Contact

Back to Insights

How to Get Your Products Recommended by ChatGPT

, ,

Consumers are increasingly using ChatGPT to research products, compare alternatives and decide what to buy. Instead of searching for a particular product name, they can describe a need:

ChatGPT can interpret these requirements, research suitable products and present recommendations. For retailers, this creates a new product-discovery channel—and a new visibility challenge.

There is no switch that guarantees your products will be recommended. However, you can make it substantially easier for ChatGPT and other AI search platforms to find, understand and confidently present them.

The foundation is good product data, consistent content, appropriate structured data and a technically sound website.

How does ChatGPT discover products?

ChatGPT can use several sources when researching products, including publicly accessible retail websites, merchant-supplied product information and other relevant sources across the web.

When a shopper asks for a recommendation, the system needs to determine:

OpenAI says ChatGPT shopping research may use public product pages and structured merchant product data. It may compare attributes such as price, features and reviews before presenting suitable options.

This means visibility depends on more than mentioning a popular keyword. Your product information must be accessible, detailed, consistent and relevant to the shopper’s needs.

This work is often described as generative engine optimisation, or GEO. For retailers, GEO is best understood as an extension of good ecommerce SEO and product-data management—not an entirely separate marketing discipline.

1. Create complete, useful product content

A product page with a title, price and two-line description gives an AI system very little information to work with.

Strong product pages clearly explain:

Avoid relying entirely on manufacturer descriptions used by dozens of other retailers. Supplement supplied copy with original information based on your customers’ actual questions.

For example, a product page for a hiking boot could answer:

These details help shoppers make decisions. They also give search engines and AI systems more context for matching the product to specific, conversational requests.

2. Keep product information consistent

Product information is often distributed across multiple systems and pages:

If these sources disagree, it becomes more difficult to determine which information is reliable.

A product should not be described as “waterproof” on its product page, “water-resistant” in its feed and “weatherproof” in a buying guide unless those terms accurately describe different properties.

The same principle applies to prices, availability, product names, colours, dimensions, model numbers and shipping conditions.

Create a reliable source of truth for your product catalogue, ideally through a product information management system or a well-governed ecommerce platform. Establish consistent rules for:

Consistency does not mean copying exactly the same description onto every page. It means keeping the underlying facts aligned while adapting the presentation to each context.

A category page may summarise a product. A buying guide may explain when to choose it. The product page may provide the complete specification. All three should agree about what the product is and what it can do.

3. Strengthen your product feed

A structured product feed gives commerce platforms a dependable, regularly updated view of your catalogue.

Depending on the platform, a feed can include:

OpenAI allows eligible merchants to provide product data for ChatGPT shopping experiences.

Before pursuing a new integration, review the quality of the data you already supply to Google Merchant Center, marketplaces and other sales channels. A feed containing vague titles, missing identifiers, incomplete attributes or stale availability information will remain a weak foundation wherever it is distributed.

Pay particular attention to:

Product-feed optimisation is no longer only a Google Shopping concern. Accurate, structured catalogue data is becoming part of the infrastructure for AI-assisted shopping.

4. Add the right structured data

Structured data labels information on a webpage in a standardised, machine-readable format.

It does not guarantee that ChatGPT will recommend a product, and valid markup does not guarantee a special Google search result. Its purpose is to reduce ambiguity and help machines understand the entities and relationships on your website.

Product schema

Product structured data should be a priority on product detail pages. Depending on the product and page, it can describe:

The structured data must match the product information visible to shoppers. Do not mark up a price, review or availability status that the page does not display or support.

Retailers with product variants should also investigate ProductGroup and related variant markup. This can clarify that several URLs or offers represent sizes, colours or configurations of the same underlying product.

Article schema

Use Article or an appropriate subtype on buying guides, product comparisons and editorial content.

Article markup can identify information such as:

This helps distinguish editorial resources from product and category pages. Clear authorship and accurate update dates are particularly important for content covering changing products, prices or technical requirements.

Breadcrumb schema

BreadcrumbList markup describes a page’s place within the website hierarchy.

For example:

Home → Outdoor Equipment → Hiking Boots → Waterproof Hiking Boots

Breadcrumbs help users navigate the site and give search engines additional context about the relationships between departments, categories, subcategories and individual products.

They should represent a useful customer journey rather than simply reproducing a complicated URL structure.

