Product markup thin enough to get skipped
An agent comparing products needs price, availability, identifier, brand and image to place you in a comparison — listings missing those get dropped rather than ranked low.
Dropped, not ranked low
Search degrades gracefully: thin markup ranks worse. Agent comparison does not. An agent building a shortlist across five retailers needs the same fields from each. A listing without a parseable price cannot be compared on price, so it leaves the set.
The failure mode is silence — you do not appear at all, and nothing tells you why.
The fields that matter
offers.price and offers.priceCurrency. offers.availability as a schema.org URL, not the word "in stock". A sku or gtin so the agent can tell your listing and a competitor's are the same physical product. brand. image. aggregateRating if you have genuine reviews.
gtin is the one most often skipped and the most valuable: it is what makes cross-site matching possible at all.
{
"@context": "https://schema.org",
"@type": "Product",
"name": "Aeron Chair, Size B",
"sku": "AER1B23DW",
"gtin13": "0885370123456",
"brand": { "@type": "Brand", "name": "Herman Miller" },
"image": "https://acme.com/img/aeron-b.jpg",
"description": "Ergonomic office chair with PostureFit SL.",
"offers": {
"@type": "Offer",
"price": "1395.00",
"priceCurrency": "USD",
"availability": "https://schema.org/InStock",
"url": "https://acme.com/products/aeron-b"
},
"aggregateRating": {
"@type": "AggregateRating",
"ratingValue": "4.6",
"reviewCount": "1284"
}
}Keep it true
Markup that disagrees with the page is worse than missing markup. A stale price in JSON-LD produces an agent quoting a number you will not honour — a support problem and a trust problem at once.
Generate structured data from the same source as the rendered page. Never maintain it by hand.
Does this apply to your site?
The free checker runs this check and eighteen others, then generates the files to fix whatever fails.
Run the checkerRelated
- UCP, ACP and AP2: the agentic commerce stackThree protocols, three different jobs — discovery, checkout and payment authority. What each does, who's behind them, and what to implement first.
- JSON-LD is how AI answer engines know what you areStructured data turns a page into facts an engine can quote with confidence. Which schema types matter, and the @graph mistake that makes valid markup look invalid.