AggregateRating

The average rating derived from all aggregated individual ratings. This is the primary numeric value displayed in rich results.

• Data Type: Number or Text
• Requirement: Required for rich results
• Scale: Typically 1-5, but can be defined by bestRating
• Example: A product with 4.2 out of 5 stars uses "ratingValue": "4.2"

Search engines use this value to render the visual star rating in SERPs. The value must be a precise representation of the calculated average, not a rounded integer, to maintain accuracy.

The total number of individual reviews or ratings that contributed to the aggregate score. This property signals statistical significance and social proof.

• Data Type: Integer
• Requirement: Strongly recommended
• Best Practice: Update dynamically as new reviews are submitted
• Example: "reviewCount": "847" indicates 847 people submitted a rating

A high review count increases user trust and click-through rate. Search engines may suppress rich results if the count appears artificially inflated or static.

These properties define the upper and lower bounds of the rating scale, allowing the system to normalize values across different scoring systems.

• bestRating: The highest possible value (e.g., 5, 10, 100)
• worstRating: The lowest possible value (e.g., 1, 0)
• Default: If omitted, the scale is assumed to be 1-5
• Example: A 10-point scale uses "bestRating": "10" and "worstRating": "1"

Explicitly defining the scale prevents misinterpretation when a site uses a non-standard rating system, such as a percentage-based score.

The entity that is being rated, specified as a nested Schema.org type. This property establishes the semantic relationship between the rating and the subject.

• Data Type: Thing (typically Product, LocalBusiness, CreativeWork, or Organization)
• Requirement: Required for valid markup
• Nesting: Embed the full entity markup within this property
• Example: "itemReviewed": { "@type": "Product", "name": "Widget X" }

This linkage is critical for entity reconciliation. It tells the search engine exactly which product, service, or business the aggregate score applies to, preventing attribution errors.

The total number of ratings that contributed to the aggregate score, distinct from reviewCount which counts full-text reviews. Use this when users can submit a score without writing a review.

• Data Type: Integer
• Distinction: ratingCount = simple scores; reviewCount = written reviews
• Use Case: Star-only rating widgets without comment fields
• Example: A product with 1,200 star ratings but only 340 written reviews uses both properties

Providing both counts gives search engines a more granular understanding of user engagement. If your system only collects one type, use the property that matches your data model.