Overview
Introduction
Review schema marks up a single review, complete with who wrote it, what it's about, and a rating, so search engines can potentially show it as a review snippet. It has three separate nested objects (itemReviewed, author, reviewRating), which is more structure than most people want to write by hand for one review.
This generator collects the handful of fields a review actually needs and assembles the correctly-nested JSON-LD, including toggling the type of thing being reviewed (a Product, a LocalBusiness, a Book, and so on) without having to remember schema.org's exact type names.
What Is Review Schema Generator?
A form-driven generator for Review JSON-LD: what's being reviewed (with a type toggle for common reviewed-item types), the reviewer's name, a rating with a configurable scale, review body text, and an optional publish date.
It produces one Review node per use; for summarizing many reviews into a single average, use aggregateRating on the item's own schema (for example, the Product Schema Generator's rating fields) instead.
How Review Schema Generator Works
The reviewed item becomes a nested itemReviewed object using whichever type you selected (Product, LocalBusiness, Book, Movie, or Recipe) with just a name, since the full details of that item usually already live in its own separate schema elsewhere on the page.
The rating fields build a nested reviewRating object with ratingValue plus bestRating/worstRating (defaulting to a 5-point scale, 1 to 5, if left blank), and the author name is nested as a Person object.
When To Use Review Schema Generator
Marking up a single detailed review that's genuinely published on the page, written by a real reviewer.
Adding editorial or expert reviews to a product, restaurant, book, or recipe page alongside the item's own schema.
Often used alongside Structured Data Validator, Product Schema Generator and Local Business Schema Generator.
Features
Advantages
- Handles all three levels of nesting (itemReviewed, author, reviewRating) from flat form fields, avoiding a common source of hand-written JSON-LD errors.
- Lets you set a custom rating scale (bestRating/worstRating) instead of assuming every site uses a 5-point scale.
- Supports the five most commonly reviewed schema.org types via a simple toggle, without needing to know their exact type names.
Limitations
- Only builds a minimal itemReviewed reference (a name and type); it doesn't pull in or merge with a separate, more detailed Product/LocalBusiness schema elsewhere on the page.
- Doesn't verify the review is genuinely visible and attributable on the page, which is a strict requirement of Google's review snippet policies, not just a best practice.
Examples
Best Practices & Notes
Best Practices
- Only mark up reviews that are genuinely written by a third party and visible on the page - never generate this for marketing copy or a review you wrote about your own product.
- Set bestRating/worstRating explicitly whenever your site doesn't use a plain 1-5 scale, to avoid the rating being misinterpreted.
- Keep the reviewBody text here identical to what's actually displayed on the page, since mismatches between visible content and structured data violate Google's policies.
- Pair this with the item's own full schema (Product, LocalBusiness) elsewhere on the page rather than relying on this generator's minimal itemReviewed reference alone.
Developer Notes
bestRating and worstRating default to "5" and "1" respectively when left blank, matching schema.org's own documented default scale, but they're always emitted explicitly in the output rather than omitted, since relying on an implicit default is what most commonly causes rating-scale misreads downstream.
Review Schema Generator Use Cases
- Adding a single expert or editorial review to a product page
- Marking up a customer testimonial that includes a numeric rating
- Publishing a book, movie, or recipe review with structured rating data
Common Mistakes
- Marking up a self-written marketing blurb as a Review, which Google's guidelines treat as a policy violation when discovered.
- Leaving bestRating/worstRating at the 1-5 default while entering a rating on a different scale (like 8/10), producing a rating that reads as almost-perfect when it isn't.
Tips
- If you're marking up several reviews for the same item, generate each one separately and place them as separate <script> blocks or entries in an @graph array, rather than trying to merge them into one Review node.
- Run the output through the Structured Data Validator tool afterward to confirm itemReviewed, author, and reviewRating all came through.