Structured Data and Rich Snippets Guide to Boost SEO

Structured data and rich snippets: what actually shows up in search

You've probably had this conversation with a client or a boss: "Can we get the stars to show up in Google?" And you say yes, then you open the page source, stare at a wall of JSON-LD that someone pasted from a tutorial in 2019, and realize half of it is invalid. I've been there. Twice this year alone.

Structured data is the code you write. Rich snippets are what Google sometimes decides to display. Those two things are not the same, and treating them as the same is the single most common mistake I see on sites that should be doing much better in search results.

What is the difference between rich results and structured data?

Structured data is the input. Rich results are the output — and the output is optional.

Concretely: structured data is a snippet of code, typically JSON-LD dropped in the <head> or <body> of a page, that tells a crawler "this is a recipe," "this is a product priced at €49," "this article was written by X and published on Y." Schema.org defines the vocabulary. Google's documentation defines which of those vocabulary items it actually reads.

A rich result — often called a rich snippet, though the terms aren't perfectly interchangeable — is the visual upgrade you see on the results page: review stars under a product, a cooking time next to a recipe link, a price range, a sitelinks block. Google generates these when three things line up:

1. The markup is valid and complete for that content type
2. The page's visible content matches the markup (no hidden prices, no invented ratings)
3. Google's systems decide that showing the enhancement improves the result

That third condition is where most people lose. I've marked up a perfectly valid page and watched the stars never appear. I've also seen stars show up on pages where the markup was technically incomplete. Google runs its own quality and trust signals on top of the technical layer, and those signals are not published.

Why the distinction matters in practice

If you conflate the two, you end up promising stakeholders something you can't deliver. "We'll add the schema and get stars" is a promise you don't control. "We'll add valid Product schema with aggregateRating, and monitor whether Google chooses to display it" is honest.

That framing change alone saved me one very uncomfortable quarterly review. The client understood the boundary. The stars showed up six weeks later. Nobody had to pretend I'd willed them into existence.

What is a rich snippet?

A rich snippet is any search result that shows more than the classic blue-link block of title, URL, and description. The formula is simple: standard result + extra visual or textual information = rich snippet.

The most recognizable forms:

• Review stars and rating counts, from Product, Recipe, or LocalBusiness markup
• Price and availability on e-commerce results
• Recipe cards with cook time, calorie count, and a thumbnail
• Event listings showing date and location
• Breadcrumb trails replacing the raw URL with a readable path
• Sitelinks search boxes, which let users query your site directly from the results page
• Video thumbnails and duration for pages with VideoObject markup

Notice what's missing from that list: FAQ accordions. That's not an oversight. More on that below.

One thing worth internalizing early: Google's documentation uses rich results as the umbrella term, and rich snippets is the older, looser label that stuck in common usage. If you're reading a 2020 blog post that insists on a hard distinction between the two, it's probably chasing a distinction Google itself stopped caring about.

Does FAQ schema still work?

No — not in the way you remember it. Google deprecated FAQ rich results for the vast majority of sites in August 2023.

Before that date, adding FAQPage schema to a page with a visible Q&A block could earn you an expandable accordion in the SERP. Sites with a moderate amount of content could occupy a shocking amount of vertical space. Then Google announced that FAQ rich results would only appear for "well-known, authoritative government and health websites," and that was that for everyone else.

What this means for you in 2026:

• Adding FAQPage markup won't get you accordions unless you're a government or health authority site
• It's not harmful to keep valid FAQ markup on the page — it can still help machines parse your content
• Building a content strategy around FAQ rich results is a strategy built on a feature that no longer exists for you
• The same fate hit HowTo rich results, which Google also wound down

I spent a full weekend in early 2023 migrating roughly 40 pages to add FAQ schema across a client's blog. By that autumn, the accordions were gone. The pages still rank. The markup did nothing measurable for the SERP appearance. That's a lesson in why you should never build a strategy around a single SERP feature you don't control.

So what do you do instead?

Keep FAQ schema if it's already there — it costs nothing and can help with entity clarity. But shift your energy toward Article, Product, Organization, Person, and BreadcrumbList markup. Those still influence how search engines and AI systems understand your content, even when they don't produce a visible enhancement.

What are three types of structured data?

There are dozens of schema types, but if you're getting started, these three cover the majority of commercial use cases.

