How to Get Your Products Recommended by ChatGPT and Other AI Assistants

This guide draws on what OpenAI, Google, Perplexity and Microsoft have published about their shopping surfaces, as of October 2026. For more on selection, see how AI shopping agents decide which products to recommend and buy.

How do ChatGPT and other AI assistants choose which products to recommend?

AI assistants choose products from the data they can read: structured data, your product pages, and what reviews and other sites say about you.

OpenAI says ChatGPT selects product results independently, and they are not ads or influenced by OpenAI partnerships. Its help pages list these inputs:

• Structured metadata. Price, product description and similar data from first-party and third-party providers, plus other third-party content.
• Merchant ranking. When several merchants sell the same item, they are ranked on factors like availability, price, quality and whether they are the maker or primary seller.
• Review summaries. These are model-generated from reviews on public websites, and OpenAI does not verify the reviews or ratings.
• Shopping research. This ChatGPT mode may draw on merchant data shared through the Agentic Commerce Protocol, publicly available product information and other retail sources. OpenAI says its results are organic, read product pages directly, cite sources and avoid low-quality or spammy sites.

Google's guidance on AI features says no additional requirements or special optimizations are needed to appear in AI Overviews or AI Mode. Its AI optimization guide says Merchant Center feeds can help products be visible in AI responses.

Perplexity says merchants with deeper product details, such as availability, reviews, pricing and specifications, are more likely to be recommended. Its algorithm sets listings on relevance and ratings, and advertisers cannot pay to list products in its related products section.

Can shoppers still buy inside ChatGPT, and what are ACP and UCP?

Mostly not. OpenAI's merchant page says it is moving away from a standalone Instant Checkout, so shoppers discover products in ChatGPT and complete purchases on merchant-owned websites or apps.

When OpenAI announced the change on 24 March 2026, it said: "We've found that the initial version of Instant Checkout did not offer the level of flexibility that we aspire to provide, so we're allowing merchants to use their own checkout experiences while we focus our efforts on product discovery".

OpenAI's shopping help page still says ChatGPT may show an Instant Checkout option for some eligible products and merchants.

Underneath sit the protocols:

• ACP from Stripe and OpenAI. The Agentic Commerce Protocol is a common language for how agents and businesses transact, including checkout and sharing payment credentials. The business stays merchant of record and controls which products can be sold, how they are presented and how orders are fulfilled.
• UCP from Google. The Universal Commerce Protocol handles checkout on AI Mode and Gemini, once merchants configure shipping, returns and their feed, join a waitlist and get Google's approval. Google said in September 2026 that UCP already enables direct checkout for hundreds of thousands of brands and retailers.
• UCP feeds at Microsoft. Microsoft Advertising made UCP-ready feeds in Microsoft Merchant Center generally available in the U.S. in April 2026, to help catalogs surface in Copilot.

The protocols change where the order is placed. Which product gets named still comes from the data and pages the assistant reads.

Do you need a product feed for ChatGPT and Google?

For ChatGPT a feed is optional but worth having: OpenAI says it gives you more control over how products appear and keeps information more accurate and current. For Google and Microsoft, a Merchant Center feed is a documented route into their AI surfaces.

ChatGPT: OpenAI's feed and crawler

A feed does not replace crawling, so cover both:

1. Check your route in. OpenAI's merchant page says merchants apply to share a feed and join a waitlist; Shopify and Etsy catalogs are already integrated. Shopping in ChatGPT is live for U.S. users.
2. Fill the required fields. OpenAI's feed spec has 9 required fields: item_id, title, description, url, brand, seller_name, image_url, availability and price. It asks for titles and plain-text descriptions of at most about 150 and 5,000 characters respectively.
3. Add the optional fields. These cover variants, attributes, shipping, returns and reviews. OpenAI says richer data helps products surface with more accurate information, and lists better placement as a benefit.
4. Allow the crawler. OpenAI's crawler documentation says sites that opt out of OAI-SearchBot will not be shown in ChatGPT search answers. A blanket AI-bot block at your CDN can drop your store from those answers; see the robots.txt mistake that blocks AI crawlers.

Google and Shopify

Consider submitting conversational attributes in Merchant Center to add richer context to listings. In testing with lululemon, described in Google's holiday shopping update, attributes the brand submitted were used 50% of the time in relevant AI Mode product recommendations. That is a single brand's test, not a benchmark.

Shopify said in September 2026 that Meta now sits alongside its other AI channels in Agentic Storefronts, with products shared with Meta by default through Shopify Catalog.

Where each channel gets your products

What should your product pages and schema include?

Every product page should state what an assistant needs in plain HTML. Repeat it in Product and Offer structured data. Google's AI guidance says structured data should match the visible text.

• Product markup for identity. Name, brand, image, SKU and a GTIN where you have one help an assistant match the item across your site, retailers and review pages.
• Offer markup for what changes. Price, currency, availability, shipping and returns.
• Live price and availability. Pull them from inventory and never hardcode "InStock". Google's merchant listing documentation warns that dynamically generated markup can make Shopping crawls less frequent and less reliable, which can be an issue for fast-changing price and availability.

Reviews belong in the initial HTML

Google renders pages with a headless Chromium that executes JavaScript once its resources allow. Its JavaScript SEO guide still recommends server-side or pre-rendering because not all bots can run JavaScript.

In data Vercel published in December 2024, ChatGPT and Claude crawlers fetched JavaScript files but did not execute them, so they could not read client-side rendered content. Render review text and ratings in the initial HTML, not only in a widget.

