Schema markup: a practical guide for business websites

Schema markup is code you add to your website that helps search engines understand what your content means, not just what it says. It is written in a format called JSON-LD and placed in the head section of your page. It does not change how your page looks to visitors, but it can change how it appears in search results.

What schema is

Schema.org is a collaborative vocabulary maintained by Google, Microsoft, Yahoo and Yandex. It defines a standard set of types and properties that webmasters can use to mark up their content. When a search engine reads schema markup, it understands that this text is a phone number, this is an address, this is a question and this is its answer — rather than having to infer those relationships from context.

The preferred format is JSON-LD, a JavaScript-based syntax that lives in a <script> tag in the page head or body. JSON-LD is easier to implement, maintain and debug than the older Microdata or RDFa formats, and Google explicitly recommends it.

Schema is not a direct ranking factor. Google has stated this repeatedly. What it does is enable rich results — enhanced search listings with star ratings, FAQ accordions, breadcrumb trails, product prices, event dates, and more. Rich results improve click-through rates significantly, and a higher click-through rate can lead to better rankings indirectly through user engagement signals.

Organisation schema

Organisation schema is the foundation block. It should appear on every page of your site, typically via a single block referenced by @id to avoid duplication. It tells search engines your company name, logo, contact details, social profiles and founding date. AI answer engines use Organisation schema as a primary source when deciding whether to cite a business as the answer to a query.

Include all required fields: name, URL, logo. Strongly recommended fields: address, contactPoint with telephone and email, sameAs for social profiles, foundingDate, and description. If your business has certifications such as ISO 9001 or PCI DSS, add them using hasCredential with EducationalOccupationalCredential.

Implement Organisation schema once and reference it from other schema types using "publisher": {"@id": "https://yourdomain.com/#org"}. This creates a connected entity graph that search engines use to associate all your content with your business.

LocalBusiness schema

If your business has a physical location where customers can visit, add LocalBusiness schema. It extends Organisation with location-specific properties: opening hours, payment methods, price range, and geo-coordinates.

Use the most specific subtype that fits your business — Dentist, Restaurant, Store, ProfessionalService — rather than the generic LocalBusiness. The extra specificity helps search engines match your listing to the right queries. Each physical location needs its own LocalBusiness block with unique address and geo coordinates.

Keep the opening hours current. An incorrect opening time displayed in search results creates a poor user experience and reduces the chance of the rich result being shown. Update LocalBusiness schema whenever your hours change.

Article schema

Article schema should be on every blog post, news article, and guide. The most commonly used subtype is Article or BlogPosting. Required fields: headline, image, datePublished, author. Strongly recommended: dateModified, description, publisher reference, mainEntityOfPage, wordCount, inLanguage.

Article schema is the primary mechanism by which search engines identify a page as a piece of content worth surfacing for informational queries. It also feeds AI answer engines, which use the headline, description and datePublished fields to decide freshness and relevance. An article without Article schema is less likely to be cited in AI answers than an otherwise identical article with it.

Use a real author reference with a name and URL to the author page. AI answer engines and search quality systems weigh author identity, and a fabricated or missing author is a credibility signal in the wrong direction.

FAQPage schema

FAQPage schema enables an accordion-style rich result in search where users can expand questions and see answers directly on the results page. It is one of the highest-impact schema types because it occupies more visual space in search results and directly answers the user's query before they click.

Each FAQ item needs a Question with a name (the question text) and an AcceptedAnswer with a text property. The entire FAQ block sits inside a mainEntity array. The answer text in the schema must match the visible answer on the page — Google's guidelines explicitly prohibit marking up content that is not visible to users.

One FAQPage block per page. Do not combine FAQPage with QAPage on the same page, as the two types serve different purposes. FAQPage is for your own FAQ content; QAPage is for user-generated questions and answers.

Product schema

Product schema is essential for e-commerce pages. It enables price, availability and review rich results in search. Required fields: name, description, offers with price and priceCurrency. Strongly recommended: image, brand, aggregateRating only if you have genuine reviews, review for individual reviews.

Use Offers.Availability correctly — indicate InStock, OutOfStock, PreOrder or Discontinued based on actual inventory. Incorrect availability markup can trigger a manual action if Google detects a pattern of promising items that are not available. For services rather than physical products, use Service schema instead of Product.

Testing tools

Two tools cover almost all schema testing needs. Google's Rich Results Test takes a URL or code snippet and shows whether Google can parse the markup, which rich results it enables, and any errors or warnings. The Schema.org validator provides full technical validation against the schema vocabulary and catches issues the Google tool may not flag.

Validate every page that uses schema whenever you update the markup. Schema errors are common — missing required fields, incorrectly formatted dates, non-matching visible and hidden content, improperly nested types. A validation pass before publishing catches these and prevents the markup from being silently ignored.

If you would rather not manage schema yourself, we include structured data setup with every website we build and every SEO campaign we run.