ChatGPT Shopping: The Complete Guide for Merchants and Buyers (2026)

---

What is ChatGPT Shopping?

ChatGPT Shopping is the in-chat purchase experience inside ChatGPT, where a user can ask a question like "find me a soft cotton bedsheet set under $200" and the assistant returns specific product recommendations with live prices, reviews, and (in many flows) a Buy Now button that completes the purchase without redirecting to a third-party site.

OpenAI shipped the first public version of this in October 2025. It's built on a combination of:

• Web retrieval — ChatGPT searches the web in real time to find currently-available products
• Structured commerce data — merchants with machine-readable product catalogs are surfaced more reliably
• The Agentic Commerce Protocol (ACP) — OpenAI's open spec for letting merchants accept agent-driven purchases
• Partner integrations — direct catalog feeds from Shopify, Stripe, and a growing list of platforms

It is not an ad network. ChatGPT doesn't sell placement; recommendations are based on relevance, structured data quality, brand trust signals, and buyer-intent matching.

---

How ChatGPT Shopping works for buyers

From the buyer's side, the experience is simple:

1. User asks ChatGPT for a product (with or without constraints — budget, brand, attribute, etc.)
2. ChatGPT returns 3-5 specific products with images, prices, brief descriptions, and source links
3. User can ask follow-ups ("show me only ones under $50", "what about for sensitive skin", "compare these two")
4. For supported merchants, a Buy button completes the purchase in-chat using the buyer's saved payment method
5. For unsupported merchants, the buyer is sent to the merchant's site to complete checkout

The shift from "10 blue links" to "1 recommendation with a buy button" is the biggest UX change in commerce since the App Store. And the merchants on the inside of that recommendation are capturing share that historically went through Google search ads and Amazon.

---

How ChatGPT Shopping works for merchants

There are three tiers of merchant participation today:

Tier 1 — Visible (no integration)

You don't have to do anything. If ChatGPT's web retrieval finds you when answering a buyer's question, you might appear. But this is a low-probability and low-quality outcome:

• You're competing with thousands of other stores
• The AI agent doesn't know your current prices reliably
• You can't be purchased inside the chat (buyer redirects out and likely abandons)
• You appear only when your existing SEO is strong enough to rank in normal web retrieval

This is the default for ~99% of online merchants today. It's also the worst tier.

Tier 2 — Recommended (AEO-optimized)

You've added structured data, JSON-LD schema, an LLMs.txt file, machine-readable product feeds, and consistent entity facts across the web. ChatGPT can now identify your products with confidence, quote accurate prices, and surface you in relevant answers.

Buyers still have to click out to your site to complete the purchase — but you're being recommended in the chat, which is the discovery half of the win.

See our AEO Complete Guide for the full playbook.

Tier 3 — Transactable in-chat (full ChatGPT Shopping)

You've implemented Tier 2 and connected your catalog to ChatGPT's commerce flow via the Agentic Commerce Protocol (ACP) — or a partner platform that handles ACP on your behalf. Buyers can now complete the purchase inside the chat, using their ChatGPT-stored payment method. The agent passes the order to your store; your fulfillment runs as normal.

This is where the channel actually moves revenue. Stage 2 gets you recommended. Stage 3 closes the loop.

---

What's required to be transactable inside ChatGPT

To participate at Tier 3, a merchant needs:

1. A machine-readable product catalog with prices, stock status, variants, shipping policies, and return windows exposed in structured form (JSON-LD or feed-based)
2. Real-time inventory and pricing APIs — the agent has to verify "yes, this product is still $42.99 and in stock" before confirming the buyer's purchase
3. Agent-payment protocol support — at minimum ACP, ideally also UCP, MPP, card-network direct flows, and stablecoin rails so you're transactable regardless of which AI assistant or buyer wallet is used
4. Webhooks for order events so when an agent places an order, your fulfillment system gets the data the same way it would from a manual Shopify checkout
5. Refund and dispute handling consistent with the agent platform's policies
6. Brand and product entity data consistent across your site, social profiles, Wikipedia (if applicable), and review platforms — so the agent can confidently identify and recommend you

For Shopify, WooCommerce, BigCommerce, Magento, and Squarespace stores, xpay handles all six layers as a single integration — typically live in under 24 hours, no checkout rebuild, no replatforming.

---

Getting your store on ChatGPT — step by step

Step 1. Audit your current ChatGPT visibility

Open ChatGPT and ask 5-10 product questions a buyer in your category would naturally ask. Things like:

• "Best [your-category] brand for [common use case]"
• "Where to buy [specific product attribute] under $X"
• "Compare top brands of [your-category]"
• "Recommend a [your-category] for [buyer persona]"

Note which brands ChatGPT recommends. If yours doesn't appear in any of them — you're at Tier 1 visibility (or below).

Step 2. Run an AEO checklist on your site

Use our AEO technical checklist or our free AI shopping readiness audit. Fix the basics: JSON-LD schema on products, LLMs.txt, real-time stock/pricing feed, consistent entity facts.

This step alone usually takes you from Tier 1 to Tier 2 within 1-2 weeks.

