aggregateRating (schema property)

How aggregateRating works in JSON-LD

{
  "@type": "Service",
  "name": "Resocial SEO Services",
  "aggregateRating": {
    "@type": "AggregateRating",
    "ratingValue": "4.8",
    "reviewCount": "47",
    "bestRating": "5",
    "worstRating": "1"
  }
}

The properties: ratingValue (the score), reviewCount (total reviews), bestRating and worstRating (the scale).

Why aggregateRating matters

When valid, Google sometimes displays the review stars and rating in SERP, visible to users before they click:

Resocial SEO Services
★★★★★ 4.8 (47 reviews) · resocial.us/services

CTR studies show 15-35% lift on listings with visible review stars vs without. This is a real conversion lever.

Critical: only ship aggregateRating with a verifiable review system

Google’s guidelines require that aggregateRating must be:

1. Aggregated from actual reviews present on the page
2. The reviews must be from real users (or a verifiable source)
3. The reviews must be visible to humans, not hidden in markup-only

Shipping aggregateRating without these conditions is a guidelines violation. Penalties include:

• Removal of rich results
• Manual action (“Spammy structured data”)
• In severe cases, ranking demotion

Several agencies still ship fake-rating schema. It works in the short term, fails badly when caught.

The right way to ship reviews

Two patterns that work:

Pattern A: Native review system

• Page has a review submission form
• Reviews stored in database
• Reviews displayed visibly on the page
• aggregateRating calculated from displayed reviews

Pattern B: Third-party aggregator import

• Connect to Yotpo, Bazaarvoice, Trustpilot, G2, Capterra
• Reviews flow back to your site (verifiable source)
• aggregateRating calculated from synced data

Both produce legitimate rich results. Neither violates guidelines.

What types support aggregateRating

Most relevant for SEO:

• Product (ecommerce)
• Service (B2B services)
• LocalBusiness and sub-types (local SEO)
• Recipe, Movie, Book, Course (content categories)
• Organization (limited use; usually better on Service/Product)

Article does NOT support aggregateRating in Google rich results.

Common implementation mistakes

• Hardcoded 5.0 stars with no real reviews
• Schema rating doesn’t match visible page rating (deceptive)
• Reviews are hidden behind authentication
• Rating count is suspiciously high relative to traffic
• Reviews are clearly self-generated (similar wording, all 5 stars)

Google’s spam detection catches all of these patterns. Penalty risk: real.

Resocial perspective

We ship aggregateRating ONLY for clients with verifiable review systems. The work isn’t the schema, it’s connecting to or building the underlying review infrastructure first. Schema-only reviews are an anti-pattern we audit out of every engagement. See the Schema Markup Complete Guide for the full implementation playbook.