JSON-LD structured data adoption grew from 34% to 41% of pages

HTTP Archive’s Web Almanac, an annual analysis built on its own crawl of millions of live pages, tracked structured-data format usage across the web and found JSON-LD growing from 34% of pages in 2022 to 41% in 2024. That places JSON-LD, the format Google recommends, behind two older and broader formats in absolute share: RDFa at 66% of pages and Open Graph at 64%. Twitter meta tags sat at 45%, Microdata at 26%, and Facebook-specific meta tags at 7%. The chapter, written by Andrea Volpini, treats these as adoption rates for each format independently, not as a single pie that sums to 100%, since a page commonly ships more than one of them at once.

What the Web Almanac actually measured#

HTTP Archive runs its crawl through its own WebPageTest infrastructure rather than sampling a handful of sites by hand. Its 2025 methodology page states the July 2025 crawl alone covered 16,213,084 websites, split into 15,426,398 mobile pages and 12,155,374 desktop pages, each rendered in a real Chrome environment and logged into BigQuery for public querying. The 2024 structured-data chapter ran on the equivalent crawl a year earlier, under the same continuous methodology HTTP Archive has used since the Almanac began. This is what makes the 34%-to-41% JSON-LD figure a genuine year-over-year comparison rather than two studies with different sampling built from scratch.

What the numbers do and do not represent#

The crawl set is HTTP Archive’s own list of URLs, built from the Chrome UX Report and its own history, which skews toward pages popular and stable enough to be crawlable at scale - not a uniform random sample of every page that exists. The chapter also measures format presence, not correctness or completeness: a page counted as having JSON-LD may carry a single minimal Organization block rather than the rich markup a publisher might assume when they read “41%.” And because formats overlap, a rise in JSON-LD adoption does not mean RDFa or Open Graph declined by a matching amount; each format’s share moves on its own.

Why we care#

JSON-LD’s growth is real, but it is still the newer format catching up, not the leader. Two-thirds of crawled pages already carry RDFa or Open Graph, mostly from social-preview meta tags rather than deliberate schema.org work, so a 41% JSON-LD share undercounts how much of a publisher’s actual competitive set already ships some form of structured markup.

The comparison is trustworthy because it is one continuous methodology, not two. HTTP Archive’s crawl infrastructure and dataset shape are consistent enough between 2022 and 2024 that the before-and-after figure reflects real adoption change, not a change in how the data was gathered.

Presence is not quality. A rising JSON-LD percentage says more pages carry the format at all; it says nothing about whether that markup validates, matches on-page content, or covers the entity types that actually earn rich results.