Is Google really ready to embrace JSON-LD for products?

Why does this announcement represent a turning point for e-commerce merchants?

Google has long favored microdata and RDFa format for structuring product page information. JSON-LD, though recommended by W3C and easier to implement, took time to be recognized as a viable format for all data types.

This announcement formalizes what many SEOs have already observed: JSON-LD is gradually becoming the standard for communicating with Google. The reason? It resides in a script block separate from the HTML, making it easier to maintain and reducing the risk of integration errors. No more need to twist the HTML code to squeeze itemprop attributes everywhere.

What does this gradual deployment actually mean?

When Google refers to expanding support, it means that not all Schema.org types are treated equally. Product, Review, Organization? Generally okay. But more specific properties like real-time inventory data or certain advanced attributes may still be overlooked.

The engine needs to learn to interpret each new data type, verify its relevance, and detect potential spam. This is not an instantaneous technical switch; it’s a machine learning process that takes time. Hence, the importance of monitoring what Google actually displays in the SERPs, not just what your schema validator confirms as syntactically correct.

Which product data is prioritized in this rollout?

Basic properties have been functioning for some time: name, image, price, availability. These are the fundamentals that allow for the display of rich snippets with stars and prices.

More complex data, such as multiple variants, detailed return conditions, or loyalty programs remain in an artistic blur. Google tests, measures the impact on user experience, and rolls out case by case. If your implementation doesn’t generate a rich snippet despite valid markup, it’s likely that this specific property is not yet utilized by the engine.

• JSON-LD becomes the recommended format for structured product data, easier to maintain than microdata
• Support is partial and gradual: not all Schema.org properties currently generate enriched results
• Syntactical validation does not guarantee utilization: you must test actual display in the SERPs
• Basic properties (price, availability, reviews) work well, advanced attributes remain uncertain
• Monitoring is essential to detect when Google starts utilizing new properties

Is this statement consistent with what we observe in the field?

Yes and no. The transition to JSON-LD is indeed well underway for classic data types. E-commerce sites that have migrated their microdata to clean JSON-LD generally find that it works as well or even better in terms of rich snippet display rates.

But the devil is in the detail of “not yet applied to all possible structured data.” Google remains deliberately vague about which properties are supported and which are not. We find ourselves doing reverse engineering: implementing, waiting, checking if it displays, iterating. It's frustrating for teams wanting to plan their developments neatly. [To be verified] for each new property you test.

Why doesn’t Google communicate a comprehensive support list?

Because this list is constantly changing. The machine learning behind the utilization of structured data evolves continuously. A property ignored today can become relevant tomorrow if Google decides to launch a new search feature.

There’s also an anti-spam struggle: officially announcing that a property is supported opens the door to mass abuse. Google prefers to remain vague to keep some leeway in detecting manipulations. The result: we work with partial documentation and must rely on practical experimentation.

What are the risks of implementing unsupported JSON-LD?

No direct technical risk. Google simply ignores what it does not understand or does not wish to utilize. Your additional JSON-LD markup will not penalize your site or significantly slow down the crawl, as long as the code remains clean and valid.

The real risk is investing development time for hypothetical benefits. If you spend three weeks structuring complex data that Google doesn’t display, that’s three weeks you could have spent on optimizations that would have had measurable impact. Prioritize what already works, and test the rest on a small scale in parallel.

What should you prioritize implementing on an e-commerce site?

Start with the basic Schema Product properties: name, image, description, offers (with price, priceCurrency, availability). This is the bare minimum to hope for rich snippets in search results. Without this, you do not exist in the race for clicks on commercial queries.

Then add the AggregateRating markup if you have customer reviews. Stars in the SERPs mechanically increase CTR; they are one of the most reliable quick wins in e-commerce SEO. Ensure that you comply with Google’s guidelines on reviews: they must be authentic, verifiable, and not synthetically generated.

How can you verify that Google is actually utilizing your JSON-LD?

The Rich Results Test from Google Search Console tells you if your markup is syntactically correct, but not if it will be displayed. The only real validation is to search for your products on Google and see what actually displays.

Use the site:votredomaine.com operator combined with the exact name of your products to force the display. If after several weeks of indexing you still do not see a rich snippet while your markup is valid, it means Google has chosen not to utilize it. This could be a matter of perceived content quality, competition on the query, or simply that this property is not yet supported.

What implementation errors most often block display?

The inconsistencies between visible HTML and JSON-LD remain the classic trap. If your displayed price on the page is €99 but your JSON-LD states €89, Google will ignore your markup or worse, consider that you are trying to manipulate.

Missing or invalid data also poses problems: a non-loading image URL, a non-standard currency, an availability status that does not match the Schema.org vocabulary. Google is more demanding than a simple syntactical validator. The engine cross-references your structured data with what it actually sees on the page, and any suspicious divergence nullifies the rich snippet.

• Implement Product, Offer, and AggregateRating in JSON-LD on all product sheets
• Ensure strict consistency between structured data and visible HTML content
• Test actual display in the SERPs, not just syntactical validation
• Regularly monitor error reports in Search Console
• Document which properties actually generate rich snippets for your sector
• Prioritize basic properties before experimenting with advanced attributes