Step 3. Connect to the agent-payment layer

This is where you go from being recommended to being purchasable in-chat. The options:

• OpenAI's direct ACP integration — apply at chat.openai.com merchant program (currently invitation/queue-based for most categories)
• A platform that handles ACP for you — services like xpay implement ACP + other agent payment rails as a managed layer, so you don't need to build it from scratch
• Direct via Stripe — Stripe has a one-line ACP integration if you're already a Stripe merchant on Shopify Plus

Step 4. Track agent-driven traffic and orders

Tag agent referrals in your analytics. ChatGPT, Perplexity, Claude, and Gemini all use identifiable user-agent strings. Segment your traffic, track conversion, and compare against your other channels.

Step 5. Iterate on what's working

The AI commerce channel is still evolving weekly. Track:

• Which buyer-intent prompts surface you vs. competitors
• Which product categories generate agent orders
• What ratio of agent-orders convert vs. abandon
• Refund/dispute rate compared to your other channels (it should be similar — if it's higher, the agent's pre-purchase information may be misaligned with your actual product)

---

What changed in October 2025 (and why now is the inflection)

OpenAI's official ChatGPT Shopping launch (October 2025) changed three things at once:

1. The Agentic Commerce Protocol (ACP) became real and open. Any merchant or platform can implement it. Before October 2025 there was no standard; now there is.
2. The buy button moved into the chat itself. Before, ChatGPT could recommend; now it can transact.
3. Search volume for "chatgpt shopping" started a sustained ramp. From ~720 monthly searches in mid-2025 to over 1,300 by April 2026 — and accelerating.

Concurrent moves:

• Anthropic's Claude is rolling out similar shopping capabilities (announced Q1 2026)
• Perplexity has had "Buy with Pro" since 2024, recently expanded
• Google's Gemini is integrating product results from Google Shopping more directly into agent flows

The pattern is unmistakable: every major AI assistant is becoming a commerce surface. The merchants who get instrumented for it in the next 6-12 months take the early share. After that, it's competitive — and the cost of catching up rises.

---

What ChatGPT looks for when picking a product to recommend

This is the core question for any merchant. Based on observed behavior + published guidance from OpenAI:

1. Specificity of structured data — clear price, in-stock status, brand, variants
2. Brand entity consistency — your brand should be identifiable as the same entity across web, social, review sites
3. Recency of inventory data — agents avoid recommending products that may be out of stock
4. Review density and recency — products with many recent reviews from real users carry more confidence
5. Match to buyer constraints — if the buyer specified "under $50," only products that clearly meet that constraint are recommended
6. Source diversity — agents prefer recommending brands cited across multiple independent sources (your site + a press mention + a review site beats your site alone)
7. Purchase friction — agents are starting to weight transactable-in-chat options higher than out-redirect options (the user experience is meaningfully better)

These factors compound. A brand at the 90th percentile across all 7 wins almost every recommendation. A brand at the 50th percentile across all 7 wins almost none.

---

Common merchant mistakes

1. Treating ChatGPT Shopping as a future bet. It's already moving money in apparel, beauty, supplements, food, and electronics. Waiting until "the channel is proven" means waiting until your competitors have entrenched.
2. Doing AEO without payment integration. You optimize to be recommended, the buyer asks ChatGPT to buy, the chat redirects them to your site, they abandon at checkout. Half a loop is worse than none — you trained the agent that your brand doesn't convert.
3. Not maintaining real-time data. Stale prices and stockouts make agents stop recommending you within days. The agent doesn't get a second chance; once your accuracy drops, you're dropped.
4. Astroturfing reviews. Agents detect and penalize inauthentic review signals more aggressively than Google ever did. Don't.
5. Trying to game keyword/prompt matching. Stuff your product titles with "AI-recommended" or "ChatGPT favorite" and you get filtered out for low-trust signals. Sound natural; be specific; let the structured data speak.

---

What's coming in the next 12 months

• Multi-merchant cart support — ChatGPT will let buyers add products from multiple merchants in a single agent-mediated transaction
• Subscription/recurring purchases — agent-managed subscriptions for replenishable products (food, beauty, supplements)
• Bundled comparison purchases — "buy the cheaper one from merchant A and the better-rated one from merchant B in the same transaction"
• Buyer-side agents shopping for you — autonomous agents acting on a buyer's saved preferences without explicit prompts
• Cross-engine portability — same buyer preferences and payment method across ChatGPT, Claude, Perplexity, and Gemini

Each expansion makes the merchant-side optimization more valuable, not less. The work you do for AEO + agent-payment now pays off across every emerging agentic-commerce surface.

---

Get your store on ChatGPT in 24 hours

If you run a Shopify, WooCommerce, BigCommerce, Magento, or Squarespace store, xpay handles the full ChatGPT Shopping integration end-to-end:

• AEO-grade structured data on your catalog
• LLMs.txt + product feed publishing
• ACP integration so buyers can complete the purchase in-chat
• Multi-rail payment (cards, ACP, UCP, stablecoin) for cross-engine compatibility
• Real-time inventory + pricing sync
• Order webhooks into your existing fulfillment

No replatforming, no checkout rebuild, native plug-in install. Free to install. Free until your first AI-driven sale.

→ Get started: xpay.sh/merchants → Free AI shopping readiness audit → Platform-specific guides: Shopify · WooCommerce · BigCommerce · Magento · Squarespace

---

Frequently asked questions

Related reading

• AEO: The 2026 Playbook for Showing Up in ChatGPT, Claude, Perplexity, and Gemini
• The 2026 Merchant's Playbook for Agentic Commerce
• Agentic Storefront for Shopify: First Principles + Setup Guide

---

Last updated 2026-05-13.