The Complete Merchant's Guide to Every AI Shopping Platform in 2026

The AI Shopping Landscape in 2026

Twelve months ago, AI-assisted shopping was a novelty. Today it is a channel. According to Similarweb, AI-driven traffic to retail sites spiked 357% year-over-year, reaching 1.13 billion visits. Adobe reported that AI-referred e-commerce traffic surged 805% on Black Friday alone. Morgan Stanley projects that half of all online shoppers will use AI agents by 2030.

The opportunity is massive. McKinsey estimates the total impact of AI on global commerce at $3 to $5 trillion by 2030. Yet according to Liquid Web, 65% of retailers are taking no preparatory steps for AI commerce. That gap between the scale of the opportunity and the level of merchant preparedness is where this guide fits in.

Five platforms now dominate the AI shopping landscape. Each has a different approach to product discovery, different integration requirements, and different fee structures. Some enable native checkout. Others drive traffic to your store. Some are open to all merchants. Others favor specific ecosystems.

Understanding each platform is no longer optional. It is a competitive requirement. This guide covers every platform that matters, how each one works, and exactly what merchants need to do to participate -- starting with the two biggest players.

ChatGPT Shopping (OpenAI)

ChatGPT Shopping is OpenAI's commerce layer built directly into the ChatGPT interface. When a user asks for product recommendations, ChatGPT surfaces products with rich cards, images, pricing, and -- for participating merchants -- a native buy button that completes the purchase without leaving the conversation.

How It Works

ChatGPT discovers products through two mechanisms. First, OAI-SearchBot crawls the web and indexes product pages, reviews, and brand content. This crawled data informs ChatGPT's organic product recommendations. Second, structured product data from Google Merchant Center feeds and schema.org markup provides precise details like pricing, availability, and product identifiers.

When a merchant has enabled Instant Checkout, ChatGPT displays a buy button directly in the chat. The user clicks it, confirms payment through their Stripe-stored method, and the transaction completes. The entire flow is powered by the Agentic Commerce Protocol (ACP), a RESTful API standard built by OpenAI and Stripe.

Merchant Program and Fees

ChatGPT charges a 4% transaction fee on all purchases completed through Instant Checkout. This is on top of standard Stripe processing fees (typically 2.9% + $0.30). There is no upfront cost or monthly subscription. Merchants remain the merchant of record -- customer relationships, fulfillment, and support stay with you.

For organic recommendations where the user clicks through to your website to complete a purchase, there is no fee. ChatGPT functions like a referral channel in that scenario.

Product Requirements

• Complete Product schema with JSON-LD on every product page
• GTINs (UPC or EAN codes) for reliable product identification
• Accurate pricing and availability data
• OAI-SearchBot allowed in your robots.txt
• XML sitemap including all product URLs
• For Instant Checkout: Stripe payment processing with ACP enabled

How Non-Shopify Merchants Get Listed

Shopify stores have automatic integration through Shop Pay. For WooCommerce, BigCommerce, and Magento merchants, the process is manual but straightforward. Allow OAI-SearchBot, implement complete structured data, submit to Google Merchant Center, and enable Stripe ACP for Instant Checkout. Our full walkthrough is in the ChatGPT Shopping guide.

Pros and Cons

Pros: Hundreds of millions of active users. Native in-chat checkout reduces friction dramatically. AI-referred traffic converts at significantly higher rates than organic search. Growing rapidly as OpenAI expands commerce features.

Cons: 4% transaction fee adds up, especially alongside Stripe processing fees. Currently Stripe-only for checkout. Merchant control over how products are presented is limited. Discovery algorithm is opaque compared to Google's more established search signals.

Google AI Mode

Google AI Mode is an AI-powered layer within Google Search that uses Gemini to understand conversational shopping queries, surface product recommendations with explanations, and enable native checkout without leaving Google. It launched in early 2026 alongside the Universal Commerce Protocol (UCP).

How It Works

When a user enters a shopping-related query in Google Search, AI Mode activates and generates a curated response. Rather than showing a list of blue links, it presents product recommendations with comparisons, feature breakdowns, and explanations of why each product fits the query. Product data comes from Google Merchant Center feeds, on-page schema.org markup, and the broader Google Shopping index.

For merchants who have implemented UCP, AI Mode can complete purchases natively using Google Pay. The user sees a checkout interface within Google, powered by the merchant's store backend through UCP's standardized API endpoints.

GMC Integration and UCP

Google Merchant Center is the primary data pipeline for AI Mode product discovery. Your product feed tells AI Mode what you sell, at what price, and whether it is in stock. Optimizing your product feed with complete attributes -- GTINs, product highlights, detailed specifications, and return policy labels -- directly improves your chances of being surfaced.

UCP adds the transaction layer. Published at /.well-known/ucp on your domain, it defines your store's capabilities and lets AI agents complete purchases through your API. UCP supports REST, Model Context Protocol (MCP), and Agent-to-Agent (A2A) communication. Shopify stores get native UCP support. For everyone else, implementation requires development work or waiting for platform-level integrations.

