Missing structured data: the most common reason
One of the main reasons products do not appear in AI applications is missing or incomplete structured data. AI systems like ChatGPT, Perplexity and Gemini read Schema.org markup directly — without that markup they have to interpret the prose, which is far less reliable.
What is structured data?
Structured data is a set of standardized formats for labelling information on web pages in a machine-readable way. The most important format for products: Schema.org Product markup as JSON-LD. It includes:
- Product name (
name) - Brand (
brand) - Description (
description) - Price and availability (
offers) - Product properties (
additionalProperty)
More on optimizing product data: Optimize product data — Schema.org in practice →
JavaScript rendering: invisible to AI
Many shop systems and marketplaces load product content dynamically via JavaScript. That is invisible to users — but a problem for AI crawlers:
- Products in React/Vue/Angular apps: often only visible after JS execution
- Dynamic filters and product variants without their own URLs
- Session-dependent content (login gates, cookie walls)
- Products only on Amazon, eBay or other marketplaces — the URL belongs to the platform
Standalone, static product pages with a fixed URL — alongside your shop. Feed-AI generates these automatically, with complete Schema.org markup and a publicly crawlable URL.
Entities and semantic context
AI systems think in entities — identifiable concepts with clear properties. A product is an entity: it has a name, a brand, a category and measurable properties. The clearer and more consistent that entity is defined, the more often it gets named in AI answers.
- Consistency: product name, brand and model number must appear exactly and uniformly across the page.
- Semantic context: state the use case, target audience and product category in both the prose and the metadata.
- Key terms: use technical specifications and industry-standard terms — not generic marketing phrases.
Building authority and trust
AI systems favour products from sources with recognizable authority. That does not necessarily mean high backlink counts — it means clear, factually verifiable information:
- Complete company information: imprint, contact details and a clear link between the product and its manufacturer.
- Consistent presence: the same product data on your company website, press releases and industry directories.
- Up-to-date content: regularly updated product data (prices, availability, new properties) signals relevance.
The right AI SEO strategy for B2B
B2B products often have specific requirements: technical specifications, certifications, minimum order quantities. AI systems process this information particularly well when it is presented in a structured form:
- Technical specifications as structured data: dimensions, weights, materials and certifications in
additionalProperty. - Describe use cases: concrete use cases help AI systems match products to user questions.
- Build in an FAQ structure: common questions about the product and its use are processed especially well by AI systems.
For a complete overview: How AI systems recommend products: the guide →
Conclusion: a checklist for AI visibility
To make sure your products appear in ChatGPT, Perplexity and Gemini:
- Schema.org Product markup fully implemented (JSON-LD)
- A standalone public URL with no JavaScript dependency
- Factual, exact product descriptions without marketing fluff
- Consistent entities (name, brand, model) across all pages
- Regular updates to the product data
Frequently asked questions
Why doesn't my product or service appear in ChatGPT? +
The most common reasons: missing Schema.org markup, JavaScript-rendered content that is invisible to AI crawlers, no permanent public URL, and marketing language instead of factual data about the product or service. Feed-AI solves all four problems automatically.
How long does it take for my product or service to appear in AI systems? +
After optimization it typically takes 2 to 8 weeks (according to the OpenAI GPTBot documentation and the Perplexity AI Docs, as of 2025). Perplexity with real-time web search responds faster than ChatGPT, which is based on training cycles. With Feed-AI you can measure visibility across all three systems from the Pro plan onward.
Does Schema.org markup really help with AI visibility? +
Yes — Schema.org Product markup is the single most important factor. AI systems read JSON-LD directly and can use it to identify the product name, brand, price and properties precisely, without having to interpret the prose.