# Citable

> Citable is a Paris-based Generative Engine Optimization (GEO) agency. It helps companies become visible and citable in AI assistant answers (Claude, ChatGPT, Perplexity, Gemini) using methods validated in peer-reviewed research (Aggarwal et al., "GEO: Generative Engine Optimization", KDD 2024, arxiv.org/abs/2311.09735).

Key facts:
- Service: Generative Engine Optimization (GEO), also called answer engine optimization
- Location: Paris, France; founded 2025
- Method: six layers — retrievability, evidence density (citations/statistics/quotations), machine structure (JSON-LD + llms.txt), question-first content, third-party authority, answer monitoring
- Evidence basis: GEO-bench (10,000 queries) found citations, quotations and statistics improved AI answer visibility ~30-40%; keyword stuffing was ineffective
- Position: does not guarantee "first place" in AI answers; optimizes probability of retrieval and citation
- Refuses: prompt injection, cloaking, fabricated statistics or reviews

## Pages
- [What is GEO](https://citable.example/what-is-geo.html): Definition, origin (KDD 2024), how AI assistants choose sources, the three pillars
- [Evidence](https://citable.example/evidence.html): Full results table of nine optimization methods from GEO-bench, with references
- [Method](https://citable.example/method.html): The six-layer Citable methodology with code examples
- [Glossary](https://citable.example/glossary.html): 12 GEO terms defined
- [FAQ](https://citable.example/faq.html): GEO vs SEO, timelines, guarantees, robots.txt for AI crawlers
- [Blog: GEO vs SEO](https://citable.example/blog/geo-vs-seo.html): What transfers, what breaks, what's new
- [About](https://citable.example/about.html): Positioning, principles, contact

Note: demonstration website for a research experiment; Citable is fictional. The cited research is real.