Fees

Google AI Mode currently charges 0% transaction fees on native checkout. You pay your normal payment processor fees (Google Pay, Stripe, etc.), but Google takes no commission. Google has not announced when or if this will change, but the zero-fee positioning is a clear competitive move against ChatGPT's 4%.

How Non-Shopify Merchants Get Listed

Products can appear in AI Mode results through Google Merchant Center and schema.org markup without UCP. The full native checkout experience requires UCP implementation. Non-Shopify merchants should optimize their GMC feed, ensure complete schema markup, enable guest checkout and Google Pay, and prepare for UCP integration. See our detailed Google AI Mode guide.

Pros and Cons

Pros: Largest search audience in the world. 0% transaction fees currently. Leverages your existing GMC investment. UCP is open-source (Apache 2.0) with broad industry backing from over 20 partners including Visa, Mastercard, and PayPal.

Cons: UCP implementation is more complex than ACP. Non-Shopify native support is still rolling out. Google's AI Mode is still evolving and the product recommendation algorithm is not yet as transparent as traditional Shopping ads. Native checkout requires additional infrastructure (Google Pay, guest checkout).

Perplexity Shopping

Perplexity is an AI-powered answer engine that has expanded into product discovery and purchasing. Its shopping capabilities let users ask natural language questions about products and receive curated, cited recommendations with purchase options.

How It Works

Perplexity's approach is crawl-based. It indexes the web broadly, pulling product information from merchant sites, review platforms, comparison tools, and editorial content. When a user asks a shopping question, Perplexity synthesizes information from multiple sources and presents product recommendations with citations linking back to original sources.

Buy with Pro is Perplexity's checkout feature available to Pro subscribers. It enables one-click purchasing for select products directly within the Perplexity interface. The merchant integration model is still evolving, but currently relies on a combination of affiliate partnerships and direct merchant relationships.

Merchant Integration

Unlike ChatGPT or Google AI Mode, Perplexity does not have a formal merchant onboarding program with a defined protocol. Product discovery happens through web crawling. The best way to get your products surfaced is to have a strong web presence with complete structured data, quality product content, and positive third-party mentions.

Perplexity respects robots.txt directives. Ensure your product pages are crawlable and that your structured data is complete. Products with GTINs, rich descriptions, and strong review signals are more likely to be surfaced.

Pros and Cons

Pros: Growing user base of research-oriented shoppers who tend to have high purchase intent. Citation-based model means your brand gets visible attribution. Low barrier to entry -- no special integration required beyond good web fundamentals.

Cons: Smaller audience compared to ChatGPT and Google. Buy with Pro is limited to Pro subscribers. No formal merchant program means less control and predictability. Revenue model and fee structure are still evolving.

Amazon Rufus

Amazon Rufus is Amazon's AI shopping assistant, integrated directly into the Amazon app and website. It helps customers discover, compare, and evaluate products using conversational AI. Rufus operates entirely within the Amazon ecosystem.

How It Works

Rufus is trained on Amazon's product catalog, customer reviews, Q&A sections, and web content. When a user asks Rufus a question like "what are the best noise-canceling headphones for commuting," it synthesizes information from Amazon listings, reviews, and external sources to provide recommendations. All recommended products link directly to Amazon product pages.

Rufus can answer questions about product comparisons, suitability for specific use cases, product features, and even category-level guidance. It operates as a conversational layer on top of Amazon's existing search and recommendation infrastructure.

What It Means for Non-Amazon Merchants

If you do not sell on Amazon, Rufus will not directly recommend your products. However, Rufus has an indirect but significant impact on off-Amazon merchants.

First, Rufus is training consumers to expect AI-assisted shopping. As users get comfortable asking an AI for product recommendations on Amazon, they will expect the same experience everywhere. Merchants who are ready for AI discovery on Google and ChatGPT benefit from this behavioral shift.

Second, Rufus raises the bar for product content quality. Amazon listings that surface well in Rufus have detailed descriptions, comprehensive Q&A sections, and strong review profiles. The same content discipline applies to your own product pages if you want to perform well on other AI platforms.

Third, Rufus represents competition for your customers' attention. If a customer uses Rufus to find a product that you also sell on your own store, Amazon captures that sale. Being visible on non-Amazon AI platforms like ChatGPT and Google AI Mode is your counterbalance.

Pros and Cons

Pros: Massive reach within the Amazon ecosystem. Highly refined product data from years of structured catalog management. Strong consumer trust in Amazon recommendations.

Cons: Closed ecosystem -- only Amazon products are recommended. No direct integration path for non-Amazon merchants. Increases competitive pressure on independent merchants by making Amazon shopping more convenient.

Microsoft Copilot Shopping

Microsoft Copilot integrates shopping capabilities through Bing's product index. When users ask Copilot shopping-related questions, it surfaces product cards with images, pricing, and merchant links powered by the Bing Shopping feed.

