What product schema markup does
Search engines
Use it for rich results: price, availability, and review stars in the listing, which lifts click-through.
Shopping surfaces
Use it to populate product listings and comparison experiences, where the structured fields are the only thing being read.
AI assistants
Use it to extract facts when composing recommendations. This is the fastest-growing reason to care, because an assistant that cannot determine your price or availability filters you out of consideration entirely.
Three audiences consume it, and the third is new. Our how LLMs find products guide covers that retrieval process and where structured data enters it.
AI Search Grader by bCloud AI
Grade your ecommerce search in 10 quick questions
31% of ecommerce searches return zero results — and most shoppers who hit a dead end leave for a competitor. How does your store's search stack up?
Answer 10 short questions and get your AI search score, plus a personalized report to fix the gaps. Free, takes about 2 minutes.
No signup needed to take the quiz.
Understanding intent…
Scoring your answers across relevance, AI, experience, and insights.
Your AI search score is ready
Tell us where to send your personalized report. You'll see your score and recommendations right away.
Your score by pillar
Personalized recommendations
Fix the gaps in weeks, not quarters
bCloud AI replaces keyword-only search with hybrid AI retrieval — sub-200ms responses, 99.99% uptime, and conversion lifts of up to 40% across 50+ implementations.
The properties that matter
Product schema markup defines dozens of Product properties. A handful do most of the work.
Required for rich results.
name, image, and an offers block containing price, priceCurrency, and availability. Without these you get nothing.
Strongly recommended.
description, sku, brand, gtin (or mpn for parts without a GTIN), and aggregateRating with review if you have genuine reviews.
Increasingly valuable for AI.
additionalProperty entries carrying the attributes that do not have dedicated fields: material, waterproof rating, capacity, compatibility, dimensions. A shopper asking for a jacket rated to 20,000mm can only be matched if that number exists somewhere machine-readable.
For variants.
Use ProductGroup with hasVariant rather than cramming everything into one Product. Size and colour variants have different SKUs, prices, and availability, and flattening them produces wrong answers.
Offer details worth adding.
priceValidUntil, shippingDetails, and hasMerchantReturnPolicy. Google uses these for shopping experiences, and they answer questions shoppers genuinely ask.
Attribute depth is where most catalogs under-invest, and it is precisely what determines whether you match a specific request rather than a generic one.
What changed in 2026
Three developments worth knowing, because advice written before them is now misleading.
Google phased out the FAQ rich result.
FAQ markup no longer produces the expandable Q and A display in search results for most sites. That does not make FAQ content worthless, since research still finds FAQ sections correlating with higher AI citation rates, particularly when each answer is self-contained. Implement it for retrieval value, not for a SERP feature that is gone.
AI-specific reporting arrived.
Google introduced generative AI performance reporting in Search Console in 2026, and announced AI performance insights in Merchant Center covering discovery across AI Mode and AI Overviews, with metrics on visibility, product terms, attributes, and completeness.
Feeds became content.
For ecommerce, your Merchant Center feed and your on-page product schema markup are two expressions of the same data, and inconsistency between them creates doubt. Audit both together.
That word completeness is the signal. Google is now telling merchants explicitly that attribute completeness affects AI discovery, which moves catalog enrichment from housekeeping to a reported metric.
How to implement it
Use JSON-LD for product schema markup. Google recommends it, it is the easiest to maintain, and it sits in a script tag rather than being entangled with your HTML. Microdata and RDFa work but are harder to change without breaking layout.
