Structured data is not a ranking factor in the way people hope. It is a clarity tool. Used well, it removes ambiguity about what a page is, who made it and what it costs. Used badly, it is a block of JSON nobody reads.
Here are the three types that consistently earn their place.
1. Organization (or LocalBusiness)
This is the entity definition. It tells a machine that your brand is a thing with a name, a URL, a logo and a set of official profiles.
{
"@context": "https://schema.org",
"@type": "Organization",
"name": "Example Ltd",
"url": "https://example.com",
"logo": "https://example.com/logo.png",
"sameAs": [
"https://www.linkedin.com/company/example",
"https://en.wikipedia.org/wiki/Example"
]
}
The sameAs array is the part that matters most for AI visibility. It is how you connect your site to the profiles that corroborate it. Get the list right, keep it short, and make sure every URL is live.
For a business with premises, use LocalBusiness (or a subtype) and include the address, telephone, opening hours and area served. Match the visible page exactly.
2. Article (or BlogPosting)
Every editorial page should declare itself as one, with a headline, a description, a publication date, a modified date and an author.
{
"@context": "https://schema.org",
"@type": "Article",
"headline": "The schema that actually earns citations",
"datePublished": "2026-11-16",
"dateModified": "2026-11-16",
"author": { "@type": "Person", "name": "Real Name" },
"publisher": { "@type": "Organization", "name": "Example Ltd" }
}
Two rules. The author must be a real, named person with a page, and the dates must be true. Fake freshness is easy to detect and expensive to recover from.
3. FAQPage
Only when the page genuinely contains the questions and answers, visible on the page. FAQ markup on a page with no FAQ is a guidelines violation, not a shortcut.
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [{
"@type": "Question",
"name": "How much does an SEO audit cost?",
"acceptedAnswer": { "@type": "Answer", "text": "Between £1,500 and £6,000..." }
}]
}
The benefit is extraction. A clean question-answer pair is easy for a model to lift and attribute.
What to skip
- BreadcrumbList on a site with no hierarchy. It adds nothing.
- Product markup without prices or availability.
- Review markup for reviews you wrote about yourself.
- Multiple competing Organization blocks on the same page. One, consistently.
The test
Paste your page into a schema validator, then read the output as though you were a machine with no other context. Could you tell what the page is, who made it, and what it is about?
If yes, the markup is doing its job. If not, more markup will not help.