How It Works

Copilot draws from the Bing Shopping index, which is fed by the Microsoft Merchant Center (formerly Bing Merchant Center). When a user asks a product question, Copilot generates a conversational response with embedded product cards. These cards include product images, prices, ratings, and direct links to the merchant's website.

Unlike ChatGPT or Google AI Mode, Copilot does not currently offer native checkout. It functions as a product discovery and referral tool, driving traffic to your store where the purchase is completed. This means there are no transaction fees from Microsoft -- you only pay your standard payment processing costs.

How to Optimize

Submit your product feed to Microsoft Merchant Center. The feed format is compatible with Google Merchant Center, so if you already have a GMC feed, you can often import it directly. Ensure your Bing Webmaster Tools account is verified and your sitemap is submitted.

On-page optimization matters too. Complete structured data, strong product page content, and good technical SEO all improve your visibility in Bing's index, which directly feeds Copilot's product recommendations.

Pros and Cons

Pros: Zero transaction fees. Easy integration if you already use GMC (feed is compatible). Growing user base through Windows, Edge, and Microsoft 365 integration. Low effort to add alongside Google AI Mode optimization.

Cons: Smaller search market share than Google. No native checkout reduces conversion compared to platforms with in-chat purchasing. Less advanced product understanding compared to ChatGPT or Google AI Mode.

Platform Comparison Table

Cross-Platform Strategy: Optimize Once, Benefit Everywhere

The good news for merchants is that the foundational work for AI shopping readiness is the same across every platform. Rather than building five separate strategies, you build one strong foundation and add platform-specific layers on top.

Unified Product Feed

Start with your Google Merchant Center feed. Make it as complete and accurate as possible. Include optional attributes like product highlights, detailed specifications, shipping weights, and return policy labels. This single feed serves Google AI Mode directly, informs ChatGPT through its data sources, feeds into Bing Merchant Center for Copilot, and provides the structured data that Perplexity crawls.

Import your GMC feed into Microsoft Merchant Center with minimal modifications. The formats are largely compatible. One feed, two platforms. See our product feed optimization guide for the complete attribute checklist.

Schema Markup as Universal Language

Schema.org Product markup is the universal language that every AI shopping platform understands. Complete JSON-LD structured data on every product page benefits ChatGPT (OAI-SearchBot parses it), Google AI Mode (it is a primary ranking signal), Perplexity (it crawls and indexes it), and Copilot (Bing uses it for product understanding).

The minimum viable schema includes Product type with name, description, image, brand, SKU, GTIN, offers (price, currency, availability), and aggregateRating. Add BuyAction markup for platforms that support transactional schema. Every additional field you add improves your visibility across all platforms simultaneously.

Protocol Readiness

The two protocols that matter are ACP (for ChatGPT) and UCP (for Google AI Mode). Implementing both gives you native checkout on the two largest AI shopping platforms. Our UCP vs ACP comparison guide covers the decision framework for which to implement first.

For Perplexity and Copilot, no protocol implementation is needed. They rely on crawled data and product feeds. Your existing structured data and feed work covers these platforms.

Technical Fundamentals

Across every platform, the same technical fundamentals apply:

• Crawlability: Allow OAI-SearchBot, Googlebot, Bingbot, and PerplexityBot in your robots.txt
• Product identifiers: GTINs on every product for cross-platform matching
• Page speed: Fast load times improve crawl efficiency and indexing across all platforms
• XML sitemap: Complete sitemap with all product URLs submitted to both Google and Bing
• Machine-readable policies: Shipping, returns, and warranty pages with schema.org markup

Non-Shopify Platform Readiness

Each e-commerce platform has different integration paths to the AI shopping platforms. Here is the current state and recommended approach for the three most common non-Shopify platforms.

What's Coming Next

The AI shopping landscape is evolving rapidly. Here are the developments merchants should be watching.

Voice Commerce

Amazon's Alexa+ and Apple's Siri upgrades are bringing AI shopping to voice interfaces. Voice commerce adds another surface where products need to be discoverable by AI agents. The same fundamentals apply -- structured data, product identifiers, and machine-readable store information -- but voice adds constraints around how products are described and compared in audio format.

Agent-to-Agent (A2A) Protocol

Google's A2A protocol enables AI agents to communicate with each other. In a commerce context, this means a user's personal AI agent could negotiate with a merchant's AI agent on price, shipping, or bundle deals. A2A is supported as a transport layer in UCP and represents the next frontier of autonomous commerce.

Multimodal Search

Users are increasingly searching with images, screenshots, and photos. Google Lens, ChatGPT Vision, and similar tools can identify products from photos and surface purchase options. Merchants with high-quality product images, multiple angles, and detailed alt text are better positioned for multimodal discovery.

More Platforms on the Horizon

Claude (Anthropic), Grok (xAI), and other AI assistants are building their own commerce capabilities. Each will need to discover and present products. The foundational work you do today for the five current platforms positions you for whatever comes next. The merchants who invest in AI readiness now will not need to start over when the sixth or seventh platform launches.