Product Structured Data on WooCommerce: Seven Failures Worth Checking
Variations with no offers, availability that contradicts the page, invented identifiers and duplicated JSON-LD. What to verify on three products instead of trusting the theme.
WooCommerce outputs Product structured data by default, and that default is wrong often enough that rich results fail silently. The markup looks present in the page source, so nobody investigates until a product shows no price in search.
These are the failures worth checking, in order of how often they occur.
1. Variations with no offers
A variable product should either expose an AggregateOffer with a price range, or list each variation as its own offer. What commonly happens: the theme outputs a Product with a single Offer whose price is the parent product’s price — which on a variable product is often empty or a placeholder.
Result: price missing, or a price that does not match anything a customer can buy.
Decide once and implement consistently:
- AggregateOffer — one entity,
lowPriceandhighPricefrom the variation range,offerCountmatching the number of purchasable variations - Per-variation Product — each variation as a separate entity with its own SKU and price
What not to do: output both. Duplicate entities with conflicting prices cause validation errors and teach search engines to distrust the page.
2. Availability that does not match the page
The most common mismatch is a product that shows “in stock” in the markup because stock management is disabled, while the page shows “out of stock” from a custom field. The other direction also happens after a migration, when stock is tracked at a different level than before.
The mapping is narrower than most people assume: InStock, OutOfStock, PreOrder, BackOrder. Everything else is invalid. A custom status like “made to order” needs a decision, not a new value.
3. Missing identifiers
Google wants gtin, mpn and brand for most product categories. Handmade, private-label and custom items often have none of them.
The honest solution is identifier_exists: false, not fabricated values. Inventing GTINs is a fast route to a suspended Merchant Center account, and the same data usually feeds both channels.
SKU is useful but not a substitute — it is your identifier, not a global one.
4. Reviews marked up without being reviews
Aggregate ratings are attractive in search results and therefore tempting. Two failures follow:
- Self-serving reviews. Ratings on your own organisation, marked up as reviews, are against Google’s guidelines and get rich results removed
- Empty or placeholder reviews. An
AggregateRatingwith review count 0, or reviews with no author, fails validation
If a product has no reviews, output no rating. If it has three, output three. Padding a count to look established is the kind of thing that is checked when someone complains.
5. Price that differs from the visible price
Structured data must match what the page shows. This breaks in three specific ways:
- Multi-currency stores outputting the base currency while the page displays a converted one — see the multi-currency considerations
- Sale prices output as regular price, so the markup shows a higher number than the page
- Tax-inclusive prices on the page, tax-exclusive in the markup
The rule is simple: whatever an anonymous visitor sees, including the currency, must be in the markup.
6. Organisation and Product colliding
A store homepage often carries both Organization and a Product or ItemList. That is fine, but the two need to be connected with stable @id values, and the brand reference inside the product must point at the organisation rather than duplicating it.
Where this goes wrong: two separate entities both claiming to be the organisation, with different names or URLs. That is a signal conflict, and it weakens entity recognition — the same problem described in the note on technical SEO.
7. JSON-LD injected twice
A theme, an SEO plugin and a schema plugin can each output product markup. The page validates, but search engines see two products with different prices, and behaviour becomes unpredictable.
Check by viewing the page source and counting application/ld+json blocks that describe a Product. One is correct.
How to verify without guessing
- Pick three products: a simple one, a variable one with many variations, an out-of-stock one
- View source and extract each JSON-LD block
- Compare price, currency, availability and identifiers against what the page renders
- Run the rich results test on all three
- Check the same three in the language versions, not just the default
The fifth step catches something specific: translated pages that keep the default-language price format, or lose the inLanguage property entirely. On a multilingual store, every language version needs its own valid markup — not a copy of the English one.
What good markup buys
It is worth being realistic. Structured data does not raise rankings directly. It affects how a result appears, which affects click-through, and consistent markup supports entity understanding over time.
The commercial reason to fix it is narrower: price, availability and ratings in search results change whether someone clicks. A product with a stale “out of stock” label in search is a customer who does not arrive. On a store with thousands of SKUs, that invisible filter is worth more than most on-page optimisation.
If your store has been through a migration or a theme change, the markup is the first thing to re-verify — it is the part of the page customers never see and search engines always read.