What is JSON-LD?
JSON-LD (JavaScript Object Notation for Linked Data) is Google's recommended format for structured data. It sits in its own application/ld+json script block, separate from HTML templates and easy to maintain.
JSON-LD vs. Microdata: which is better?
JSON-LD, in almost every case. Microdata nests properties in HTML attributes (itemprop, itemscope), which makes templates fragile. JSON-LD can express complete graphs and is recommended by Google.
What is an @id graph?
A schema graph connected through stable @id URIs: an Article references its author (Person) and publisher (Organization) by @id. The result is a consistent mini knowledge graph per domain, the precondition for readability by the Knowledge Graph and LLMs.
How does schema relate to the Knowledge Graph?
Schema markup is the most important declarative signal you can send to Google's Knowledge Graph. sameAs properties link your entity to Wikidata, LinkedIn, Crunchbase or ORCID, which makes it readable as a consistent node. That is a precondition for a Knowledge Panel, but whether one appears is Google's decision.
Which rich results still work in 2026?
Active: Product, Recipe, Event, JobPosting, Course, Movie, BreadcrumbList, Article, LocalBusiness, VideoObject. Retired: per Google's changelog, FAQ rich results no longer appear since 7 May 2026; HowTo went earlier. FAQPage markup stays valid as semantic structure but triggers no SERP feature.
Which tools are used for validation?
The Rich Results Test, the Schema.org Validator (validator.schema.org), Bing's markup validation and our own graph coherence parser for @id references and property coverage, automated as a CI gate before every deployment.
What does schema mean for GEO and LLM citations?
Google documents (A): no special markup is required for the AI features in Google Search; structured data is not an AI ranking switch. Our own data (C): in our audit cohort, consistent Person and Organization graphs with a curated sameAs cluster correlate with more stable entity resolution across models. That is an observation, not a Google commitment. Bottom line: schema is groundwork, not a secret GEO lever.
How is schema maintenance organized?
Drift monitoring after content updates, CI/CD validation gates before every deployment and quarterly audits with a coverage report. Schema is an ongoing process, not a one-off.
Which KPIs does a schema implementation track?
Primary: validation pass rate (target 100 %), @id graph coherence, property coverage per content type and rich result impressions in Search Console. Secondary: Knowledge Panel emergence, citation rate in AI search, sameAs cluster size and Wikidata links.
What does a professional schema implementation cost?
Three models: audit sprint (4-6 weeks), implementation project (8-16 weeks) and continuous schema engineering (6-12 month retainer). Terms depend on stack complexity, content types and markets.
How does onboarding work?
Four steps in the first 14 days: stack discovery, a crawl and validation audit, gap analysis with a property matrix per content type, and an implementation plan with @id namespace and inheritance rules.
