What is the Schema.org Review type (review schema)?
Review is the Schema.org type that describes a specific review of an item: who writes it, what it rates, the score it gives and its text. On a product page, each customer review shown on the page can be marked up as a Review node inside the product, unlike the average of all of them, which is another type.
Key takeaways
- Review describes one specific opinion, with its author, rating and text; the average of many reviews is another type, AggregateRating.
- Google requires the author, the rating and what is being rated, which is omitted when the review sits inside the product.
- Only mark up reviews the customer sees on the page, about a specific item and born on your site, not gathered from others.
- Since July 2026, Google explicitly prohibits fake reviews and incentivized reviews without disclosure, in the markup too.
- Marked-up pros and cons only count on editorial review pages, not on a store's product page.
Under the photo and the price, many product pages have a list of reviews: "Fits me perfectly, the fabric is cool", four stars, Marta, August 12. Each of those reviews, when described for machines, is a Review node. It is a Schema.org type, written in English because it is a name in the vocabulary. Review is one specific review, with author, rating and text; not the average of all of them, which has its own type, AggregateRating. It matters to you because it is the part of your product page's structured data with the most rules about where it comes from: who wrote the review, whether it is visible and whether it was paid for.
How it works
Schema.org defines Review as a review of an item, and gives a restaurant, a movie or a shop as examples. It is a creative work (CreativeWork), like an article or a video, and its own properties say what is reviewed (itemReviewed), the text (reviewBody), the rating (reviewRating), the specific aspect being rated (reviewAspect) and the points for and against (positiveNotes and negativeNotes).[1]Schema.org. Review The author and the date it inherits from CreativeWork, as author and datePublished. Schema.org makes one clarification that avoids a mix-up: reviewRating is the rating the review gives the item; aggregateRating, applied to a review, would be the rating others give the review itself.[1]Schema.org. Review
Google uses Review for the review snippet, which can show stars and a summary in rich results or in the knowledge panel. It accepts reviews of books, courses, events, movies, recipes, apps and Product, among others, and of local businesses or organizations only when the site collects reviews about others.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data What it asks for in each review:[2]Google Search Central. Review snippet (Review, AggregateRating) structured data
| Property | What Google requires |
|---|---|
author |
Required. A person or organization with a valid name, under 100 characters. "50% off until Saturday" is not a name[2]Google Search Central. Review snippet (Review, AggregateRating) structured data |
itemReviewed |
Required if the review stands alone. If it sits inside the product with the review property, it is omitted: Google understands that the containing product is being rated |
reviewRating.ratingValue |
Required. The rating, with a decimal point: 4.4, never 4,4 |
bestRating, worstRating |
Recommended. If missing, Google assumes a scale of 1 to 5 |
datePublished |
Recommended. The date of the review, in ISO 8601 format |
There is a nuance few stores take advantage of: an individual review can go without a rating if the marked-up content has an author and a date.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data A review written without stars is still a valid Review.
The rules weigh as much as the properties. Google asks that marked-up reviews be visible on the page, with their text and rating; that they be about a specific item, not a category or a list; that, if you mark up several reviews, you also mark up their average; and that you not gather reviews or ratings from other sites.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data Since January 15, 2025 it recommends accepting only ratings that arrive with a comment and the author's name, and on November 12, 2025, in a clarification about nesting, it added that you should not use several ways of indicating what is being rated at the same time.[3]Google Search Central. Latest documentation updates On July 24, 2026 it added one more guideline: no fake reviews and no incentivized reviews without disclosure.[3]Google Search Central. Latest documentation updates The guideline gives two examples: reviews that don't come from a real experience with the product, and reviews written in exchange for money, discounts, vouchers or free products without saying so clearly and visibly.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data
In a store
A made-up example. A shirt's product page in an invented store shows two customer reviews, with name, date, stars and text, and the average of the two. The product node carries the two reviews and their average:
{
"@context": "https://schema.org",
"@type": "Product",
"name": "Calella linen shirt, blue",
"review": [
{
"@type": "Review",
"author": { "@type": "Person", "name": "Marta G." },
"datePublished": "2026-08-12",
"reviewRating": { "@type": "Rating", "ratingValue": "4" },
"reviewBody": "It fits me perfectly and the fabric is cool. It wrinkles quite a bit."
