OpenAI announced today that ChatGPT is now a visual shopping platform, powered by an expanded Agentic Commerce Protocol (ACP) that connects merchants directly to 300+ million users. Target, Sephora, Nordstrom, Lowe's, Best Buy, Home Depot, Wayfair, and Walmart have already integrated. For retailers, this is a new customer acquisition channel with higher-intent shoppers. For enterprise tech leaders, it's a live case study in how agentic AI protocols are becoming infrastructure.
This isn't about adding chatbots to your website. It's about plugging your catalog into a conversational AI layer that sits between Google search and your checkout page — and capturing customers before they hit your competitors' sites.
What Changed: Visual Shopping Meets Conversational AI
OpenAI rolled out three major updates to ChatGPT's shopping experience:
Visual product browsing: Users can now see product images, compare options side-by-side, and upload photos to find similar items. Instead of text-only recommendations, ChatGPT displays rich product cards with prices, reviews, and key features.
Agentic Commerce Protocol expansion: ACP now handles product discovery, not just checkout. Merchants share product feeds, promotions, and inventory data through ACP, and ChatGPT surfaces relevant products in real-time conversations. Third-party providers like Salesforce and Stripe are supported, so merchants can integrate using existing systems.
Merchant checkout control: OpenAI discontinued its Instant Checkout feature and now redirects users to merchant websites for payment. Walmart went a step further, building an in-ChatGPT app experience with account linking, loyalty rewards, and Walmart-branded checkout.
Why This Matters for CMOs and CROs
ChatGPT users aren't browsing casually — they're asking "Which laptop should I buy for video editing under $1,500?" That's higher purchase intent than SEO traffic or social media clicks. If your catalog isn't in ACP, those conversations send users to competitors who are integrated.
Photo by Tima Miroshnichenko on Pexels
How Merchants Integrate: Three Paths to ACP
OpenAI built ACP to work with existing e-commerce infrastructure. Here are the integration options:
For Shopify merchants: Zero lift. Shopify automatically syncs product catalogs to ChatGPT via the Shopify Catalog integration. Millions of merchants are already live.
For Salesforce Commerce Cloud or Stripe users: Connect your existing product feed through your platform's ACP adapter. Most implementations take 1-2 weeks, depending on catalog size and data quality.
For custom platforms: Build a direct ACP connection by exposing your product catalog as a structured feed (JSON or XML). OpenAI provides API documentation and a testing sandbox. Expect 4-6 weeks for initial integration, including feed validation and QA.
The ROI Story: Higher-Intent Shoppers, Lower Acquisition Costs
Why should retailers care about ChatGPT as a sales channel? Because the customers arriving from ChatGPT have already done their research, compared options, and narrowed their choices. They're not browsing — they're deciding.
Reduced decision time (hours → seconds): Traditional online shopping means opening 10+ tabs, reading "best of" listicles, and comparing spreadsheets. ChatGPT collapses that process into a single conversation. Users describe what they need, refine criteria conversationally, and see side-by-side comparisons instantly.
Higher purchase intent: A user who asks "Which standing desk under $800 supports dual monitors?" is closer to buying than someone who Googles "best standing desks." ChatGPT conversations are goal-oriented, not exploratory.
Lower acquisition costs: If ChatGPT drives conversions at rates comparable to paid search but without CPC costs, the CAC math changes significantly. Early merchants report that ACP-driven traffic behaves more like retargeting campaigns than cold acquisition.
What CTOs Need to Know About ACP
ACP is a structured product feed protocol, not a proprietary API. That means:
- Your catalog stays on your infrastructure (OpenAI doesn't host products)
- Real-time inventory updates prevent overselling
- You control pricing, promotions, and availability
- Integration works with existing CDNs and image optimization
Walmart's In-ChatGPT App: The Premium Integration Play
While most merchants redirect users to external checkout, Walmart built a native ChatGPT app experience. Users can link their Walmart accounts, view personalized recommendations based on purchase history, and earn loyalty points — all without leaving ChatGPT.
This is the premium tier of ACP integration: full account linking, loyalty program access, and a branded environment inside ChatGPT. It requires deeper engineering investment (estimated 8-12 weeks for custom app development), but it keeps users in the conversational flow instead of redirecting them mid-purchase.
For retailers with strong loyalty programs and high repeat purchase rates, native app integration makes sense. For everyone else, redirecting to merchant checkout after product selection is the faster, lower-risk path.
The Strategic Shift: OpenAI Exits Checkout, Focuses on Discovery
OpenAI's decision to kill Instant Checkout is significant. It signals that OpenAI sees its value in product discovery, not payments processing. Merchants wanted control over checkout (to preserve branding, upsell opportunities, and loyalty programs), and OpenAI listened.
This is a more sustainable business model for OpenAI. Instead of competing with Stripe and Shopify on payments, it becomes the layer between search and checkout — capturing high-intent users and handing them to merchants ready to buy.
For merchants, it means you keep the customer relationship. You control the checkout experience, capture payment data, and own the post-purchase engagement. OpenAI gets you the lead; you close the sale.
What Questions Should Your Team Be Asking?
If you're evaluating ACP integration, here are the key questions to answer internally:
Is our target customer using ChatGPT for product research? B2C retailers selling consumer electronics, home goods, and fashion are seeing the most traction. B2B merchants selling complex enterprise products may not see meaningful traffic yet.
Can our product catalog integrate cleanly with ACP? If you have a clean product feed (structured SKU data, accurate pricing, up-to-date inventory), integration is straightforward. If your catalog data is messy, expect delays.
Do we have the engineering resources to build and maintain an ACP integration? Shopify merchants get zero-setup integration. Everyone else needs 1-12 weeks of engineering time, depending on complexity.
What's the opportunity cost of not integrating? If your competitors are in ACP and you're not, users asking "Which [your product category] should I buy?" will see competitor products, not yours.
Decision Guide: Should You Integrate with ACP?
Integrate now if:
- You're on Shopify (zero setup required)
- You sell consumer products in competitive categories (electronics, fashion, home goods)
- Your competitors are already integrated
- You have clean product data and structured inventory feeds
Wait and watch if:
- You sell complex B2B products requiring sales consultations
- Your catalog data is messy (outdated SKUs, inconsistent pricing)
- You don't have engineering bandwidth for integration + maintenance
- Your target customer isn't using ChatGPT for product research yet
Want to calculate your own AI ROI? Try our AI ROI Calculator — takes 60 seconds and shows projected savings, payback period, and 3-year ROI.
Continue Reading
Related articles on agentic AI and enterprise automation:
- Oracle's Free Agentic Platform: No-Code AI Agents Built Into Your ERP — Oracle launched a free no-code agent builder for Fusion customers, directly competing with Salesforce and Microsoft.
- [Cursor's Hidden Chinese Model: A Wake-Up Call for AI Vendor Due Diligence](/article/cursor-chinese-model-vendor-transparency-crisis-2026) — Cursor defaulted to a Chinese AI model without disclosure, exposing gaps in enterprise vendor risk management.
- LangGraph vs Google ADK: Which Enterprise AI Framework Should You Choose? — Side-by-side comparison of LangGraph and Google Agent Development Kit for building enterprise AI agents.
Thoughts on ACP integration for your business? Connect with me on LinkedIn, Twitter/X, or via the contact form.
Photo by