Search engines can read your words, but they still need help distinguishing a product from a review, a business from an author, or an article from ordinary page text. Schema markup is a form of structured data that adds context in a machine-readable format. It may help systems interpret a page for voice search and make it eligible for rich results in search results.
Used properly, this context helps Google understand what your page describes and how its entities relate to one another. It can support better search presentation, but it doesn’t guarantee a higher ranking, a rich result, or an AI citation. Accurate content, not markup alone, still determines visibility in voice search.
Key Takeaways
- Schema markup is structured data that helps search engines interpret a page’s content, entities and relationships.
- Use the most specific accurate Schema.org type for the page, such as
Article,Product,Organization,LocalBusinessorBreadcrumbList. - JSON-LD is the preferred implementation format, but every value must match information visible on the live page.
- Valid markup can make a page eligible for rich results, but it does not guarantee enhanced listings, higher rankings, voice search visibility or AI citations.
- Test structured data before publishing and monitor it in Google Search Console after deployments, template changes and content updates.
What schema markup means for your website
Schema markup is structured data added to a webpage using Schema.org, a shared semantic vocabulary. Schema.org (also searched for as ‘schema org’) labels information such as a product name, price, author, business address, event date, review score, or breadcrumb trail.
Most websites use JSON-LD (also written as “json ld”), a script format separate from visible page content. Google recommends JSON-LD because it keeps data separate from HTML elements and is easier to maintain than embedded Microdata. RDFa and Microdata are also supported formats.
Google’s introduction to structured data explains how these formats help search engines interpret page content.
A page about a Shopify service, for example, might contain several related entities:
- An Organisation identifies the company publishing the page.
- A WebPage identifies the URL and page type.
- A Service describes what the company offers.
- A BreadcrumbList shows where the page sits within the website.
These connected entities make the structured data easier to understand. They may also influence how information appears in search results and help voice search systems identify a business or service.
However, structured data markup must describe content users can see on the page. Hiding a price when none is shown, or adding reviews that don’t exist, can mislead users and voice search systems.
Rich results are not the same as featured snippets
Schema markup provides structured data that can make a page eligible for rich results, enhanced listings in search results with extra information. A product result might show price, availability and review details. An event result might show a date and venue. Breadcrumbs can also appear beneath a result’s title.
People often use “rich snippets” as a general term for these enhanced listings. Google tends to use “rich results” for the wider group of search features it supports.
A featured snippet is different. Search engines select a passage from a page and display it above or alongside normal search results, often in response to a question. A spoken answer from voice search is another feature, not the same as a featured snippet. A page doesn’t need this markup to appear in a featured snippet.
The distinction matters because adding structured data won’t force Google to display a particular format or guarantee a voice search response. Eligibility for rich results depends on the schema type, the page content, Google’s guidelines and the search query.
A valid schema item can produce no visible enhancement. Validation confirms that Google can read the markup, not that Google will display it.
It also isn’t one of Google’s direct organic ranking factors. It can improve the way a result communicates its value, which may support click through rates when the feature appears. The page still needs useful content, sound technical seo, good links and a strong match for the searcher’s intent.
Schema markup types worth considering
Choose the most specific of the schema types that describes the page’s main subject. Structured data markup should represent the real page and content type. It can help systems interpret content for search results and, where relevant, voice search, but more markup isn’t automatically better.
Editorial content, blog posts and news stories suit article schema. Use Article, BlogPosting or NewsArticle where appropriate. Include a clear headline, author, image and publication or modification date. Google recommends an identifiable author and an ISO 8601 dateModified value, preferably with a timezone.
An individual product page can use product schema. It can describe the product name, image, brand, SKU, offers, currency, availability and genuine reviews. The price and stock status in the markup must match the page and remain current.
The company behind a website can be identified with organization schema, the official Schema.org Organization type. Add the official name, logo, URL and relevant sameAs profiles. Consistent business details across your website and reputable third-party profiles help search engines connect the right entity to the right brand in a knowledge graph.
A real local business can use LocalBusiness schema for a physical location. Use accurate address details, telephone number, opening hours and service information. A national business without a public premises shouldn’t add a false local address simply to gain local visibility. That won’t guarantee a local voice search result.
BreadcrumbList, also known as breadcrumb markup, describes the site’s hierarchy. It helps systems understand how a page relates to its parent categories.
Use review schema with particular care. Mark up reviews that are visible on the page and relate to the item being reviewed. Never create review scores or copy third-party ratings without following the relevant rules.
FAQPage and HowTo markup should describe genuine FAQ or step-by-step content. Google has limited FAQ rich-result eligibility for many commercial websites, so the markup doesn’t guarantee an FAQ feature.
Google’s structured data search gallery lists the search features and rich results it currently supports. Check it before choosing a type, as support and eligibility can change.
How to implement schema markup on WordPress or by hand
Start with the page, not the code. Identify its main purpose and select the most specific applicable type before adding schema markup. A product page shouldn’t receive Article markup, and a category page shouldn’t pretend to describe one individual product.
On WordPress, a WordPress plugin such as Yoast SEO can create foundational structured data for the organisation, webpage and breadcrumbs. WooCommerce can add product information to product pages. These tools save time, but inspect the output after installation. A second WordPress plugin generating different Product or Organisation entities can create duplication and conflicting values. Accurate entity data may also help systems interpret content for voice search, but it doesn’t guarantee visibility.