},
{
"@type": "Review",
"author": { "@type": "Person", "name": "Jordi P." },
"datePublished": "2026-09-03",
"reviewRating": { "@type": "Rating", "ratingValue": "5" },
"reviewBody": "I ordered the M and the size is the usual one."
}
],
"aggregateRating": {
"@type": "AggregateRating",
"ratingValue": "4.5",
"reviewCount": "2"
}
}
What the example shows. The reviews carry no itemReviewed because they sit inside the product.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data They carry no bestRating because the store scores from 1 to 5, the scale Google assumes.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data The names are what each customer signs on the page, as is. And because there are several reviews, the node also carries their average, with the same numbers the customer sees.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data If the page showed only the three latest reviews out of forty, the marked-up reviews would be those three, because what the customer can't read on the page isn't marked up; the average of the forty can go in the node if the page shows it.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data
What people often get wrong
Marking up pros and cons on your store's product page expecting them to show in Google. The positiveNotes and negativeNotes properties exist in Schema.org, but Google only shows pros and cons on editorial product review pages, not on a merchant's product pages or in customer reviews.[4]Google Search Central. Product snippet (Product, Review, Offer) structured data If you want to summarize what your customers say, the place for it is the page's text.
How to check it
- Search for
"@type": "Review"in the source code of a page with customer reviews and count the nodes. - Compare them with the reviews the customer sees: the same author, the same date, the same rating and the same text. A marked-up review that can't be read on the page doesn't belong.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data
- Check each one's author. It has to be the name of a person or organization, under 100 characters, not a slogan or a promotion.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data
- If there are several reviews, also look for
aggregateRatingon the same product, with the average and the count the page shows.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data - Run the URL through the Rich Results Test to see whether it detects the review snippet and with what errors or warnings.
Related concept: AggregateRating
Review and AggregateRating are marked up together and do not replace each other. Review is one opinion; AggregateRating is the average of many, with their count. Google asks that, if you mark up several individual reviews, you also mark up their average, and says that for an evaluation of an item made by many people the type is AggregateRating.[2]Google Search Central. Review snippet (Review, AggregateRating) structured data The stars with the number of reviews come from the second; a customer's text with their name, from the first. And before being marked up, what your customers write are product reviews, which have their own entry.
To go further, the structured data reading path orders the rest of the terms, and the AggregateRating entry explains the average, the scale and where the numbers can come from.
Synonyms
- English
- review schema, review markup, review structured data
- Español
- marcado de reseña, datos estructurados de reseña
Sources
- [1] Schema.org. Review. Accessed . ↑ ↑
- [2] Google Search Central. Review snippet (Review, AggregateRating) structured data. Accessed . ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑ ↑
- [3] Google Search Central. Latest documentation updates. Accessed . ↑ ↑
- [4] Google Search Central. Product snippet (Product, Review, Offer) structured data. Accessed . ↑
Change history
- First English version.
This definition is published under CC BY 4.0. Reuse this definition freely, crediting Clione with a link to this page.
Related terms
AggregateRating
What AggregateRating is in Schema.org, which properties Google asks for to show stars, which ratings you cannot mark up and how to check your product page.
Product
What the Schema.org Product type is, which properties it carries, what Google requires for product snippets and merchant listings, and how to check it.
Product reviews
What product reviews are, what Google's reviews system evaluates, and the rules for customer reviews.
UGC
What UGC (user-generated content) is in a store: customer reviews, photos and questions, how Google and ChatGPT read it and how to moderate it.
ProductGroup
What ProductGroup is in Schema.org, which properties Google asks for, how to mark it up on a single page or one page per variant, and how to check yours.
@graph
What @graph is in JSON-LD, what the W3C specification says, how it links a product page's product, store and site, and how to review your store's.
Make your catalog easy for bots to read
Clione fills every product page with what bots read best: the JSON-LD graph with its structured data, the meta description and the FAQs. You review and press Push; then we check your live store so it all stays published.