The properties Google requires, and what they rest on

Key findings

• 30 of the 32 live rich result features need at least one required property.
• Exactly one required property, lender on Book, sits on fewer than 1,000 websites.
• 6 live features rest on a property Schema.org has not ratified yet.
• 82.86% of Schema.org's 1,529 properties are named by no documented rich result.
• applicableCountry and returnPolicyCategory sit on 100,000 to a million websites while still listed as pending.

Only one required property sits on fewer than 1,000 websites

July 2026 · six ranges, no exact numbersGoogle and the Schema.org community, Schema.org usage statistics dataset. Websites in Google's index.

The table lists the required properties of all the live features, 170 rows in all, with their usage bands and the Google page the requirements came from.

168 of the 169 measured required properties already sit above Google's smallest band, because a feature that requires what nobody uses barely works for anyone. Book breaks the pattern. It requires lender for its library-borrowing action, and lender sits on fewer than a thousand websites. Build a Book rich result exactly as documented, and you are implementing a property with almost nobody else's markup to check yours against. I went looking for more cases and found one. That is reassuring. It means the required list is safe to follow almost everywhere, and Book is the one place I would test the markup twice.

That does not make Book rare or the requirement impossible; one property on one feature simply has almost no other implementation to learn from.

6 features rest on properties Schema.org calls pending

The table lists all 10 required properties that are still pending, with the feature they belong to.

applicableCountry and returnPolicyCategory already sit on 100,000 to a million websites, and Schema.org still lists both as pending. Return policy carries 4 of the 10, applicableCountry, returnPolicyCategory, merchantReturnLink, returnPolicySeasonalOverride, more than any other feature. So returns markup already relies on wording Schema.org says could change, which is a reason to expect a revisit, not a reason to avoid it. If you sell online, that revisit is yours to plan for, and I would plan for it, not around it.

Schema.org warns that pending terms "are subject to change and should be used with caution"; its full description is quoted in the method section. You can also read Schema.org's own documentation for the pending section.

82.86% of properties have no documented rich result

Below I split the 1,529 properties Schema.org publishes by whether any documented rich result, required or recommended, names them; 262 are named.

Google documents rich results for a curated set of page types, recipes, jobs and events among them. Most of what Schema.org describes was not built around one of those, so a property with no documented rich result is one Google has not built a feature around, not one that is wrong to use. This is the figure I expected to be large. It is. Search engines can still read markup that earns no rich result, so for your site most of the vocabulary earns nothing visible, and that is not a reason to leave it out.

Neither dataset publishes the exact count of properties with no documented rich result, so if you want it, subtract 262 from 1,529.

Known limits of this data

• Google's required-property tables list one thing that is not a Schema.org property at all, xPath on the speakable feature. I keep it in the required-property table because Google asks for it, and it carries no usage figure, because there is nothing in Schema.org for it to have one against. It is not counted in any of the band figures on this page.
• Names with a dot in them, such as offers.price and ListItem.position, are Google notation for a property nested inside another type rather than Schema.org terms in their own right. The usage figure against them is for the last part of the name.