For manual implementation, you may see the format described as json ld. JSON-LD is the technically correct format name. It usually goes in the page head or body as structured data markup. You can write it yourself or use a schema generator as a starting point. Review every field before publishing, particularly prices, dates, availability, addresses and review scores.
This compact article schema example shows the basic structure:
{
"@context": "https://schema.org",
"@type": "Article",
"headline": "How to Improve an Online Shop",
"author": {
"@type": "Organization",
"name": "BBMM"
},
"dateModified": "2026-08-02T09:00:00+01:00"
}
For a product page, this product schema example keeps the offer aligned with what shoppers can see:
{
"@context": "https://schema.org",
"@type": "Product",
"name": "Example Product",
"offers": {
"@type": "Offer",
"price": "49.00",
"priceCurrency": "GBP",
"availability": "https://schema.org/InStock"
}
}
The values in these examples are illustrative. Replace them with the live page’s accurate details, and add an image, brand, SKU or review only when the page supports those properties. Accurate implementation can make a page eligible for rich results in search results, but it cannot guarantee them.
A page can contain several entities. Use @id values to connect them rather than creating unrelated copies of the same organisation. For instance, a WebPage can point to its main Article, whilst the Article can identify its author. This approach creates clearer relationships within your site’s knowledge graph. It may help voice search systems interpret those relationships, but it doesn’t control Google’s graph or guarantee visibility.
Test, publish and monitor your structured data
Before publishing, run the URL through Google’s Rich Results Test to complete initial schema validation for the page’s schema markup. It can show whether the page contains recognised structured data and whether it qualifies for rich results in search results.
Then check the live URL after deployment. A staging site may behave differently if it blocks crawlers, requires a login or uses temporary URLs.
Google’s general guidelines require structured data to represent the visible page accurately. Pages must remain accessible to search engines, and this markup shouldn’t sit behind a noindex directive or a blocked resource.
Use Google Search Console to monitor enhancement reports and inspect individual URLs. After deployment, return to Google Search Console to check important pages.
Treat errors and warnings differently:
- Fix errors involving missing required properties or invalid values.
- Review warnings and add recommended properties where they improve clarity. Output from a schema generator still needs review against the visible page.
- Compare the structured data markup with the visible page after template, plugin or catalogue changes.
- Repeat schema validation after deployment, template changes, migrations, redesigns and product feed updates. Include voice search queries in a small test set, then check for unintended effects on voice search visibility.
Warnings don’t always prevent eligibility, but they can point to incomplete data. An Article page, for example, becomes more useful when its author and modification date are clear. A product page becomes less trustworthy when its schema says “InStock” while the page says the item is unavailable.
Does AI search affect your visibility?
Schema markup can help systems identify a page’s subject, relationships and business identity. That may support retrieval and entity recognition across search engines, including Google AI Overviews, ChatGPT, Gemini and other answer engines. It doesn’t guarantee that any platform will mention your brand or cite your page.
AI systems also assess a page’s wording, authorship, sources, freshness and consistency. An accurate LocalBusiness entity can help, but it won’t correct contradictory opening hours across your website and business profiles. organization schema can identify a company through the official Organization type. Clear contact details, credible information and consistent organisation, local-business and author signals may help systems connect entities in a knowledge graph.
For answer engine optimisation, make important pages easy to quote. Put a direct answer near the start, use descriptive headings, keep paragraphs self-contained and support claims with named sources. Structured data should reinforce those signals, not replace them.
Measure AI visibility with a small set of real customer prompts, including qualified voice search questions. Record whether your brand appears, which pages receive citations, whether competitors appear and whether a voice search answer represents your business accurately. Combine those observations with google search console, GA4 referrals, branded searches, qualified leads and assisted conversions. A ranking report alone won’t show every way a customer encounters your business.
Frequently Asked Questions
What is schema markup?
Schema markup is structured data added to a webpage using the Schema.org vocabulary. It gives search engines machine-readable context about entities such as products, articles, businesses, authors and events.
Does schema markup improve SEO rankings?
Schema markup is not a direct organic ranking factor and cannot guarantee higher rankings. It may improve how a result communicates its information and make the page eligible for certain rich results.
Which schema markup format should I use?
JSON-LD is generally the preferred format because it keeps structured data separate from the visible HTML and is easier to maintain. Microdata and RDFa are also supported, but the chosen format must accurately describe the page.
How do I test schema markup?
Use Google’s Rich Results Test to check whether the page contains recognised structured data and may qualify for supported rich results. After publishing, inspect the live URL and monitor relevant reports in Google Search Console.
Can schema markup guarantee a voice search or AI citation?
No. Accurate schema markup may help voice search systems and AI platforms understand a page’s subject and relationships, but visibility still depends on content quality, relevance, trust signals and each platform’s own systems.
Conclusion
Schema markup gives search engines clearer information about your products, content, organisation and locations. It can improve eligibility for rich results, but it can’t guarantee a particular appearance or raise rankings on its own.
Use the most specific accurate type, connect related entities carefully and keep the data aligned with the visible page. Validate it before publishing, monitor it after changes and treat accurate content as the foundation. Accurate structured data supports interpretation across traditional search visibility, voice search and AI-generated answers, but it can’t guarantee a result.
