When we launched navetteforyou.com — a private airport transfer service in Marrakech, Morocco — we knew that structured data would be critical for visibility in search results and AI-powered answers. Here is a walkthrough of how we implemented Schema.org JSON-LD for a transport service, and what we learned along the way.

Why Structured Data Matters for Transport Services

Google's rich results, AI Overviews, and platforms like ChatGPT and Perplexity increasingly pull structured data to answer user queries directly. For a niche service like private airport transfers, appearing in these answers can mean the difference between a booking and obscurity.

A tourist searching "Marrakech airport transfer" might never scroll past the first three results. But if your structured data is correct, your service can appear in knowledge panels, rich cards, and AI-generated recommendations.

The Entity Graph Approach

Rather than sprinkling disconnected JSON-LD blocks across pages, we built a unified entity graph. The operator (Serenity Morocco Tours) is defined once as an Organization entity with a canonical @id, and every page references that same node.

export const OPERATOR_ID = "https://www.serenitymoroccotours.com/#organization"; export const OPERATOR_SAME_AS = [ "https://www.instagram.com/serenity_morocco_tours/", "https://www.trustpilot.com/review/serenitymoroccotours.com", "https://www.wikidata.org/wiki/Q142170383", "https://www.youtube.com/@navetteforyou", // ... 15+ verified profiles ] as const; 

Every route page then emits a TouristTrip or Service node that references the operator by @id. Google sees one consistent entity across hundreds of pages rather than fragmented, conflicting markup.

Key Schema Types We Use

1. TouristTrip for Route Pages

Each route (Marrakech to Essaouira, Airport to Hotel, etc.) gets a TouristTrip with:

  • itinerary describing waypoints
  • provider pointing to the operator @id
  • offers with price ranges (we use AggregateOffer for variable pricing)

2. FAQPage for Common Questions

Every route page includes an FAQ section marked up with FAQPage. This is the single highest-ROI schema type for transport services — Google regularly pulls these into featured snippets.

3. BreadcrumbList for Navigation

Static breadcrumbs help Google understand the site hierarchy: Home → Routes → Marrakech to Essaouira.

4. WebSite with SearchAction

We added a SearchAction to the homepage schema, pointing to our route search. This enables sitelinks search in Google results.

The llms.txt Strategy

Beyond traditional Schema.org, we implemented /llms.txt — a plain-text file that tells AI crawlers what the site offers, the service area, and how to cite us. When ChatGPT or Perplexity crawls the site, this file provides a clean, structured summary.

Validation Pipeline

We built a static price-check script that runs before every deploy:

bun scripts/static-price-check.ts # 633 pages checked, 17113 amounts verified # PASS: no invalid prices found 

This prevents accidental publication of incorrect pricing in structured data — a real risk when generating hundreds of route pages programmatically.

Results

Within weeks of deploying the entity graph:

  • Google indexed 600+ pages with valid rich results
  • The Wikidata entity (Q142170383) was picked up by knowledge graph queries
  • AI platforms started citing the service in Marrakech transport recommendations

Key Takeaways

  1. One entity, many references. Define your business once, reference it everywhere by @id.
  2. FAQPage is gold. For service businesses, FAQ schema drives more traffic than any other type.
  3. Validate before deploy. Automated checks prevent embarrassing structured data errors.
  4. Feed the AI crawlers. An llms.txt file is trivial to maintain and positions you for the AI search era.
  5. Build sameAs deliberately. Every verified profile you link strengthens entity confidence.

Marrakech Transport is a private airport transfer and day trip service operated by Serenity Morocco Tours. Book via navetteforyou.com or WhatsApp +212 701 664 704.