FAQ schema

Frequently asked questions can add valuable, concise information to product pages, category pages and buying guides. Where appropriate, FAQPage structured data can identify genuine questions and answers that are visible on the page.

However, FAQ schema should not be treated as a shortcut to prominent search results. Google now limits FAQ rich-result visibility primarily to authoritative government and health websites. For most retailers, adding FAQ markup is unlikely to produce an expanded FAQ result in Google.

The real value is in creating useful, clearly structured answers for shoppers. Only use FAQ markup when the page genuinely contains a set of questions with one authoritative answer supplied by the business. Do not mark up customer forums, fabricated questions or content hidden from users.

5. Build strong internal links

A search engine or AI system should not have to discover every product in isolation.

Internal links establish relationships between your products, categories and supporting content. They also help people continue their research without returning to a search engine.

A strong internal-linking structure might connect:

Use descriptive anchor text. “Compare our lightweight travel prams” provides more context than “click here.”

Internal linking is especially valuable for products buried deep within a large catalogue. If an important product is only accessible through filters, pagination or an internal search box, crawlers may struggle to find and understand it.

Review whether priority products receive appropriate links from prominent, indexable pages. Strong internal linking should reflect commercial importance and customer needs, not merely website architecture.

6. Improve page speed and accessibility

Having excellent product information is not enough if the page is difficult to access or use.

Slow product pages create friction for customers and can make large catalogues less efficient to crawl. Google includes Core Web Vitals within its broader assessment of page experience, although speed alone will not make an irrelevant page rank.

Common ecommerce performance problems include:

Optimise images, remove unnecessary scripts and ensure essential product information is present in the rendered page. Test important templates on mobile devices and slower connections, not only on a fast office network.

Use PageSpeed Insights, Lighthouse and Google Search Console’s Core Web Vitals report to identify recurring template-level problems.

The objective is not a perfect score for its own sake. It is a fast, stable page that lets shoppers—and automated systems—access important product information without unnecessary obstacles.

7. Create content around real shopping questions

Product pages capture only part of the buying journey.

Many ChatGPT searches are comparative or problem-led:

Build content that answers these questions using genuine expertise and specific product evidence.

Useful formats include:

Avoid producing dozens of nearly identical articles designed only to target slight keyword variations. Each page should fulfil a distinct customer need and contain enough original value to justify its existence.

Where products are mentioned, link directly to the relevant product or category page. Keep the product claims in editorial content consistent with the catalogue’s source data.

8. Build trust beyond your own website

AI product research may consider information from multiple sources. What your website says about a product can be compared with reviews, publications, marketplaces and other public information.

Retailers should therefore invest in genuine reputation and authority signals:

Do not manufacture reviews or create artificial recommendation websites. Short-term manipulation can damage customer trust and make the retailer appear less reliable.

9. Measure AI visibility as a developing channel

AI-search measurement is less mature than conventional organic-search reporting, but retailers can still establish a baseline.

Start by monitoring:

Use a repeatable set of prompts representing different customer needs. Test them periodically, but do not treat a single response as a fixed ranking result. AI answers can vary according to context, location, available sources and the wording of the request.

The more valuable question is not simply, “Did ChatGPT mention us today?” It is, “Can AI systems consistently access accurate, useful and commercially relevant information about our products?”

A practical ChatGPT product-visibility checklist

Before investing in more experimental GEO tactics, make sure that:

Get your products ready for AI-assisted shopping

Getting products recommended by ChatGPT is not about finding a single optimisation trick. It requires a connected product-information ecosystem.

Your website content, structured data, product feed and supporting articles should reinforce one another. When product information is complete, consistent and easy to access, search engines and AI platforms have a stronger foundation for understanding where each product belongs and which customers it may suit.

If you are unsure whether your catalogue is ready, an AI commerce audit can identify gaps in your product data, structured markup, internal linking, content and site performance—and prioritise the improvements most likely to increase product discovery.

Want to know whether ChatGPT can properly understand your products? Request an AI Commerce Product Visibility Audit.

PREVIOUS ARTICLE

The Complete Guide to eCommerce Product Data

Your Product Data Is More Important Than Your eCommerce Platform When businesses begin planning an eCommerce project, most of the discussion centres around selecting the right platform. Should you choose Shopify or Adobe Commerce? Should you implement a new ERP integration? Should you invest in a Product Information Management (PIM) system? While these are all […]