Schema Markup for Beginners: Speaking Google’s Language
Everything else in this module helps Google infer what your page is. Schema markup lets you state it outright: “this is a recipe, it takes 40 minutes, it’s rated 4.8 by 200 people.” Explicit beats inferred — for rich results in the SERP, and increasingly, for the AI systems assembling answers from your content.
What Schema Markup Is
Schema markup (also called structured data) is a standardised vocabulary — maintained at schema.org by the major search engines jointly — for labelling your content in machine-readable form. It lives in your page’s code, invisible to visitors, and answers the questions Google’s indexing systems otherwise have to guess (Lesson 1.1): What type of thing is this page about? What are its properties?
Two payoffs:
- Rich results: the enhanced listings you learned about in Lesson 1.4 — star ratings, prices, event dates, recipe times shown right in the SERP. Eligibility for most of them requires the matching schema. Rich results occupy more space and earn more clicks at the same ranking position.
- Machine understanding: unambiguous data about your content and business feeds Google’s Knowledge Graph, helps entity recognition (the E-E-A-T groundwork from Lesson 1.2), and gives AI-driven features clean facts to retrieve. This payoff has no visual badge — and in 2026 it’s arguably the bigger one.
JSON-LD: The One Format You Need
Schema can be written in three formats, but Google recommends one and so does everyone else: JSON-LD — a small script block, usually in the page’s <head>, completely separate from your visible HTML. Here’s a real, complete example for a local business:
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "Bakery",
"name": "Sweet Layers Home Bakery",
"image": "https://example.in/photos/storefront.jpg",
"address": {
"@type": "PostalAddress",
"streetAddress": "12, Sunshine Society, Kothrud",
"addressLocality": "Pune",
"addressRegion": "MH",
"postalCode": "411038",
"addressCountry": "IN"
},
"telephone": "+91-98XXXXXXXX",
"priceRange": "₹₹",
"url": "https://example.in",
"openingHours": "Mo-Sa 10:00-20:00"
}
</script>
Read it top to bottom — it’s self-explanatory, which is exactly the point. @type declares what the thing is; the remaining properties describe it. Every schema type works the same way, just with different properties.
The Schema Types Worth Using in 2026
Schema.org defines hundreds of types; a handful earn their setup time for a typical site:
| Type | Use on | What you get |
|---|---|---|
| Organization / LocalBusiness | Homepage (site-wide) | Entity identity: feeds the Knowledge Graph and your knowledge panel; for local businesses, reinforces the Google Business Profile connection (Module 7). |
| Article | Blog posts, guides | Clear authorship, dates and publisher data — machine-readable E-E-A-T signals (Module 5), plus eligibility for article-style displays. |
| Product + Offer + Review | Product pages | Price, availability and star ratings in the SERP — among the highest-CTR rich results that exist. |
| Course | Course/lesson pages | Course info displays; directly relevant if you publish structured learning content (like the site you’re reading). |
| Recipe / Event / JobPosting | Niche-specific | Their own dedicated rich result formats — near-mandatory in those niches to compete visually. |
| BreadcrumbList | All pages | Clean breadcrumb trails in results instead of raw URLs. Usually automatic via your SEO plugin. |
| FAQPage | FAQ sections | Special case — read on. |
The FAQ schema nuance (updated 2026)
FAQ schema used to produce expandable Q&A dropdowns in the SERP. Google restricted that visual treatment to government and health sites back in 2023, and in mid-2026 deprecated the FAQ rich result entirely — the reports and testing support are being retired through 2026.
So should you delete FAQ schema? No — and this distinction makes you smarter than most of the industry. Google explicitly confirmed that only the visual SERP enhancement is gone; the FAQPage markup itself remains valid structured data. Google’s AI features work by retrieval — pulling specific content from indexed pages to assemble answers — and cleanly labelled question-answer pairs are ideal retrieval material. Combined with the question-heading pattern from Lesson 3.2, FAQ markup keeps feeding the machine layer even with no badge to show for it.
Plugins vs Manual: The Practical Setup
For WordPress users, the honest answer is: your SEO plugin already does most of this. Rank Math (and its peers) automatically outputs Organization/LocalBusiness, Article and BreadcrumbList schema from settings you fill in once, and offers per-post schema choices for Product, Course, Recipe, FAQ and more. The setup checklist:
- Complete the plugin’s site-identity settings — business name, logo, type (choose the most specific: “Bakery” beats “LocalBusiness”), address, social profiles. Ten minutes, done once, applied site-wide.
- Set default schema per post type — Article for posts, and the appropriate type for any custom content.
- Add specific schema per important page — Product on product pages, Course on course pages, FAQ where you have real FAQ sections.
- Go manual only for what plugins can’t express — unusual types or multi-entity pages. Paste a JSON-LD block (like the example above) via your header/footer code tool, and you’ve matched what agencies charge for.
Testing: Never Ship Unvalidated Schema
Broken schema silently does nothing (or worse, triggers warnings in Search Console). Two free validators:
- Google’s Rich Results Test — paste a URL or code; shows which rich results the page is eligible for and flags errors and missing fields. Test here first.
- Schema.org Validator — checks the markup’s structural validity regardless of Google’s rich-result support; useful for types (like FAQ now) that are valid data without a visual result.
Also monitor Search Console → Enhancements, which reports schema errors Google encounters across your whole site during normal crawling — your ongoing early-warning system.
Key Takeaways
- Schema markup states explicitly what Google otherwise infers — content type and properties, in machine-readable form.
- It’s not a ranking boost: it unlocks rich results and feeds machine understanding — including AI retrieval, which has no badge but growing value.
- JSON-LD is the only format you need — a self-contained script block, readable enough to write by hand.
- The golden rule: schema describes visible content truthfully — invented data is a manual-action magnet.
- Priority types: Organization/LocalBusiness, Article, Product, Course, BreadcrumbList — and keep FAQ markup despite the retired visual result, because retrieval-based AI features still read it.
- Plugin first, manual for the exceptions, validate always — Rich Results Test before shipping, Search Console Enhancements as the ongoing monitor.