Structured data is code added to a page that labels its content in a standard vocabulary, usually schema.org, so machines can read facts without guessing from layout. It can make pages eligible for rich results in Google, such as review stars, breadcrumbs or product details, and helps search engines connect a page to entities like an organization or software product. Google recommends the JSON-LD format.
Formats
There are three formats: JSON-LD (a script block), Microdata and RDFa (attributes inside the HTML). Google supports all three but recommends JSON-LD because it sits apart from the visible markup and is easier to maintain. The vocabulary itself comes from schema.org, which defines types such as Organization, SoftwareApplication, Article, BreadcrumbList and FAQPage.
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "SoftwareApplication",
"name": "Example CRM",
"applicationCategory": "BusinessApplication",
"operatingSystem": "Web",
"offers": {
"@type": "Offer",
"price": "29.00",
"priceCurrency": "USD"
}
}
</script>Types that suit SaaS sites
Mark up what the page actually is. Most SaaS sites only need a handful of types.
- Organization on the homepage: name, logo, URL, sameAs links to official profiles
- SoftwareApplication on product or pricing pages
- Article or BlogPosting on blog posts, with author and dates
- BreadcrumbList on docs and deep pages
- VideoObject for product demos
What it does and does not do
Valid structured data makes a page eligible for rich results; it does not guarantee them. Google decides whether to show them, and it changes which types it supports. It restricted FAQ rich results to well-known government and health sites in 2023, removed HowTo rich results, and has retired other types since, so check the current Search Gallery before building around a feature.
Google has said structured data is not a general ranking factor. Its value is in eligibility for enhanced results and in clearer understanding. How much AI search systems use structured data is not well documented; it is reasonable to keep it accurate, but do not expect it to drive AI citations by itself.
Rules to follow
Markup must describe content visible on the page and must be truthful. Fake reviews, self-serving ratings or prices that do not match the page can lead to a manual action that removes rich result eligibility. Test with Google's Rich Results Test and the Schema Markup Validator, and monitor the enhancement reports in Search Console.
Example
Breadcrumbs for a deep docs tree
A fictional data pipeline SaaS, Pipestream, has docs four levels deep. Search results show long, unreadable URLs for docs pages. Adding BreadcrumbList markup generated from the docs navigation, matching the visible breadcrumb trail, makes the pages eligible for a breadcrumb path in results.
Common mistakes
Marking up content that is not on the page
Reviews, ratings or FAQs in JSON-LD must also appear to visitors. Hidden markup breaks Google's guidelines.
Copying the same markup to every page
Sitewide Product or SoftwareApplication markup with identical data on blog posts adds noise. Match the type to the page.
Never re-testing
Template changes and CMS updates break markup silently. Watch Search Console's enhancement reports.
Questions
Google has said it is not a general ranking signal. It can earn rich results, which may change how a listing looks and how often it is clicked.
Google recommends JSON-LD. It is easier to generate and maintain, especially in JavaScript frameworks.
Most sites no longer get FAQ rich results in Google. The markup is harmless if accurate, but do not build pages around the feature.
