Structured Data, Explained: What JSON-LD Actually Does (and Why Crawlers and AI Need It)

Structured data is code that tells search engines and AI tools exactly what your content means, instead of leaving them to guess. It uses a format called JSON-LD to label your pages as an organization, an article, a service, or a set of questions and answers. In 2026, that clarity is what gets you understood, and increasingly, cited.

What structured data actually does

Every page you publish gets read on two levels. There is the text layer, which is what you and I see. And there is the semantic layer, a set of plain, explicit claims about what that text represents. Without structured data, crawlers and AI have to infer the semantic layer from context. They are getting better at guessing, but a guess is still a guess. Structured data lets you skip the guessing and state the facts directly.

Without structured data:
  "Open Mon to Fri, 9 to 5"  ->  [just words]  ->  the crawler guesses

With structured data (JSON-LD):
  "Open Mon to Fri, 9 to 5"
     + labels: openingHours, dayOfWeek, opens, closes
     ->  [explicit meaning]

Without structured data:
  "Open Mon to Fri, 9 to 5"  ->  [just words]  ->  the crawler guesses

With structured data (JSON-LD):
  "Open Mon to Fri, 9 to 5"
     + labels: openingHours, dayOfWeek, opens, closes
     ->  [explicit meaning]

Without structured data:
  "Open Mon to Fri, 9 to 5"  ->  [just words]  ->  the crawler guesses

With structured data (JSON-LD):
  "Open Mon to Fri, 9 to 5"
     + labels: openingHours, dayOfWeek, opens, closes
     ->  [explicit meaning]

This matters for two audiences at once. For traditional search, clear labels make you eligible for richer search features and help engines understand your site. For AI tools like ChatGPT, Perplexity, and Google's AI answers, structured data acts as a bridge between your content and the model's understanding, feeding the entity models they rely on. Google itself says it uses structured data to understand a page and to learn about the people, companies, and things described on it. When you consistently label who you are and what you do, you are quietly teaching the whole web's knowledge layer about your business.

JSON-LD vs microdata vs RDFa

There are three ways to add structured data to a page. You really only need to care about one of them.

Google has recommended JSON-LD for years, and the reason is practical. Because JSON-LD sits in its own block, separate from your visible content, you can add, change, or remove it without touching your page layout. It is easier to maintain, easier to validate, and far less likely to break something. Unless you have a specific reason to do otherwise, use JSON-LD.

The vocabulary itself, meaning the list of available types and properties, comes from Schema.org, a shared project backed by Google, Microsoft, Yandex, and others. It is huge: as of 2024, over 45 million domains use it across hundreds of billions of tagged items. You will never need most of it. Most service businesses do everything they need with seven to ten types.

The schema types that matter for a founder-led-business

You do not need to learn the whole dictionary. You need the handful of types that describe who you are, what you publish, and what you sell. Think of them as a pyramid, with your identity at the base.

• Organization and ProfessionalService are your foundation. They establish your entity: name, URL, logo, description, contact details, and social profiles. Put them on your homepage and keep them consistent everywhere. This is what feeds your profile in Google's knowledge layer.

• Article marks up your blog posts and guides. It tells crawlers who wrote a piece, when it went live, and when it was last updated. Those last two, datePublished and dateModified, matter because freshness influences how informational content surfaces.

• Service describes each thing you offer, and points back to your Organization as the provider. On a service page, this makes a clean, machine-readable claim: this business provides this service.

• FAQPage maps your questions to your answers in a format machines can lift directly. More on its 2026 plot twist below, because it is not the slam dunk it used to be.

• BreadcrumbList tells crawlers where a page sits in your site's hierarchy. It only works if your site actually has a clean hierarchy, so treat your information architecture as the real foundation. Schema can describe a logical structure, it cannot invent one.

What quietly changed in 2026 (read this before you over-invest)

Here is where a lot of older advice, including the earlier version of this very article, needs an honest update. Three myths are worth retiring.

That first point is the big one. Google's John Mueller has said it plainly and repeatedly: structured data will not make your site rank better. It makes you eligible for certain search features and helps engines understand you, which can lift engagement over time, but it is not a lever you pull for rankings.