Type	Best for	Typical rich result	Effort to implement
Product	E-commerce product pages	Price, availability, review stars, shipping info	Medium — needs price, currency, availability, and rating data that stays in sync
Article / NewsArticle / BlogPosting	Editorial content, blogs, news	Top stories carousel, author info, publication date	Low — most CMS platforms generate this automatically once configured
BreadcrumbList	Any site with hierarchical navigation	Readable breadcrumb path replacing the URL	Low — one snippet per page, rarely changes

Beyond those three, the ones I reach for most often are LocalBusiness (for any business with a physical location), Event, Recipe, and Organization (usually on the homepage only, to anchor the brand entity).

Don't try to mark up everything. I've seen sites with fifteen schema types crammed onto a single product page, half of them conflicting. Pick the types that match what the page actually is, implement them cleanly, and stop.

How to test and validate your markup

Two tools matter, and people mix them up constantly.

The Rich Results Test checks whether a URL or code snippet is eligible for Google's supported rich result types. It tells you which enhancements it detects and flags errors that would block them.

The Schema Markup Validator checks your markup against the schema.org vocabulary itself, regardless of whether Google cares about that type. It's useful for validating types that Google doesn't support for rich results — you'll get a clean bill of health without a promise of any SERP feature.

Use both. I run every new template through the Rich Results Test first, because that's where the commercial value lives. Then I run the Schema Markup Validator to catch structural problems the first tool ignores.

Common errors I keep finding

• Mismatched prices. The markup says €49, the page shows €55. This is the fastest way to trigger a manual action.
• Missing required properties. Product schema without offers or priceCurrency is dead weight.
• Invented review counts. Never fabricate an aggregate rating. Google catches this and it's a policy violation.
• Stale data. Markup gets updated manually once, then the CMS changes the page and nobody touches the schema again.
• Duplicate conflicting blocks. A page with two Product JSON-LD blocks specifying different prices.

That last one cost me a full afternoon of debugging once. The developer had added one block via a plugin and another directly in the template. Google read both. The Rich Results Test flagged the conflict, but only after I inspected the raw HTML — the visual test showed green.

Why structured data matters more in AI-driven search

The deprecation of FAQ and HowTo rich results might make you think structured data is a fading investment. The opposite is closer to true.

AI-generated search summaries and chat-based search tools lean heavily on structured information to identify entities, attributes, and relationships. When a machine is trying to decide whether your page is about a specific product, service, or person, clean markup gives it a shortcut. Pages with consistent Product, Organization, and Person markup tend to get parsed more confidently than pages where the same information only exists as prose.

This isn't a visible win you can screenshot for a client deck. It's the kind of thing that shows up as sustained visibility across multiple surfaces, and it compounds slowly. Which is annoying, because slow compounding is exactly what nobody wants to fund.

Still true though. The sites I've worked on that kept their markup maintained over years consistently outperform the ones that bolted on schema, chased a SERP feature for a quarter, and abandoned the whole thing.

Getting started without overcomplicating it

Here's the order I'd follow if I were starting fresh on a site today.

1. Add Organization markup to the homepage. One block. This anchors your brand entity for everything else.
2. Add BreadcrumbList sitewide. Most CMS platforms handle this with a config toggle. If yours doesn't, it's still a fifteen-minute snippet.
3. Pick your money pages and mark them up with the appropriate type — Product, Article, LocalBusiness, whatever fits.
4. Validate everything with both the Rich Results Test and the Schema Markup Validator.
5. Monitor in Search Console. The Enhancements reports show you which valid markup is actually producing rich results and which is being ignored.
6. Re-check quarterly. Markup drifts as pages change. Set a reminder.

Skip the temptation to build a fully automated schema generation pipeline in your first month. I tried that on a mid-sized e-commerce site and ended up with 300 pages of technically valid but factually wrong markup, because the generator pulled from a product feed that was itself out of date. Manual, targeted markup on the pages that matter beats automated markup everywhere, at least until you've proven the manual version works.

A final thought on expectations

You will not get stars on every product. You will not get an FAQ accordion. You will not be able to tell a stakeholder exactly when a rich result will appear, because Google doesn't publish that logic and changes it without notice.

What you can do is make your content unambiguously legible to machines that increasingly decide what humans see. That's the durable part of the work. The stars are a nice bonus when they show up — and a reminder that you never really owned them in the first place.