Schema.org Product + JSON-LD

Listrar summary

Schema.org is the shared vocabulary of the machine-readable web — founded by the major search engines in 2011 — and JSON-LD is the W3C format it is most often written in. Mark a page up as schema.org/Product and its name, GTIN, brand, materials and certifications become data any crawler or agent can consume without scraping. No law mandates it; every serious consumer of product pages rewards it. Listrar server-renders validated Product JSON-LD on every public passport page, so the same canonical record that satisfies a regulator is legible to a search index and an AI shopping agent.

Who does it apply to?

• Anyone publishing product pages they want machines to understand
• Brands whose products should surface correctly in search and in AI answers
• Marketplaces ingesting product data from thousands of sellers
• Agent builders who would rather parse data than HTML

Main obligations

• None in law — the disciplines are self-imposed
• Emit valid JSON-LD typed as schema.org/Product, with gtin, brand, offers and the fields that matter
• Render it server-side: crawlers and agents do not reliably execute your JavaScript
• Keep the markup and the visible page telling the same story — divergence is how trust is lost

Key dates

2 June 2011

Schema.org founded by Google, Bing and Yahoo

16 January 2014

JSON-LD 1.0 becomes a W3C Recommendation

16 July 2020

JSON-LD 1.1 Recommendation

2023 →

AI agents join search engines as the main readers of structured product data

Products in scope

• Any product with a public page — physical or digital
• Especially products whose GTIN, certifications and material claims deserve to be read exactly, not inferred

Obligations by actor

Brand

Publish Product JSON-LD on every product page, validated, not hand-maintained

Registry

Generate the markup from the canonical record so page and data never diverge — Listrar's approach

Marketplace

Ingest and re-emit structured data rather than re-typing it

Agent

Read the JSON-LD island; fall back to scraping only when there is none

Documents and evidence to keep

• Validation reports (Schema.org validator, search-console rich-result checks)
• The generation path from record to markup — one source, no hand edits
• Change history of the record, which becomes the change history of the markup

Sanctions and risks

• No legal sanction — the cost is invisibility
• Pages without structured data lose rich results and get paraphrased, sometimes wrongly, by agents
• Markup that contradicts the page is treated as spam by search engines

Key references

• schema.org/Product — the Product type and its properties
• JSON-LD 1.1 — the W3C Recommendation

Official specification

• Schema.org — vocabulary home