How to Submit Your Products to AI Shopping Platforms in 2026: Perplexity, ChatGPT, and Google AI Mode

Something has shifted in how people shop online, and most merchants haven’t caught up yet.

Adobe Analytics measured a 693% increase in traffic from AI sources to US retail websites during the 2025 holiday season. AI-referred shoppers were 33% less likely to bounce and converted 31% better than traffic from other sources. Perplexity grew shopping-intent queries by 5x within months of launching its merchant program. ChatGPT serves over 700 million weekly users, and a meaningful percentage of those interactions are now product research.

The merchants winning in this new channel aren’t the ones with the biggest ad budgets. They’re the ones whose product data is clean, complete, and structured for AI agents to read. The merchants losing are the ones who assumed their existing Google Shopping feed would carry over.

This guide covers what’s actually required to submit to the three platforms that matter most right now: Perplexity, ChatGPT, and Google AI Mode. We’ll cover the technical specs, the application process, and the question most merchants are getting wrong — what to do if you’re on Shopify and “already integrated.”

The landscape, in one paragraph

Each AI shopping platform takes a slightly different approach to merchant data. Perplexity accepts CSV feeds in the Google Merchant Center format via SFTP, plus relies on Schema.org markup on your product pages. ChatGPT uses its own structured feed specification (CSV, TSV, XML, or JSON) ingested through OpenAI’s developer infrastructure with 15-minute refresh cycles. Google AI Mode draws from your existing Google Merchant Center feed plus your site’s structured data and content. All three favor merchants with complete product data, GTINs, real-time pricing and stock, structured Schema.org markup, and substantive product descriptions written for natural language queries.

The three platforms differ in checkout integration (Perplexity has “Buy with Pro” via PayPal, ChatGPT uses the Agentic Commerce Protocol with Stripe, Google AI Mode supports Shop Pay), but for getting your products discovered, the fundamentals overlap heavily. Optimize for one platform and you’ve done most of the work for the others.

Platform 1: Perplexity Merchant Program

Perplexity’s merchant program launched in November 2024 alongside the Buy with Pro checkout feature. It’s free to join and currently accepts merchants who sell and ship to the United States.

Eligibility

You need to be a US-based merchant or one shipping to US customers. There’s no minimum size requirement, but Perplexity’s program description suggests they’re prioritizing established retailers — small Shopify stores often get auto-included via Perplexity’s Shopify integration without needing the formal program.

How to apply

The application is a five-step form that takes under five minutes. Perplexity asks for:

• Business name and contact details
• Website URL
• Shopping vertical (electronics, apparel, beauty, etc.)
• Estimated monthly order volume
• Business registration documents

You can apply at the Perplexity Merchant Program form linked from their shopping help center. Perplexity reviews applications and reaches out with next steps if accepted.

Feed requirements

Perplexity accepts product feeds in CSV format following the Google Merchant Center feed specification. If you already have a Google Shopping feed, you have a head start — but don’t assume it’s ready as-is. Perplexity puts more weight on certain fields than Google does.

Required fields:

• GTIN (UPC/EAN) — non-negotiable for product matching across platforms
• Product title — brand + product type + key differentiators (material, audience, use case)
• Product description — natural language that anticipates questions
• Real-time pricing and inventory — stale data causes products to be dropped from results entirely
• High-quality images — hero on plain background, plus multiple angles
• Customer reviews and ratings — surfaced via AggregateRating schema

Feeds are delivered via SFTP or other secure server methods.

Schema.org markup matters more than you think

Perplexity ingests Schema.org Product structured data directly from your product pages. This is in addition to the feed — not a replacement for it. Without Product, Offer, Review, and AggregateRating schema on your pages, Perplexity’s AI has a harder time parsing your products for conversational matching.

If you’re not currently injecting Schema.org markup, this is one of the highest-leverage additions you can make. UCP Radar generates and hosts a Schema.org loader script that auto-injects this data — you paste a single script tag into your product page template and the loader handles the rest. More on that below.

What Shopify merchants need to know

Through Perplexity’s Shopify partnership, products from Shopify stores can appear in Perplexity Shopping results without joining the Merchant Program. This is genuinely useful — but it’s also where most Shopify merchants get the wrong impression.

The Shopify integration is a passive sync. Whatever data exists in your Shopify store flows through to Perplexity. If your titles are generic (“Style #18383”), if your descriptions are thin, if your products are missing GTINs — Perplexity sees that exact data and ranks accordingly. Auto-integration isn’t a free win. It’s a free exposure of whatever data quality you already have.

The Merchant Program adds a custom dashboard with shopping insights, free API access for building Perplexity-powered search on your own site, and (according to Perplexity’s program description) “increased chances of being a recommended product.” For Shopify merchants with optimized data, both paths work. For Shopify merchants with mediocre data, neither path produces results — they need to fix the data first.