On AI citations, it is tempting to repeat the confident numbers floating around. Some AEO agencies report content with clean structured formatting is 28 to 40 percent more likely to be cited, and the directional logic is sound: machines cite what they can clearly parse. But the most careful study we found complicates the hype. A cross-platform analysis of 730 AI citations from ChatGPT and Gemini found that, once you account for domain authority, schema on its own did not reliably predict more citations. In other words, structured data helps machines understand you, but it does not substitute for being a credible, well-linked source. We are flagging the rosier stats as directional, not gospel.

So should you still bother with FAQ markup now that the rich results are gone? Most practitioners say yes, just lower your expectations. As one SEO put it in a widely shared LinkedIn thread, most sites simply kept their FAQ code after the change and saw no harm, because accurate markup still adds context that machines can read, even when it no longer paints a fancy result in the SERP. Keep it for clarity, not for decoration.

How to add JSON-LD correctly (no developer required for the basics)

The reassuring truth: you do not need to hand-write code. Free generators will produce the JSON-LD for you, and on most platforms adding it is closer to a five minute job than a engineering project. Where people go wrong is not the syntax, it is the logic. Watch for these:

• The copy-paste trap. Do not paste the same static schema onto every page. If your Organization block hardcodes your homepage URL but lives on a service page, you have just sent a contradictory signal. Each page's schema should describe that page.

• Schema that disagrees with the visible content. Your markup has to match what a human actually sees on the page. If they disagree, you can trigger problems rather than benefits.

• Skipping validation. Run your key pages through Google's Rich Results Test and the Schema.org validator before you ship. They catch the small errors that quietly break everything.

Your quick-start checklist

• Add Organization (and ProfessionalService) schema to your homepage

• Add Article schema to every blog post, with datePublished and dateModified

• Add Service schema to each service page, with the provider pointing to your Organization

• Keep FAQPage markup on real Q&A content for machine readability, even though the rich result is gone

• Add a BreadcrumbList that reflects your real site hierarchy, not a wished-for one

• Make sure each page's schema URL matches that page, not a hardcoded homepage

• Validate every key page in the Rich Results Test and the Schema.org validator

• Re-check the schema whenever you change the content

Keeping it healthy over time

Structured data is not a one-and-done task, and the simplest way to stay sane is to tie schema to content. When a page changes, reviewing its schema is part of the same job. When you launch a new content type, designing its schema is part of the launch. That one habit prevents the slow drift where your markup describes a version of your business that no longer exists. For ongoing peace of mind, Google Search Console's reports will flag pages with errors so you can fix them before they matter.

That is really the whole philosophy. Structured data is the vocabulary that lets you speak to crawlers and AI directly, in plain, unambiguous terms. It says who you are, what you published, and what you offer, so your work is not overlooked in favor of a competitor whose identical content simply happens to be labeled more clearly.

FAQ

Do I need to know how to code to add structured data?

No. Free tools generate the JSON-LD for you, and most website platforms let you paste it in without touching anything else. The skill that matters is logical, not technical: making sure each page's labels honestly match what is on that page.

Will structured data make my site rank higher?

Not directly. Google has said clearly that it is not a ranking factor. What it does is make you eligible for certain search features and help engines and AI understand your content, which can improve engagement over time. Treat it as clarity, not as a ranking shortcut.

Is FAQ schema still worth adding now that the rich results are gone?

For most businesses, yes, with realistic expectations. Google retired FAQ rich results for everyone except government and health sites, so you will not see the expanded result anymore. But accurate FAQ markup still gives machines clean question-and-answer pairs to read, and most sites kept theirs in place without any downside.

Does structured data actually get me cited by ChatGPT and other AI tools?

It helps, but it is not a guarantee. Clear markup makes your content easier for AI to parse, which is a good thing. The strongest available study found that authority and credibility mattered more than schema alone. So add structured data, and keep earning the trust and links that make you a source worth citing.

JSON-LD, microdata, or RDFa, which should I use?

JSON-LD, in almost every case. Google recommends it, it lives in its own block separate from your visible content, and it is the easiest to maintain and validate. Microdata and RDFa still work, but there is rarely a reason to choose them today.

How often should I check my structured data?

Whenever your content changes, and as a light monthly habit through Search Console. Tying schema review to content updates is the easiest way to stop your markup from quietly going stale.

Want this done right, not just done?

Structured data is the difference between hoping the machines understand you and making sure they do. If you would rather have your entity, services, and content labeled cleanly and validated properly, that is exactly the kind of foundation our Search Presence (SEO, AEO & GEO) work is built on. Let's make your site easy to understand, and easy to cite.