What does an export-focused independent store need first to get recommended by ChatGPT?
First, let crawlers in and let them read your pages. OpenAI's official docs state that a site must allow OAI-SearchBot to appear in ChatGPT search answers. Then the product page itself: price, specs, shipping and returns, the things buyers ask about, must be written as text and structured data, not hidden in images or in pop-ups that appear only after a click.
To be honest, no company has published rules for how products get recommended. What follows is based on Google's official field requirements and what we have observed of AI citations. For a product entry point on OpenAI's side, we did not find an official description.
What should a product page include for AI to understand it?
| Content | How to write it | Why |
|---|---|---|
| Name, image, price, currency | Use Product + Offer structured data; price above 0, currency as a three-letter ISO code | Google merchant listing docs |
| Stock status | Write availability, for example InStock | Google recommended field |
| Spec table | Use an HTML table, one spec per row: material, size, weight, capacity, certifications | Inference: tables are easier to extract than paragraphs |
| Shipping cost and delivery time | Write it as text on the page, and use shippingDetails in structured data: shipping cost, destination country, handling and transit days | Optional Google field; if you write it, fill it in completely |
| Return policy | Write it as text on the page, and use hasMerchantReturnPolicy in structured data: applicable countries, policy category, return days | Google recommended field |
| Brand and company information | State the company, address and contact details clearly on the About page | Inference: used to confirm you are a real merchant |
Structured data must match the text visible on the page; Google's docs are explicit about this. If you write a price of 24.90 and the page shows 29.90, the information simply contradicts itself.
How exactly do you write the structured data?
Below is a JSON-LD example for a product page, with fields matched to Google's merchant listing docs. The name, URLs and price are placeholders; replace them with your own.
{
"@context": "https://schema.org",
"@type": "Product",
"name": "Insulated Water Bottle 750 ml",
"image": ["https://your-store.com/img/bottle-1x1.jpg", "https://your-store.com/img/bottle-4x3.jpg"],
"description": "Stainless steel, keeps drinks cold 24 h, contains no BPA.",
"sku": "BTL-750-BLK",
"offers": {
"@type": "Offer",
"url": "https://your-store.com/products/bottle-750",
"price": "24.90",
"priceCurrency": "USD",
"availability": "https://schema.org/InStock",
"shippingDetails": {
"@type": "OfferShippingDetails",
"shippingRate": { "@type": "MonetaryAmount", "value": "4.99", "currency": "USD" },
"shippingDestination": { "@type": "DefinedRegion", "addressCountry": "US" },
"deliveryTime": {
"@type": "ShippingDeliveryTime",
"handlingTime": { "@type": "QuantitativeValue", "minValue": 1, "maxValue": 2, "unitCode": "DAY" },
"transitTime": { "@type": "QuantitativeValue", "minValue": 5, "maxValue": 9, "unitCode": "DAY" }
}
},
"hasMerchantReturnPolicy": {
"@type": "MerchantReturnPolicy",
"applicableCountry": "US",
"returnPolicyCategory": "https://schema.org/MerchantReturnFiniteReturnWindow",
"merchantReturnDays": 30
}
}
}
Google suggests putting the site-wide return policy in Organization markup, and using product-level policies only to override exceptions. For more types, see Does structured data help AI search?.
Why write separate return and shipping policy pages?
Buyers often ask AI things like “can I return it?” or “how long to ship to a given country?” If your product page only says “returns supported”, AI has no detail to cite. A separate policy page should state clearly:
- Which countries and regions it applies to, and where you do not ship
- Where the return period starts counting, and how many days it lasts
- Who pays return shipping, and how long a refund takes to arrive
- Which items cannot be returned
- The last updated date, stated on the page
The answers to these questions should sit in the policy page's text, linked from every product page.
Why do third-party reviews matter for e-commerce too?
On 2026-10-26 we reviewed GammaSlide (a software product, not e-commerce): AI read 28 cited web pages, of which 19 were pages listing multiple similar products. Inference: the same likely holds for e-commerce; when AI answers “best brand of X”, it is more likely to cite lists and reviews written by others. We have no review data for e-commerce sites; this point is extrapolated from software product data, not measured on e-commerce.
1. Find the review sites cited in your category
Ask AI questions like “best brand in a category” or “is a category worth buying”, and note the URLs it cites.
2. Send samples or offer trials to reviewers
Provide real products and accurate spec information; do not ask for only positive reviews.
3. Put real buyer reviews on your own pages
Use Review or AggregateRating structured data; the numbers must come from real reviews and match what the page displays.
How do you check when you are done?
1. Check the markup with Google's Rich Results Test
Paste in a product page URL and look for errors and warnings.
2. Check that robots.txt allows OAI-SearchBot
Also check PerplexityBot and Claude-SearchBot. For the full set of checks, see AI visibility audit checklist.
3. Ask AI with your buyers' questions
For example “ship to US, 750 ml insulated bottle, which brand”. Note whether you appear and whom it cites.
FAQ
How do I get ChatGPT to recommend my products from an export-focused independent store?
First allow OAI-SearchBot, then write price, specs, shipping and returns into the product page and structured data, and work to get third-party reviews. There is no approach that guarantees a recommendation.
Should I add Product structured data?
We suggest it. Google's docs say appearing in AI Overviews needs no special markup, but merchant listings themselves need Product and Offer. We did not find an official statement on whether OpenAI reads it. Structured data must match the visible text on the page.
Does OpenAI have a product feed I can submit?
We only saw third-party articles relaying that OpenAI is exploring letting merchants submit product feeds directly. We did not check OpenAI's original text, so this is unverified. Go by OpenAI's official sources.
Where should the return policy go?
On a separate policy page, linked from every product page. In structured data, use hasMerchantReturnPolicy; Google suggests putting site-level policies under Organization.
What if no review site is willing to cover me?
First get your own pages right: specs, shipping, returns and contact details all in place. Then offer samples and accurate information to reviewers in your category. Do not pay to pass off a placement as an independent review.
Sources
- Google · Merchant listing (Product, Offer) structured data （Checked 2026-11）
- Google · AI features and your website （Checked 2026-11）
- OpenAI · Overview of OpenAI Crawlers （Checked 2026-11）
- kagentx evaluations: the data on the case study pages; each example gives its date.