Platform 2: ChatGPT Shopping (OpenAI Merchant Program)

OpenAI launched ChatGPT Shopping with Instant Checkout in September 2025. It’s a more technical program than Perplexity’s, with a published Product Feed Specification and a developer-oriented integration model.

Eligibility

ChatGPT Shopping is currently live in the US with plans to expand. Merchants on Shopify and Etsy are auto-integrated through Shopify Catalog and Etsy’s direct partnership — no application needed for those merchants. Merchants on other platforms can apply through the merchant portal at chatgpt.com/merchants.

The feed specification

OpenAI publishes a detailed Product Feed Specification at developers.openai.com/commerce/specs/feed/. The feed is a structured file in CSV, TSV, XML, or JSON format. Key attributes:

• Basic data: product ID, SKU, title, description, GTIN, product URL
• Variants: group_id to link related SKUs (size, color, etc.)
• Pricing & availability: price, sale price, real-time inventory, geo-specific pricing for regional availability
• Media: primary image plus additional images, videos, and 3D models if available
• Merchant details: brand name, seller URL, store policies
• Shipping: methods, costs, delivery times
• Reviews: aggregate rating and review count
• Control flags: enable_search and enable_checkout to control whether products are discoverable and/or purchasable inside ChatGPT

The 15-minute refresh cycle

This is the part that trips up merchants used to Google Merchant Center’s daily fetch model. OpenAI accepts feed updates as often as every 15 minutes. Merchants who don’t update their feeds frequently — particularly on price changes and stock updates — get penalized fast. ChatGPT treats your feed as the source of truth, so a stale feed means stale prices and out-of-stock items appearing in conversations.

Application process

For non-Shopify and non-Etsy merchants, the application flow is:

1. Submit your business details at chatgpt.com/merchants for verification
2. OpenAI reviews and approves merchants for feed ingestion
3. Approved merchants receive a unique SFTP endpoint with authentication credentials
4. Build a sample feed (start with ~100 representative products) following the spec
5. Submit, validate, and set up automated refresh cycles

If you already manage a Google Shopping feed, the conversion isn’t difficult — most fields map directly. The main additions are the variant grouping system and the discovery/checkout flags.

What Shopify merchants need to know

ChatGPT auto-integrates Shopify stores through Shopify Catalog. Major retailers including Target, Sephora, Nordstrom, Lowe’s, Best Buy, The Home Depot, and Wayfair have integrated for discovery. For Shopify merchants, this means your products are technically eligible to appear without any submission work.

But — and this is the same trap as Perplexity — auto-integration syncs whatever data is in your Shopify product records. If your descriptions are short, your titles are SKU-style, or your GTINs are missing, ChatGPT’s ranking model sees incomplete data and ranks competitors above you. The auto-integration gives you a starting line, not a finish line.

The OpenAI merchant program (for non-Shopify/non-Etsy stores) gives you direct control over the feed, including custom fields ChatGPT uses for ranking that aren’t part of the standard Shopify catalog sync. If you’re a serious merchant on a non-Shopify platform, this is worth pursuing.

Platform 3: Google AI Mode

Google’s approach is the most fragmented of the three. AI Mode draws from multiple sources:

• Your existing Google Merchant Center feed (the same one powering Google Shopping)
• Schema.org structured data on your product pages
• Page content and context that Googlebot crawls

There’s no separate “Google AI Mode merchant program” to join. If you’re already submitting a clean feed to Google Merchant Center, you’re already a candidate to appear in AI Mode results. Google AI Mode also supports Shop Pay checkout through its Shopify partnership for eligible merchants.

What this means in practice

Most merchants who are doing Google Shopping right are already doing Google AI Mode right. The optimizations are largely the same: complete product data, accurate GTINs, real-time pricing and inventory, comprehensive Schema.org markup, well-written product descriptions.

The one nuance: AI Mode tends to favor product pages with more substantive content than traditional Google Shopping. A product page with 50 words of description ranks fine in classic Google Shopping if the feed is clean, but tends to underperform in AI Mode against pages with 200+ words of helpful content (specifications, use cases, comparisons, FAQ-style content). This is why AI Mode optimization often ends up looking like content marketing applied to product pages.

The Shopify question, addressed directly

If you’re on Shopify, you’ve probably read marketing copy from your platform or your apps saying you’re “already integrated with AI shopping platforms.” This is technically true. It’s also misleading.

Shopify auto-integration syncs your existing product data to Perplexity and ChatGPT. It does not:

• Improve weak product titles
• Add missing GTINs
• Generate better descriptions for AI agents to parse
• Inject Schema.org markup on pages that don’t have it
• Optimize images for AI visual search
• Add the structured commerce attributes (variant groupings, geo-pricing, marketplace seller info) that AI platforms use for ranking

The auto-integration is a pipe. It moves your data from Shopify to the AI platforms. If your data is excellent, the pipe works great. If your data is mediocre, the pipe faithfully delivers mediocre data.

