Should you really mark up each customer review individually with structured data?

What does Google really say about review markup?

Google allows the use of an aggregate rating schema rather than individual markup for each review. Specifically, instead of marking up 150 reviews one by one with Review, you can use AggregateRating with an average score and a total number of reviews.

This approach applies when the volume of reviews is significant. Google does not specify an exact threshold, but the implication is clear: if you have 5 reviews, mark them up individually. If you have 300, the aggregate rating becomes acceptable and even recommended.

What is the difference between Review and AggregateRating?

The Review type marks an individual review with its author, rating, text, and date. It is granular but time-consuming to maintain. The AggregateRating type synthesizes all reviews into an average score, a total number of ratings, and possibly a range of scores.

This distinction is crucial for rich snippets. An AggregateRating displays stars and the review count in the SERPs. An individual Review can generate a review carousel if Google decides to display it, but this is not guaranteed.

Why does Google favor aggregation for large volumes?

The technical reason is clear: to avoid structured data overload in the HTML. A product with 500 individually marked reviews increases code weight, slows crawling, and complicates parsing on the search engine side.

The other reason relates to signal quality. Google wants a representative score, not a dump of 500 reviews where 80% are never displayed. Aggregation forces sites to consolidate a clear metric rather than spam marking.

• One AggregateRating per product/page suffices if the volume of reviews exceeds a few dozen.
• Individual markup remains relevant for editorial reviews or sites with little UGC.
• Google does not provide a precise quantitative threshold for switching from one mode to another.
• The AggregateRating must reflect real and verifiable reviews, not an invented average.
• The number of reviews (reviewCount) is mandatory in the AggregateRating schema to prevent abuse.

Does this statement align with observed practices on the ground?

Yes, but with significant gray areas. Most major e-commerce sites have been using AggregateRating for years without any issues displaying stars. Amazon, Cdiscount, Fnac: all aggregate. None individual mark up 10,000 product reviews.

However, Google remains deliberately vague on the threshold that justifies aggregation. Three reviews? Ten? Fifty? This lack of precision forces practitioners to interpret. My observation: below 10 reviews, individual markup remains valuable to drive potential carousels. Above 30, aggregation becomes the norm.

What nuances need to be added to this rule?

Google says “significant volume” but never quantifies it. This is typical of their communication: a soft rule without verifiable metrics. [To be verified]: Does the threshold vary by sector? Do product reviews vs. local reviews vs. book reviews follow the same logic? There is no official data on this.

Another critical point: aggregation works for traditional rich snippets, but some Google formats (like the review carousel in Google Shopping or Discover) may prioritize individually marked reviews. If your strategy relies on these specific placements, aggregation alone may limit visibility.

Finally, beware of the risk of devaluation. If you mark up an AggregateRating with 500 reviews but Google crawls your page and only detects 20 visible reviews, it might consider the markup misleading. Consistency between HTML and structured data remains a validation criterion.

In what cases does this rule not apply?

For editorial reviews or expert opinions, individual markup remains preferable. A product test written by a journalist has more SEO value if marked as Review with complete author, datePublished, and reviewBody. Google can then associate the review with an author entity and enhance E-E-A-T.

The same goes for third-party comparison or aggregator sites (Trustpilot, Avis Vérifiés, etc.). Their model relies on displaying detailed reviews. An AggregateRating alone is not enough to generate the rich snippets they aim for. They must individually mark for the review carousels.

What should you do concretely on a site with many reviews?

Implement AggregateRating at the product or main entity level. The schema should include ratingValue (average score), bestRating (generally 5), worstRating (generally 1), and reviewCount (total number of reviews). These four properties are mandatory for displaying stars.

If you have a dedicated customer reviews page or a reviews tab, you can combine AggregateRating on the product with individual Review markup for the 3-5 most recent or best-rated reviews. This hybrid approach maximizes visibility chances in different Google contexts.

Also consider the freshness of the markup. If your AggregateRating shows 4.3/5 with 120 reviews but the data is six months old, Google may ignore the markup. Set up an automatic updating system with each validated new review.

What errors should be avoided with AggregateRating?

Never invent a score or number of reviews. Google can cross-check structured data with visible HTML content. If your markup states 200 reviews but your page displays only 15, you risk a manual action for misleading structured data.

Also avoid duplicating markup. A product should only have one AggregateRating. If you use multiple structured data scripts (WordPress plugin + custom code), ensure they do not generate two competing schemas. Google will validate the first and ignore the second, or worse, reject both.

Finally, do not mix contexts. An AggregateRating on a category page aggregating ratings from 50 different products makes no sense to Google. The markup should focus on one unique entity: a product, service, business, or location.

How can I verify that my implementation is correct?

Use Google’s Rich Results Test (not the old Structured Data Testing Tool, which is outdated). Paste the URL of your product page and check that Google properly detects the AggregateRating with all required properties. If a warning appears, correct it before deployment.

You should also check the Search Enhancements report in Google Search Console. It lists the markup errors detected during crawling. If your AggregateRatings generate widespread errors, you have a structural issue (incorrect reviewCount calculation, missing bestRating, etc.).

For complex sites with thousands of products, these structured data optimizations can become technical and time-consuming. If you lack internal resources or still see your stars not appearing in the SERPs despite seemingly correct markup, consulting a specialized SEO agency can save you time and ensure proper implementation.

• Implement AggregateRating with mandatory ratingValue, bestRating, worstRating, and reviewCount
• Check consistency between the number of marked reviews and the number displayed in the HTML
• Automatically update structured data with each validated new review
• Test the implementation with the Rich Results Test and monitor Search Console
• Avoid competing schema duplication on the same page
• Reserve individual Review markup for editorial reviews or pages with little UGC