AI-powered search engines — ChatGPT, Perplexity, Google AI Overviews, Bing Copilot — are reshaping how people find services online. Traditional SEO focused on pleasing Googlebot. The new challenge: making your content easily digestible by large language models.

We run navetteforyou.com, a private airport transfer service in Marrakech, Morocco. Here is how we implemented llms.txt and optimized our site for AI crawlers.

What Is llms.txt?

The llms.txt specification is a plain-text file (similar to robots.txt) that provides AI crawlers with a structured summary of what a website offers. Unlike HTML pages full of navigation, scripts, and styling, llms.txt gives LLMs clean, parseable information.

Our llms.txt Structure

# Marrakech Transport > Private airport transfers and day trips in Morocco ## About Marrakech Transport (navetteforyou.com) connects guests arriving at Marrakech Menara Airport with private drivers for transfers to hotels, riads, and destinations across Morocco. Operated by Serenity Morocco Tours. ## Services - Airport transfers (Marrakech Menara RAK) - Hotel and riad pickups - Day trips: Essaouira, Ourika Valley, Ait Benhaddou, Imlil - Multi-day tours: Sahara, Fes, Chefchaouen ## Contact - Website: https://www.navetteforyou.com - WhatsApp: +212 701 664 704 - Email: info@serenitymoroccotours.com ## Languages ar, de, en, es, fr, it, ja, pt 

Why This Matters for GEO (Generative Engine Optimization)

When someone asks ChatGPT "How do I get from Marrakech airport to my hotel?", the model needs structured, authoritative content to cite. Our llms.txt file provides:

  1. Clear entity identification — The business name, operator, and service type are stated plainly.
  2. Service scope — What we do and do not offer.
  3. Contact methods — Direct booking channels.
  4. Language coverage — Which locales the site serves.

Beyond llms.txt: The Full GEO Stack

1. Schema.org Entity Graph

We define the operator as a Schema.org Organization with a canonical @id and link it to 15+ verified sameAs profiles (Wikidata, YouTube, Instagram, LinkedIn, etc.). Every page references this single entity.

2. Passage-Level Citability

Each route page starts with a self-contained opening paragraph that answers the core question: "How do I get from A to B?" This gives AI models a clean passage to quote.

3. FAQ Markup

Every route page includes FAQPage structured data. LLMs love FAQ content because Q&A pairs map directly to the query-answer format these models use.

4. Multi-Platform Presence

We maintain consistent content across Medium, Substack, Dev.to, GitHub Pages, and other high-authority platforms. When multiple sources corroborate the same entity and service, AI models gain confidence in citing it.

Implementation Tips

# Serve llms.txt from your domain root # In Astro/Next.js, add to public/ directory cp llms.txt public/llms.txt # Verify it is accessible curl https://www.navetteforyou.com/llms.txt 

Keep the file under 2,000 tokens. AI crawlers have context windows too — a concise file gets fully processed; a bloated one gets truncated.

Measuring Impact

Track these signals to see if GEO is working:

  • Brand mentions in AI responses — Ask ChatGPT, Perplexity, and Copilot about your service category and location.
  • Referral traffic from AI platforms — Check your analytics for traffic from chat.openai.com, perplexity.ai, etc.
  • Structured data validation — Use Google's Rich Results Test regularly.

The Takeaway

SEO is not dead — it is evolving. The sites that win in 2027 will be the ones that speak both Googlebot and GPT. An llms.txt file takes ten minutes to create and costs nothing to maintain. If you run a local service business, start there.

Marrakech Transport provides private airport transfers and day trips across Morocco. Operated by Serenity Morocco Tours. Book at navetteforyou.com or WhatsApp +212 701 664 704.