What is microdata?
Microdata is one of the three formats for marking up a web page with structured data, alongside JSON-LD and RDFa. It works inline: you add attributes such as itemscope, itemtype and itemprop to the HTML elements that already hold the content, telling a machine that this heading is the name of a product, this span is its price and this block is a review. The vocabulary comes from schema.org, the shared dictionary that Google, Microsoft, Yahoo and Yandex agreed on in 2011 so that structured data would mean the same thing everywhere.
Because the markup wraps the visible content, microdata is by definition describing what is on the page. That was its appeal: no separate data block that could drift from the content or lie about it.
Why it matters
Why microdata matters
It matters because structured data is how a page qualifies for richer treatment in search: star ratings, prices, event dates, FAQ fold-outs, recipe cards, video thumbnails. Engines can usually work out what a page is about, but they will not guess that a specific number is a price or that a block is a review. Marking it up removes the guesswork and unlocks the result formats that take more of the page and earn more clicks.
It also matters as history and as inheritance. Many sites, themes and e-commerce platforms still carry microdata in their templates. Knowing how to read it is part of auditing a site, and knowing its limits is part of deciding whether to keep it.
In practice
How microdata works in practice
- 01
Pick the schema.org type that fits the content (Product, Event, Article, FAQPage, LocalBusiness, VideoObject), then wrap the relevant HTML. The container gets itemscope and itemtype pointing at the type; each child element gets an itemprop naming the property it holds. Nested items (a review inside a product, an address inside a business) nest the same way. Validate the result with Google's Rich Results Test and the schema.org validator, which show what was understood and what is missing.
- 02
In practice most teams now add structured data as JSON-LD instead: a single script block in the head that describes the page without touching the HTML. Google recommends it, it is easier to generate from a content management system and it does not break when a template changes. Microdata remains valid and fully supported; it is simply more work to maintain. If you inherit a site with microdata, keep it working or replace it, but do not run both in ways that disagree.
At RIOT
Microdata at RIOT
The FAQ at the bottom of this page, and of every entry in this glossary, is carried as structured data so the questions can appear in results as they are written here. The format matters less than the rule behind it: the markup describes exactly what is on the page, which is the promise microdata was built to keep.
Watch out
What people get wrong about microdata
Marking up what is not there. Structured data must match the visible content; a rating in the markup with no rating on the page is a policy violation and can earn a manual action. The other mistake is mixing formats carelessly, with microdata in the template and JSON-LD from a plugin saying different things. The third is expecting rich results on demand. Valid markup makes a page eligible; it does not guarantee the engine will show the richer format.