This is why “Shopify auto-integrated” merchants often see worse AI shopping performance than smaller competitors who optimized their feeds carefully. The Shopify merchant assumed they were done. The competitor knew they had work to do, did the work, and got rewarded for it.

If you’re on Shopify, the practical advice is the same as if you were on WooCommerce or BigCommerce or a custom platform: optimize your product data first, then worry about which platforms you’re submitting to. Auto-integration without optimization is the worst of both worlds — you have visibility but you rank poorly.

What’s actually required, across all three platforms

Strip away the platform-specific details and you get a small set of requirements that apply everywhere:

Clean, structured product titles. Brand + product type + key differentiator. “Waterproof Adventure Motorcycle Jacket — Touring, All-Season, Black” beats “Style #18383” by an order of magnitude.

Detailed product descriptions written for conversational queries. AI agents match products to natural language questions like “what’s a good waterproof jacket for long-distance touring?” — descriptions that anticipate these questions in plain language outperform feature-list descriptions.

Accurate GTINs (UPC/EAN) on every product. Required for product matching and de-duplication across platforms. Missing GTINs is the single most common reason products don’t appear in AI shopping results despite being in the merchant’s feed.

Real-time pricing and inventory. Stale stock data causes immediate exclusion from results. The 15-minute refresh cadence ChatGPT supports isn’t optional — it’s a competitive baseline.

Comprehensive Schema.org markup on product pages. Product, Offer, Review, AggregateRating schema. AI agents read these directly when validating feed claims against the live site.

robots.txt that allows AI crawlers. GPTBot, ClaudeBot, PerplexityBot, and Google-Extended all need explicit access. Many default robots.txt files block them as collateral damage from generic disallow rules.

Reviews and ratings. Surfaced through AggregateRating schema. AI agents use review volume and sentiment as ranking signals — not just for trust, but for understanding what kind of customer would buy this product.

That’s the list. Get those seven things right and you’re competitive on every AI shopping platform that exists today and probably most that launch in the next year.

Where UCP Radar fits

UCP Radar is built specifically for this layer of the stack. We take your existing product feed, score it across 100 quality points, and use AI to enrich the gaps — better titles, structured descriptions, missing attributes, FAQ-style content for conversational matching. The output is a supplemental feed you can plug into Google Merchant Center (which feeds AI Mode), plus a UCP JSON Feed and llms.txt file designed specifically for AI agents, and a Schema.org markup script that auto-injects structured data into your product pages.

For Shopify merchants, this is the optimization layer that makes auto-integration actually work. You don’t need to leave Shopify or change platforms — you optimize the data Shopify is already syncing to Perplexity and ChatGPT. For non-Shopify merchants, UCP Radar produces the same optimized feed you can submit directly to OpenAI’s merchant program or Perplexity’s program.

Either way, the technical requirements above are the same. AI shopping platforms reward merchants whose data is clean and structured. Everything else — which platforms you submit to, which integrations you use, which checkout systems you support — is downstream of that.

A pragmatic order of operations

If you’re starting from scratch and want to participate in AI shopping, here’s the order that maximizes return on effort:

1. Audit your product feed quality. Run a feed quality check. UCP Radar’s 100-point UCP score is one option; there are others. Identify what’s missing — GTINs, descriptions, attributes, structured data.

2. Fix the data layer. Optimize titles, expand descriptions, add missing GTINs and variant attributes. This is the part that pays off everywhere — Google Shopping, AI shopping, on-site search, marketplaces.

3. Configure robots.txt to allow AI crawlers. Add explicit allows for GPTBot, ClaudeBot, PerplexityBot, and Google-Extended. This is a 10-minute task that unblocks the rest of your AI visibility strategy.

4. Add Schema.org markup to product pages. If you’re using a tool like UCP Radar, this is a single script tag. If you’re doing it manually, you’ll be writing JSON-LD per product page — much more painful at scale.

5. Submit to platforms where you have direct control. Perplexity Merchant Program if you’re US-based. OpenAI Merchant Program if you’re not on Shopify/Etsy and want direct feed control. Google Merchant Center (which you should have anyway) for Google AI Mode coverage.

6. Monitor and iterate. Track which products are appearing in AI shopping results, which queries surface your products, which categories are converting. The early-mover advantage in AI shopping is real — merchants who optimized first are seeing meaningful traffic shifts.

Closing thought

The merchants winning in AI shopping in 2026 aren’t necessarily the largest or the most established. They’re the ones whose product data was ready when the AI agents started looking. The bar is rising fast — what worked six months ago is being out-optimized by competitors who took the data quality work seriously. The good news is that the work itself is finite and well-defined. Get the seven requirements above right, and you’re competitive across every AI shopping platform that exists today.

The bad news is that auto-integration alone doesn’t get you there. Whether you’re on Shopify, WooCommerce, BigCommerce, or a custom platform, the data quality work is the same, and it’s the work that determines whether you rank.

Start with the audit. Everything else follows from there.