In 2026, the viewability playing field has fundamentally changed. Classic blue links now compete with AI-generated answers, AI Overviews, and chat-based search engines that synthesize information from multiple sources. In the middle of all this is a technology that has been around for over a decade but has never been more important:structured data. It's the machine-readable description of your content - and it increasingly determines whether you're seen, cited and clicked on.
What is Structured Data and Schema.org?
Structured data is a standardized way of labeling the content of a web page so that search engines and AI models can understand it without guessing. Instead of a machine interpreting that "Anna Svensson" is an author and "1,290" is a price based on context, you speak about it explicitly.
Schema.orgis the common vocabulary that Google, Microsoft, Yandex and several others founded back in 2011. It defines thousands of entity types –Product, Article, Organization, LocalBusiness, Eventsand many more – with associated properties. Schema.org is the dictionary;JSON-LDis the format in which you write the sentences.
Why JSON-LD won
There are three formats for implementing Schema.org: Microdata, RDFa, and JSON-LD. Google explicitly recommends JSON-LD, and if you're starting a new implementation in 2026, it's effectively the only option worth considering.
The reason is simple. JSON-LD (JavaScript Object Notation for Linked Data) lives in a separate<script type="application/ld+json">-block, detached from the visible HTML code. Microdata and RDFa weave attributes directly into the markup, making them fragile and difficult to maintain at scale. Instead, JSON-LD can be templated, dynamically generated, and validated separately—a critical advantage on sites with hundreds or thousands of pages.
From rich results to AI citations
Historically, structured data has been aboutrich results- the visually enhanced hits in Google with star ratings, prices, FAQ accordion and recipes. They still drive significantly higher click-through rates than a standard blue link, and for e-commerce, recipes, events and local businesses, they remain central.
But in 2026, a new dimension has become at least as important. Google's Gemini-poweredAI ModeandAI Overviewsuses schema markup to verify assertions, establish relationships between entities, and assess a source's credibility during response synthesis. Schemas that accurately describe your content increase the likelihood that you will be cited in an AI response—even when no traditional rich result appears.
GEO and AEO: the new abbreviations
Two concepts dominate the discussion.GEO(Generative Engine Optimization) is about being cited as a source by generative models such as ChatGPT, Claude and Perplexity.AEO(Answer Engine Optimization) focuses more on AI-powered search features such as Google's AI Overviews. Both build on – not replace – traditional SEO.
The common denominator is that the machine must be able to decidewhoorWhatyour source is with high security. Here haveentity disambiguationhas become the most leveraged implementation: pure Organization and Person markup, linked to authoritative identifiers viasameAs, allows an AI to reliably resolve your identity and thus quote you more often. The brands that succeed best with GEO 2026 are consistently the same ones that already have a strong SEO foundation.
The main schedule types in 2026
Schema.org hosts thousands of types, but a handful cover most real-world needs. Start with the basic trio –Organization, BreadcrumbListand the right content type – and build from there.
Organization- the company's identity: name, logo, contact details and
sameAs-links to official profiles. The basis for both knowledge panel and AI entity recognition.Website- describes the site as a whole and can activate the search box in the search result viaSearchAction.
Article / BlogPosting– for guides, news and editorial content. Enter author, date of publication and publisher.
ProductwithVictims, ReviewandAggregateRating– the backbone of e-commerce markup and product-rich results.
LocalBusiness– for physical and service area-based operations. Runs the local package and connection to the Google Business profile.
BreadcrumbList– shows the page's place in the hierarchy. The last step (the current page) should be omitted
items- property.Events– date, venue and ticket information for events.
What's changed – FAQ and HowTo
One of the biggest changes in 2026 concerns pipesFAQPageandHow To. The HowTo results were phased out as early as 2023. And theMay 7, 2026stopped showing FAQ-rich-results in Google Search completely. Google plans to remove the search appearance itself, the Search Console report, and the Rich Results Test support in June 2026, and the Search Console API support in August 2026.
Important nuance:FAQPageis still a valid Schema.org type, and Google continues tointerpretthe markup to understand the page – it just no longer provides a visual SERP effect. So you don't need to clean up existing FAQ markup, but stop expecting accordions in the search results. Additionally, in June 2025, Google retired seven less-used structured data types. The moral: build on markup that actually describes your content, not in search of a specific visual effect that might disappear.
Common mistakes that sabotage your markup
Structured data is unforgiving – small errors can silence the full effect or, worse, trigger manual action. The most common pitfalls:
Markup that doesn't match visible content.The single most important rule: schema must describe what the user actually sees on the page. Flagging hidden, misleading or irrelevant content is considered spam and may result in a manual penalty.
Made up ratings and prices.To insertAggregateRatingwithout real reviews, or prices that are not correct, violate the guidelines.
Incomplete required fields.If a mandatory property is missing, the entire object becomes invalid for rich results.
Broken or duplicate entities.Multiple Organization blocks with different tasks confuse both Google and AI. Have a single canonical source.
Forgotten
sameAs.Without links to official profiles and records, entity disambiguation – the very thing that determines AI citations – becomes more difficult.Inconsistent breadcrumb.The labels in the markup must match the visible breadcrumb trail exactly.
Tools you should use
Always validate before commissioning. Three tools go a long way:
Google's Rich Results Test– checks if your markup qualifies for Google's supported rich results.
Schema.org Validator– checks broader validity against the Schema.org vocabulary, even for types Google doesn't render.
Search Console– URL inspection and improvement reports show how Google actually interprets your live-marked pages over time.
A pragmatic workflow: write JSON-LD in a template, validate in Rich Results Test, deploy, and follow up in Search Console after indexing. For large sites, the generation from CMS data is automated so that the markup always reflects real content.
Then you get started right
Structured data is not a magic button, but a foundation. It makes content you already have comprehensible to machines – and in a world where search is increasingly answered by AI, it's the difference between being a source that gets cited and a site that gets overlooked. Start with Organization and your most important content type, keep the markup honest and in sync with the page, and validate with discipline.
Want your site to be both Google and AI readable without having to keep track of every Schema.org update? At ZORC, we build in correct, maintainable JSON-LD from the ground up and couple it with a technical SEO and GEO strategy. Count on your project inthe quote calculatoror get in touch viacontactthen we look at your structured data.
