Review schema is a structured data format that lets search engines understand and display the reviews associated with a product, a service, a creative work or a local business. It relies on a shared vocabulary, usually the one from Schema.org, and is added to the page code as machine-readable data. This way the review content is not confined to the visible text but becomes information the search engine can interpret, aggregate and use to enrich search results.
In the context of search engine optimization, review schema is one of the so-called structured data types and contributes to generating rich results. A page that correctly exposes a review can show additional elements in search results, such as the average rating, the number of reviews and, in some cases, the reviewer’s name. The goal is not just visual: a richer presentation can increase visibility and click-through rate, as long as the data is truthful and complies with search engine guidelines.
Structured data and review schema: what they are
Structured data is a set of rules and vocabularies that let you describe the content of a page in a standardized way. Schema.org is the most widely used vocabulary and defines types such as Product, Recipe, Book, Movie and LocalBusiness, each of which can be associated with reviews. Review schema is therefore the application of this vocabulary to the specific case of reviews.
According to Schema.org, a review is a type of entity that describes an author’s evaluation of an item. The typical elements are the item reviewed, the author, the rating expressed as a score or as stars, the review text and the date. When these elements are coded correctly, the search engine can link them to the main entity of the page and show a rich result.
Single review and average rating: the two structures
Review schema describes a single review, with an author and a text. AggregateRating, on the other hand, summarizes a set of reviews through an average rating and a total number of ratings. The two structures are complementary: a product page can contain several individual reviews and an aggregateRating that sums them up. To get a rich result with stars, you generally need a valid aggregateRating.
JSON-LD, microdata and RDFa: which to choose
There are several formats for adding structured data. The JSON-LD format is placed in a block inside the page and is not visible to the user; it is the format recommended by the main search engines. Microdata and RDFa, on the other hand, are integrated directly into the attributes of the visible HTML code. All formats can be valid, but the choice affects maintenance and code cleanliness.
The elements that make up a review schema
A well-built review schema contains some recurring elements. The first is the item reviewed, indicated with a property that links the review to the main entity. The second is the author, which can be a person or an organization. The third is the rating, expressed through a numerical score and a reference scale. Finally there is the review text and the publication date.
The rating deserves particular attention. Schema.org lets you express the score as a numerical value, as a percentage or as a number of stars. The scale must be declared consistently, for example from one to five. A common mistake is using different scales within the same page, creating ambiguity that the search engine can flag as non-compliant.
Who signs the review and why it matters
The review author contributes to the credibility of the content. Where possible, it is worth indicating a real name or a verifiable profile, avoiding fictitious identities. Search engine guidelines discourage self-serving reviews, meaning ratings that a business gives itself. Correctly indicating the author helps distinguish an authentic review from a manipulated one.
The markup types compared
| Format | Visibility in the code | Maintenance | Recommendation |
|---|---|---|---|
| JSON-LD | Separate block, not visible | Simple to update | Recommended |
| Microdata | Integrated into the visible HTML | Can become scattered | Accepted |
| RDFa | Integrated into attributes | Requires precision | Accepted |
| No markup | Absent | None | Not eligible for rich results |
From the main entity to markup validation
Implementation requires order and verification. First you identify the main entity of the page, then you collect the review data, then you choose the format and validate the result. Each step affects the final quality of the markup and whether the search engine accepts it.
- Identify the main content type, for example a product, a book or a local business.
- Collect the review data: author, rating, text and date.
- Choose the Schema.org vocabulary and the appropriate entity type.
- Define the rating scale consistently and declare it.
- Add the markup in the chosen format, preferably JSON-LD.
- Link the review to the main entity through the correct property.
- Validate the code with structured data testing tools.
- Fix any errors and repeat the check after every change.
Inconsistent data and duplicate reviews: what to avoid
The most frequent mistakes involve inconsistent data, such as a rating that does not match the visible text, or a missing author. Another problem is the presence of duplicate or automatically generated reviews, which search engines can ignore or penalize. Improper use of aggregateRating, applied to content that does not collect real ratings, is also a violation of the guidelines.
Best practices suggest keeping the markup aligned with the page content, updating the data when it changes, and not promising rich results that depend on algorithmic decisions. Review schema is a tool for clarity, not a guarantee of visibility. Content quality remains the main factor, while markup makes it easier for search engines to understand it.
Frequently asked questions about review schema
What exactly is a review schema?
It is a set of structured data that describes a review according to the Schema.org vocabulary, indicating the item reviewed, the author, the rating and the text. It makes the review understandable to search engines.
What is the difference between review and aggregateRating?
Review describes a single review with author and content, while aggregateRating summarizes multiple reviews with an average rating and a total number. They are often used together on the same page.
Which format is preferable for implementing it?
The JSON-LD format is generally the most recommended, because it separates the data from the visible content and simplifies maintenance. Microdata and RDFa are also accepted if implemented correctly.
Does review schema guarantee a rich result?
No, there is no guarantee. Correct markup makes the page eligible, but the decision to show the rich result is up to the search engine and depends on various factors.
Can you add self-serving reviews?
Guidelines discourage reviews that a business gives itself. It is preferable to indicate real, verifiable authors to maintain the credibility of the content.
How do you check that the markup is correct?
You use the structured data testing tools provided by search engines. These tools flag errors and warnings, letting you fix the code before publication.
From the glossary to the projectDo search engines understand what your site is about?Implementing structured data and Schema markup helps search engines and AI systems interpret your content.Discover structured data
Related entries:
Reviews Schema.org Schema markup Offer schema Product schema