Entity-based schema implementation for AI search visibility

I implemented JSON-LD schema across multiple client websites to improve search visibility in AI Overviews and AI-driven search platforms. The work focused on clearly defining entities, their attributes, and relationships so page content could be better understood by crawlers.

Process

Structured data was a new concept for me at the time so I started by researching schema, entities, and knowledge graphs.

• Reviewed official documentation and best-practice guides
• Researched schema implementations across other sites
• Developed a repeatable heuristic for myself for writing schema
• Implemented schema using:
• Inline <script type="application/ld+json"> blocks
• Google Tag Manager where appropriate

Languages

• JSON (JSON-LD)
• JavaScript

Tools

• Google Schema Markup Testing Tool
• Schema.org documentation and validator
• Google Tag Manager

Challenges

Understanding how entities are represented and connected within a knowledge graph; modeling relationships between multiple entities on a single page; identifying best practices for different page types, including:

• Landing pages
• Blog articles
• Financial product pages for a financial institution

Solutions

Used @id fields to create stable, reusable entity references across pages; structured schema as a data graph instead of isolated objects; refined my heuristic to begin every implementation by:

• Identifying core entities on the page
• Creating a simple page outline, listing entities
• Defining attributes and relationships in the outline
• Translating the simple outline in properly written JSON structured data