A minimal but genuinely useful example:
{
"@context": "https://schema.org",
"@type": "Product",
"name": "Trailhead Waterproof Commuter Jacket",
"image": ["https://example.com/jacket-front.jpg"],
"description": "Packable waterproof shell rated to 20,000mm, designed for bike commuting.",
"sku": "TH-745-B",
"mpn": "TH745B",
"brand": { "@type": "Brand", "name": "Trailhead" },
"additionalProperty": [
{ "@type": "PropertyValue", "name": "Waterproof rating", "value": "20000mm" },
{ "@type": "PropertyValue", "name": "Weight", "value": "310g" },
{ "@type": "PropertyValue", "name": "Packable", "value": "Yes" }
],
"offers": {
"@type": "Offer",
"url": "https://example.com/trailhead-commuter-jacket",
"priceCurrency": "USD",
"price": "149.99",
"priceValidUntil": "2026-12-31",
"availability": "https://schema.org/InStock",
"itemCondition": "https://schema.org/NewCondition"
},
"aggregateRating": {
"@type": "AggregateRating",
"ratingValue": "4.6",
"reviewCount": "127"
}
}Note the additionalProperty block. Those three entries are what let a machine answer “is it waterproof enough for heavy rain, and does it pack down?”, questions your prose might answer but your markup otherwise would not.
What platforms give you.
Shopify, BigCommerce, and WooCommerce all generate basic product schema markup, and most themes handle name, price, and availability adequately.
What they rarely populate.
additionalProperty, gtin, and variant structure, which is exactly the part that matters for AI discovery. Check what your theme actually outputs rather than assuming.
How to validate.
Google’s Rich Results Test and the Schema.org validator, both linked from Google Search Central’s structured data documentation.
Five mistakes that cost you
Markup that disagrees with the page.
If schema says $149.99 and the visible price is $129.99, you have introduced ambiguity at the worst possible moment, and Google treats mismatches as a quality signal. Generate schema from the same data source that renders the page.
Stale availability.
InStock on a sold-out product produces recommendations you cannot fulfill. This needs to update in real time, not nightly, the same discipline covered in our real-time indexing guide.
Flattened variants.
One Product entry covering fifteen size-and-colour combinations gives wrong prices and wrong availability. Use ProductGroup and hasVariant.
Marking up reviews you don’t have.
Fabricated or aggregated-from-elsewhere ratings are a policy violation and a genuine legal exposure. Only mark up reviews genuinely collected on your own site.
Treating it as set-and-forget.
Schema requirements change, platforms update themes, and a migration can silently drop your markup. Validate a sample monthly rather than assuming it still works.
Each of these produces markup that technically validates while quietly telling machines the wrong thing, which is worse than having none at all.
What product schema markup won’t do
Worth being direct, because expectations run high.
It won’t rank you.
Structured data helps machines understand what is there. It does not make thin content good. A perfectly marked-up page with a three-word description still loses to a well-described competitor.
It won’t guarantee rich results.
Google decides whether to display them, and eligibility is not entitlement.
It won’t guarantee AI citation.
It removes ambiguity at the filtering stage, which meaningfully improves your odds, but retrieval quality and content depth still decide the outcome. Our answer engine optimization guide covers the wider picture.
It won’t fix crawl access.
If AI crawlers cannot reach the page, the markup is invisible regardless. Check access first. Our AI crawlers guide has the diagnostic.
The honest framing: product schema markup is necessary infrastructure, not a growth lever on its own. It makes good content legible. It does not make weak content competitive.
Auditing what your platform already outputs
Most teams assume their platform handles this. Running an audit usually corrects that assumption quickly, and it takes under an hour.
Pull ten representative pages.
A simple product, a heavily-varianted one, a bundle, a sale item, an out-of-stock item, and a few from your long tail. The edge cases are where product schema markup breaks.
Run each through the Rich Results Test.
Note what is present, what is missing, and what throws warnings. Warnings are not errors, but they usually indicate a property that would help.
Check four things specifically.
Does the price in markup match the visible price, including on sale items? Does availability reflect real stock? Are variants structured or flattened? Are any attributes present beyond the basics?
Compare against your feed.
Your Merchant Center product data and on-page markup should agree. Disagreement between them is a quality signal you do not want to send, and it is common after a catalog migration.
What teams typically find: name, price, and availability handled adequately by the theme; gtin and mpn missing; variants flattened into a single entry; and additionalProperty entirely absent. That last gap is the one that matters most for AI discovery, since it is where the specifications live that let a machine match your product to a specific request.
One practical warning. Product schema markup generated by a plugin separately from your page template will drift, because the two read different data sources. Generate it from the same source that renders the page, or accept that they will disagree after the next price change.
