Structured data testing and validation tools

Phase 2 made the case for structured data; this lesson is the practical follow-up: how to check what your store actually outputs. The tools are free, the check takes a minute per page, and the findings are usually concrete.

Rich Results Test: the one that matters

Google's Rich Results Test (search.google.com/test/rich-results) fetches any URL and reports which rich-result types it found and whether they're eligible - for product pages, that's Product snippets and Merchant listings. Paste your bestseller's URL. Three outcomes:

Valid: schema present, complete, eligible for rich results. Note which fields it found - price, availability, ratings - because eligible isn't the same as maximal.

Valid with warnings: the interesting case. Warnings are missing optional fields - no aggregateRating (you have reviews but the review app isn't outputting schema?), no GTIN, no shipping details. Each warning is a specific, assignable improvement.

Errors: broken markup - often a legacy app or theme customization outputting conflicting schema. Errors can disqualify the page from rich results entirely, which makes them quietly expensive.

Schema.org validator: the second opinion

The Rich Results Test only checks what Google can show as rich results. The schema.org validator (validator.schema.org) shows everything your page declares, Google-eligible or not. Useful for seeing the full picture - including markup that matters to other consumers of structured data, like AI systems (a thread we pick up in the AI Search phase).

Search Console: validation at catalog scale

Testing pages one by one doesn't scale to two thousand products. GSC's Shopping and enhancement reports (under Indexing/Experience depending on layout) aggregate structured-data health across every indexed page - how many product pages have valid markup, and which errors repeat. Because Shopify pages are template-driven, errors almost always come in batches: one template bug, eight hundred affected URLs. Which is the good news - one fix, eight hundred pages healed.

The habit

Test one URL per template (product, collection, homepage) after any theme change or app installation - apps injecting broken schema is a classic silent regression. And glance at the GSC report monthly. Structured data breaks quietly; these tools are how you hear it.