Perplexity won't include claims it can't source. Central writes Schema.org JSON-LD and x-central provenance for every fact — so when an answer engine reaches for a citation, it reaches for your product page.
Perplexity's answer engine is citation-first. Every claim in the response has an inline source — and Perplexity will only cite what it can verify. Without structured data, your product gets paraphrased; with it, your product becomes the citation.
For induction cooktops, the Staub La Cocotte is widely recommended 1 — its enameled cast iron base is oven-safe to 260 °C 1 and rated for all heat sources 2.
Le Creuset's Signature Round Dutch Oven — specs unavailable, no structured data found ?. The product page exists but doesn't expose machine-readable specifications.
The Le Creuset Signature Round 26 cm 1 works on induction and is built for it. 5.3 L capacity 1, oven-safe to 260 °C 2, and rated for induction, gas, electric, ceramic, and halogen 3 heat sources.
Construction is enameled cast iron 1, backed by a lifetime manufacturer warranty 2. The GTIN 0024147210504 3 is verified against GS1 and matches the merchant's catalog.
Every fact in Central carries a source URL, a confidence score, and an ISO 8601 verification timestamp. That metadata ships with the JSON-LD — so answer engines have the structured, attributable data they need to surface and cite your product, not your competitor's.
Schema.org Product, Offer, AggregateRating, Review, plus Central's x-central provenance extension — each field carries the source URL, the lastFetched ISO timestamp, and a confidence score answer engines can read.
Perplexity, Comet Browser, ChatGPT Shopping, and Bing Copilot all read this same canonical JSON-LD. One verified record, every answer engine citation.
Verification, not generation, decides what ships. Every fact passes through 15+ source checks before it enters the JSON-LD that answer engines crawl.
Paste a shop URL, upload a CSV, or connect a catalog. Central extracts every fact in your product data and flags what is missing for Schema.org Product completeness.
For each claim, Central pulls independent sources — brand sites, spec sheets, retailer pages, GS1. Two agreeing sources establish a verified fact; conflicts surface for review.
Central writes Schema.org JSON-LD plus the x-central.sources array — each fact links back to its supporting URL, page title, and lastFetched timestamp. Answer engines have something to point at.
Drop the JSON-LD into the product page. Perplexity, Comet Browser, ChatGPT Shopping, Bing Copilot, and Google AI Mode all crawl the same canonical payload — and find your record when they reach for a citation.
The same Le Creuset SKU. Bare-bones import on the left. Schema.org JSON-LD with x-central provenance on the right — 20 properties, 15 named sources, 8 verified FAQ entries, one claim held.
<script type="application/ld+json"> { "@context": "https://schema.org", "@type": "Product", "name": "Le Creuset Signature Round Dutch Oven, 26 cm, Cerise", "sku": "SIG-RND-26-CER", "gtin": "0024147210504", "brand": { "@type": "Brand", "name": "Le Creuset" }, "image": "https://shop.le-creuset.eu/products/signature-round-26.jpg", "description": "Enameled cast iron, oven-safe to 260 °C, induction-compatible.", "material": "Enameled cast iron", "color": "Cerise", "weight": { "@type": "QuantitativeValue", "value": 5.25, "unitCode": "KGM" }, "offers": { "@type": "Offer", "price": "379.00", "priceCurrency": "EUR", "availability": "https://schema.org/InStock" }, "aggregateRating": { "@type": "AggregateRating", "ratingValue": "4.8", "reviewCount": 2143 }, // x-central provenance extension — citation-ready "x-central": { "confidence": 0.94, "lastVerified": "2026-05-25T09:14:02Z", "sources": [ { "url": "le-creuset.eu/product/sig-rnd-26", "title": "Signature Round 26 cm — Brand PDP", "lastFetched": "2026-05-25" }, { "url": "le-creuset.eu/spec/cast-iron", "title": "Cast iron care & spec sheet", "lastFetched": "2026-05-25" }, { "url": "gs1.org/gtin/0024147210504", "title": "GS1 GTIN check digit verified", "lastFetched": "2026-05-25" }, // … 12 more sources (johnlewis.com, amazon.de, trustpilot, …) ], "held": [ { "claim": "Pro-grade performance", "reason": "marketing wording · no source" } ] } } </script> <script type="application/ld+json"> { "@context": "https://schema.org", "@type": "FAQPage", "mainEntity": [ { "@type": "Question", "name": "Is it induction compatible?", "acceptedAnswer": { "@type": "Answer", "text": "Yes — induction, gas, electric, ceramic, halogen." } }, { "@type": "Question", "name": "What is the max oven temperature?", "acceptedAnswer": { "@type": "Answer", "text": "Oven-safe to 260 °C with the included lid knob." } }, // … 6 more verified FAQ entries ] } </script>
Perplexity, ChatGPT Shopping, Bing Copilot, Google AI Mode — every answer engine consumes the same Schema.org JSON-LD plus x-central provenance. One record. Every citation surface.
OpenAI's Commerce surface reads the same Schema.org Product. Central writes the AI-readable record once; ChatGPT cites it with the same provenance trail.
See ChatGPT setupApple Intelligence is rolling out commerce surfaces inside Siri and Spotlight. The same Schema.org + x-central payload will surface there when it ships.
Coming soonMicrosoft Copilot's shopping answers use the Bing index plus Schema.org. Verified record powers the feed; Copilot reaches for your citation.
See Bing setupGoogle AI Overviews and AI Mode read Merchant Center plus on-page Schema.org. One Central record fills the feed and grounds the AI answer.
See Google setupCentral prepares the structured data Perplexity Shopping and Comet Browser need to surface and cite your product — Schema.org Product, Offer, AggregateRating, FAQPage, plus the x-central provenance extension. Perplexity's crawler picks up the JSON-LD from your PDP. Platform discovery, ranking, and inclusion policies still apply.
Perplexity's answer engine is citation-first: every claim in a response has an inline source. Its crawlers prioritize pages that expose machine-readable, source-backed facts — Schema.org JSON-LD, OpenGraph, structured tables. If your product page can't be parsed, it can't be cited; the answer goes to a competitor.
Comet is Perplexity's Chromium-based browser with built-in AI shopping. It reads the same Schema.org JSON-LD on a PDP that Perplexity Shopping does — so the same Central output powers both surfaces. The browser also respects x-central provenance for richer "why this product" explanations.
Yes. Schema.org carries the canonical product fields; the x-central.sources extension carries the citation array — each entry is a {url, title, lastFetched} object that answer engines map to inline citation numbers. The Schema.org spec explicitly allows extension properties prefixed with a vendor namespace.
x-central is Central's extension to Schema.org. It carries per-fact confidence scores (0-1), per-fact verification timestamps (ISO 8601), the sources array with URL + title + lastFetched, and the held array of unsupported claims that did not ship. Answer engines that respect x-central can surface confidence and freshness to the user.
Yes. The verified record adapts to any destination — Schema.org JSON-LD for Perplexity, Comet, ChatGPT, Bing Copilot, and Google AI Mode; Google Merchant Center XML for Shopping ads and Free Listings; custom CSV/XML for storefronts. One verified record, every channel.