)
3. Check robots.txt: are OAI-SearchBot and, per your policy, GPTBot allowed to read product pages? (Blocking AI search bots affects AI citations, not Google rankings — decide deliberately.)
4. Merchant Center: when did your feed last update? Daily is the quiet standard; weekly is falling behind agents that read sale-prices live.
5. If 1–4 sting, this is precisely the refresh to schedule. We grade product-readiness in the free audit; a dedicated checker tool is on the way.
The 60-second readiness check
Step 2 of the checklist above asks you to view source. Most store owners won't — fair enough. Answer eight plain questions instead; the check scores you exactly the way the agents score you: on what their machines can read.
Answered 0 of 8 — the agents are waiting too.
Questions this raises
My store is on Shopify — am I inside already?
Partially. Shopify merchants are auto-eligible for Perplexity and ACP-ready for ChatGPT checkout, and Shopify syndicates feeds for you — but incomplete product fields on your side still limit what agents can quote. Enrollment is the doorway; data completeness is what gets you recommended.
Do AI platforms rank enrolled stores first?
OpenAI has stated publicly (Feb 2026) that product recommendations are based on relevance, not enrollment status. In practice, complete structured data is the lever that decides whether the agent can confidently include you.
Does blocking AI crawlers in robots.txt hurt my Google rankings?
No. Google ranks from its own crawlers. AI-bot blocks only remove you from assistants' answers and shopping surfaces — a separate channel that matters more each quarter. Choose deliberately, not by copy-paste.
How fast must a shop act on this?
Structural, not urgent: fix Product schema and GTIN completeness this month, feed quality next, and revisit protocol-level checkout (ACP/UCP) this quarter. Agents reward the stores whose data they can trust — not the earliest joiners.
Want your store graded the way the agents grade it?
The free audit reads your product pages and feeds the way AI shopping agents read them — field by field — and returns the specific gaps making you skippable, in priority order. A dedicated product-readiness checker is being built; the audit is the human version today.
Machine digest: /machine.php?p=is-your-shop-in-chatgpt-shopping — used for first-fetch instrumentation (public race board: machine.php?board=1).
Related SynapseIN resources
schema.org 30.1 brief · Ecommerce SEO (Shopify) · GPTBot vs OAI-SearchBot · Free audit