About & methodology

What it is

Schema.org, in collaboration with Google, publishes a public dataset on the real-world usage of vocabulary terms (types like Person or properties like price) across millions of web domains. This dashboard offers a visual, observatory-style reading of it.

How the data is gathered

• Gather: term frequencies are measured within Google's public web crawling infrastructure and aggregated per domain (not per page): using a term on 100 pages of the same site counts as one domain.
• Group: instead of exact (noisy) numbers, sites are grouped into popularity tiers for stability and privacy.
• Publish: a new file is published on GitHub every month.

The domain tiers

Each term is classified into one of these tiers, from the lowest to the highest number of unique domains:

The < 1K tier includes both brand-new terms and highly specialized ones (e.g. medical or government): it does not mean they are ignored.

Important notes

• The data does not distinguish between JSON-LD, Microdata or RDFa: they are counted together.
• The statistics reflect the web as indexed by Google; no crawl covers the entire web.
• The format is open: other crawlers can contribute their own statistics in the same format.

Get the data

Download the raw data for the latest month:

This project

Sources

• Dataset announcement (blog.schema.org)
• About Usage Statistics (schema.org)
• Data files on GitHub