Structured Data

Concept

Markup added to a page that labels what each value actually is, kept separate from the text a human reads.

Why it exists

What a search engine takes away from a page is text. A person can tell at a glance which number is the price and which line is the author's name, from position and size alone. A machine has none of those cues.

So the same characters can be read several ways. A 4.5 on the page could be a rating or a version number; a date could be when an event happens or when the article was published. The search engine guesses from surrounding sentences, and when the guess is wrong the value simply goes unused.

Unused shows up in the shape of the search result. Your pricing page appears as a title and one line of text while a competitor appears with a star rating and a price attached.

Structured data exists to remove that guess. The text humans read stays as it is, and a second copy is written for machines to read.

How it works

Markup invisible on screen is included in the page, pairing each value with a label a machine can read.

It usually goes into a separate block in the page source. The rendered page does not change; the crawler reads that block and takes away the values already paired with their labels. Which labels to use follows a shared vocabulary the search engines agreed on — that vocabulary is schema.org markup.

What gets labeled depends on the kind of page.

  1. An article labels the title, the author and the publication date.
  2. A product labels the price, the currency and availability.
  3. A review labels the rating and the number of raters; a frequently asked questions page labels each question paired with its answer.

There is one rule. Never label a value that does not appear on the page. Writing a 4.8 rating into the markup when the page shows no reviews violates the guidelines, and the markup gets ignored.

After writing it, check for errors with the testing tool the search engine provides. One missing value causes the whole block to go unused, so hunting for what is missing is most of the actual work.

None of this raises rankings directly. What changes is how the result looks, and a result carrying a rating or a price is called a rich snippet. The position stays the same while the click rate goes up.

How it sounds at work

"Let's add product markup to the pricing page. Click rate changes when the price shows in the result."

Change the shape of the result to win more clicks.

"Drop the rating markup. There are no reviews on the page, so that's a guideline violation."

You cannot label a value that is not there.

"The testing tool says currency is missing. That's why the price never showed."

One absent value made the whole block unusable.

At Pagedeck, Theo added title and publication date markup across the help articles, and question-and-answer pairs on the FAQ page. Two weeks later Ravi confirmed the questions expanding inside the search result, with the ranking unchanged and the click rate up.

Easily confused with

TermWhat it refers toHow it differs
Schema.org markupThe shared vocabulary the labels come fromAn agreement about which names to use. Structured data is what you actually wrote into your page following that agreement

Where to go next

  • Technical SEO

    see where comprehension aids sit among the six groupings and what comes first

  • Schema.org markup

    the vocabulary used to write it, so reading both sets the order of work

Related terms

Last updated 2026-09-16