Beyond Product: schema worth adding
Beyond product schema markup, three other types earn their place on an ecommerce site.
Organization
On your homepage: name, logo, contact points, and social profiles. This helps machines resolve your brand as an entity rather than a string, which matters for consistency across AI answers.
BreadcrumbList
On category and product pages, clarifying your site hierarchy and frequently appearing in search results.
FAQPage
On guides and support content. Despite the rich-result deprecation, the retrieval value holds, with the caveat that each answer must be self-contained to be quotable.
Our product findability guide covers the merchandising work that pairs with this technical foundation.
Schema for category and collection pages
Product pages get the attention, but category pages are where a lot of discovery traffic lands, and product schema markup alone does not serve them.
ItemList
Describes the products on a category page in order, which helps machines understand the page as a curated set rather than a wall of links. Useful for best-of and collection pages where the ordering carries meaning.
CollectionPage
Identifies the page type explicitly, distinguishing a browsable collection from an article or a single product.
BreadcrumbList
Clarifies where the page sits in your hierarchy, genuinely useful for both search display and for machines resolving your site structure.
One caution worth naming: do not apply full product schema markup to every item in a category listing. Marking up twenty products on a collection page with complete Offer blocks creates ambiguity about which product the page is actually about, and it can suppress rich results rather than earning them. Use ItemList for the collection and reserve full Product markup for the individual product pages.
For faceted category pages generated by filters, be deliberate about which get indexed at all. A store with thousands of filter combinations generating thousands of thin marked-up pages is producing noise, not visibility.
An implementation sequence
Week 1 – audit what exists.
Run ten representative product pages through the Rich Results Test. Most teams discover their theme outputs less than they assumed, particularly on variants and attributes.
Week 2 – fix the required properties.
Name, image, price, currency, availability across the whole catalog. This is the baseline.
Week 3 – add attributes.
Populate additionalProperty for the specifications shoppers filter and ask on. Prioritize your long-tail products, which have the most to gain and usually the least data.
Week 4 – structure variants and validate.
Move to ProductGroup where appropriate, then validate a sample and set a monthly recheck.
Then align your feed.
Your Merchant Center feed and on-page markup should agree. Given Google’s new AI insights reporting on attribute completeness, that consistency now has a visible payoff.
Frequently asked questions
What is product schema markup?
Product schema markup is structured data using the schema.org vocabulary that labels your product information so machines understand it explicitly: name, price, currency, availability, brand, and attributes. It powers rich results in search, populates shopping surfaces, and helps AI assistants extract facts when composing recommendations.
Which product schema properties are required?
For rich result eligibility: name, image, and an offers block with price, priceCurrency, and availability. Strongly recommended additions are description, sku, brand, gtin or mpn, and genuine aggregateRating. For AI discovery, additionalProperty entries carrying specifications matter most.
Did Google remove FAQ schema in 2026?
Google phased out the FAQ rich result, so FAQ markup no longer produces the expandable Q and A display for most sites. The markup still carries retrieval value, since research finds FAQ sections correlating with higher AI citation rates, so implement it for that reason rather than for a SERP feature.
Does product schema markup improve rankings?
Not directly. It helps machines understand what is on the page, which can improve click-through via rich results and improves your odds of AI citation by removing ambiguity. It does not make thin content competitive, since a well-marked-up page with a weak description still loses.
How do I handle product variants in schema?
Use ProductGroup with hasVariant rather than one flattened Product entry. Size and colour variants have different SKUs, prices, and availability, and collapsing them into a single entry produces wrong answers for shoppers and machines alike.
How do I validate product schema markup?
Use Google’s Rich Results Test and the Schema.org validator. Test a representative sample rather than one page, and recheck monthly, since platform theme updates and migrations frequently drop markup silently.
Does schema markup help with AI assistant recommendations?
It helps meaningfully by stating price, availability, and attributes explicitly rather than leaving a model to infer them from prose. It does not guarantee citation, since retrieval quality and content depth still decide, but it removes ambiguity at the stage where products get filtered out.
Structured data is how machines read your catalog.
bCloud AI works from the same structured product data that powers AI discovery, so clean attributes improve both at once.
bcloud.ai