FAQ markup is optional

Answer common product questions ("Is it dishwasher safe?") in visible text, where shoppers and assistants both read them. In August 2023 Google limited FAQ rich results to well-known, authoritative government and health websites.

What do AI assistants read beyond your own website?

AI assistants also read public reviews and other third-party content. OpenAI lists "other third-party content" among ChatGPT's inputs, and its review summaries come from reviews on public websites.

• Public reviews. Ask for reviews on publicly crawlable platforms, not only in a private widget.
• Consistent product names. Use the same name for each product and variant everywhere you sell, or an assistant may split your reviews across what look like different products.
• Current retailer listings. ChatGPT ranks merchants partly on whether they are the maker or primary seller. Your store can win that slot, but stale retailer listings can still shape how the product is described.
• Editorial coverage. OnlyAEO builds a ranked list of third-party sites AI cites in your space where you are not mentioned yet, with a recommended play for each (guest post, partnership, listing or review).

What content on your own site earns product recommendations?

Pages that answer the way shoppers phrase questions earn recommendations: by use case, by comparison, as a gift and as a replacement. A category grid answers none of them.

Comparison pages must show both sides, including when the other product is the better pick. Admitting your product is heavier or pricier for some buyers gives the assistant something to quote.

Each buyer persona gets a different answer, so OnlyAEO researches each one from the brand's market.

Can your own website become the source AI assistants cite?

Yes. In an OnlyAEO program for a New York edibles retailer, the brand went from 1.5% AI visibility, ranked #50, to 25% visibility and #4 of 374 brands. Its own website became the most-cited source in the category, ahead of New York State's official site.

It is a multi-location, state-licensed retailer, not an online-only DTC brand, but it faced the same sources a product brand competes with: directories, review platforms and large incumbents.

What the program did, per the New York edibles retailer case study:

• Buyer questions as targets. The audit captured the questions AI models were asked across every buyer type, and each became a content target structured for a direct answer.
• Location-aware pages. It built content for the neighborhoods and towns each store serves.
• Own site as the source. Answer capsules, structured data, comparison tables and pricing and loyalty transparency were built to make its own site the page AI cites. Pages also carried FAQ schema.

How do you measure whether AI assistants recommend your products?

Ask assistants your shoppers' questions and track products, not only the brand. A healthy brand mention rate can hide a recommendation for your discontinued model.

1. Run shopper questions in each assistant.
2. Record the products named, yours and competitors'.
3. Check each description: price, main specs, availability and the product page link.

Adobe's August 2026 AI traffic report found AI-referred retail visitors converted at a 60% higher rate than non-AI traffic. AI-referred shoppers also generated 53% more revenue per visit, a reversal from 12 months earlier, when non-AI visits were worth 128% more. These are market-wide figures, not a forecast for your store.

OnlyAEO Commerce shows where your products appear when shoppers ask AI what to buy, why the models pick what they pick, and where you lose to competitors. You get a per-product score, a category leaderboard and the trend over time.

Our content engine then writes the pages that win those placements, with brand and factual guardrails on every page. For the metrics, see measuring AI visibility for e-commerce.

Commerce sits alongside OnlyAEO's other products: AI Feed, the AI Visibility audit and report, LLM Attribution and a free llms.txt generator.

How long does it take for products to start getting recommended?

First measurable results usually take at least 60 days. Some steps also wait on other queues: OpenAI's feed program has a waitlist, and UCP checkout needs Google's approval.

OnlyAEO does not guarantee results or promise specific rankings or citations. No one outside the assistant can promise a placement, but you control the data and pages it reads.

How do you choose a partner for AI product visibility?

Choose one that tracks each product, shows the sources assistants cite, and is straight about timing.

1. Does it track products? Reports should show which of your products each assistant names.
2. Which sources does it map? OnlyAEO ranks third-party sites AI cites in your category that do not mention you yet, with a recommended play for each.
3. Is each persona researched? Each buyer persona gets a different answer, so OnlyAEO researches each one from the brand's market.
4. Who fixes the technical layer? Ask who fixes feeds, rendering and markup.
5. What does it promise? A guaranteed placement is a promise no outside provider can keep; OnlyAEO does not guarantee results, rankings or citations.

How we tested

On 13 October 2026 our content engine asked ChatGPT and Claude this article's main question. ChatGPT answered without searching the web.

Claude searched for how e-commerce brands get products recommended in ChatGPT shopping, Instant Checkout, product feeds and the Agentic Commerce Protocol. It cited agency blogs and trade publications; neither assistant named OnlyAEO.

Frequently asked questions

Can a brand pay to have its products placed in ChatGPT shopping results?

No. OpenAI says ads do not influence ChatGPT's answers: ads run on separate systems from the chat model, and advertisers cannot shape, rank or alter responses. Placement comes from your product data, your pages and what other sites say about you.

Does OpenAI take a fee when a ChatGPT shopper buys from us?

OpenAI's merchant page says there are no fees on purchases that start in ChatGPT. Purchases are completed on your own site or app, so checkout stays with you.

Is shopping in ChatGPT available outside the United States?

Not yet. OpenAI says shopping is live for ChatGPT users in the U.S. and that it plans to expand to more merchants and regions over time. Brands elsewhere can still prepare product pages, crawler access, public reviews and a Google Merchant Center feed.

What does OnlyAEO's managed service include for an e-commerce brand?

Monthly AI visibility analytics; researched, fact-checked articles published on your site in your voice and design; YouTube videos planned, scripted and packaged for your channel; articles in other languages; help getting mentioned on the trusted sites and listings AI reads; and a website health check so every AI engine can read your site.