What is Answer Engine Optimisation (AEO)?

AEO is the practice of structuring your online presence so AI assistants — ChatGPT, Perplexity, Google AI Overviews — cite your business when answering relevant queries. Unlike SEO, AI engines select sources, not pages.

How does Citura calculate my AI visibility score?

Citura analyses 6 dimensions: robots.txt/llms.txt configuration, structured data (schema.org), answer-first content depth, agent.json presence, and active citation signals. Each dimension is weighted; the aggregate is your 0–100 score.

How long before I see results with Citura?

AI citation indexes update slowly — expect meaningful movement in 4–8 weeks. Your score improves within days as optimisations are deployed; actual citations follow as AI engines re-crawl your domain.

Does Citura guarantee AI citations?

No. Citura sells the system and the signals — not a placement. Anyone promising "rank #1 in ChatGPT" is misleading you. We make your business structurally citable; the AI engine decides when to cite.

Is my data safe with Citura?

Citura only analyses publicly accessible data on your domain. No credentials are required. All processing runs on EU infrastructure (Google Cloud Run, region europe-west6).

Is llms.txt an official standard?

llms.txt is a community-proposed standard, not an official W3C or IETF specification. It was proposed by Jeremy Howard (fast.ai) in September 2024 and has been adopted by a growing number of websites and AI tools. The specification is maintained at llmstxt.org. It is voluntary — AI systems that support it benefit; those that do not ignore it without harm.

Does llms.txt replace sitemap.xml or robots.txt?

No. The three files serve different purposes. robots.txt controls crawler access (which bots can crawl which paths). sitemap.xml lists URLs and their priority for indexing. llms.txt describes site content in natural language for AI comprehension. A fully optimized domain has all three. llms.txt is additive, not a replacement.

How long should my llms.txt be?

The introductory description (the blockquote after the site name) should be 50–100 words — concise enough to be used as a direct excerpt. Total file length is flexible, but keep it under 2,000 words. AI systems that process the file for comprehension benefit from conciseness; very long files are less likely to be read in full.

Should I list every page on my site?

No — list your most important pages (5–20 is typical). Focus on pages that answer the questions your target customers ask, pages that describe your core services or products, and any resource or guide pages that establish your expertise. Exclude login pages, checkout flows, duplicate content, and thin pages.

llms.txt: What It Is and How to Create One

llms.txt is a plain-text file served at your domain root that tells AI systems what your site is about and which pages to prioritize. Learn the exact format, what to include, and how it increases AI citation probability.

llms.txt is a Markdown file served at yourdomain.com/llms.txt that describes your website to AI language models in plain language. Proposed by Jeremy Howard (fast.ai) in 2024, it functions as a machine-readable summary of your site's purpose, structure, and key pages — helping AI systems understand and cite your content more accurately.

Why llms.txt exists

AI crawlers process web pages primarily through HTML parsing, which is noisy — navigation menus, cookie banners, ads, and boilerplate text all compete with actual content. robots.txt tells crawlers where to go; llms.txt tells AI systems what your site is. The file is voluntary, plain-text, and served from the root of your domain. It takes under 30 minutes to create and can meaningfully improve how AI systems represent your business in synthesized answers.

As of mid-2026, Perplexity, several ChatGPT plugins, and an increasing number of AI agent frameworks actively check for llms.txt when analyzing a domain. Adoption is accelerating following the publication of the llmstxt.org specification.

The llms.txt format

The file uses standard Markdown with a specific structure:

# [Business/Site Name]

> [One-paragraph description: what the site is, who it serves, what problems it solves. 50–100 words. This is the excerpt AI systems use most frequently.]

## [Section 1 — e.g., Key Pages]
- [Page Title](https://yourdomain.com/page): Brief description of what this page answers

## [Section 2 — e.g., Products/Services]
- [Service Name](https://yourdomain.com/service): What it does, who it's for

## [Section 3 — Optional: About]
- [About](https://yourdomain.com/about): Founded [year], [location], [key credential]

## Optional
- [Privacy Policy](https://yourdomain.com/privacy)
- [Contact](https://yourdomain.com/contact)

A complete example: accounting firm

# Fiduciaire Dupont SA

> Fiduciaire Dupont SA is a Swiss accounting and tax advisory firm serving SMEs in French-speaking Switzerland since 1998. Specialties include VAT registration for foreign companies, Swiss payroll compliance, and annual financial statements under Swiss GAAP. Based in Geneva, serving clients across Geneva, Vaud, and Fribourg cantons.

## Services
- [VAT Registration](https://dupont-fiduciaire.ch/tva): Swiss VAT registration for foreign companies and new Swiss businesses
- [Payroll Services](https://dupont-fiduciaire.ch/salaires): Compliant payroll processing, AHV/AVS declarations, monthly reporting
- [Annual Accounts](https://dupont-fiduciaire.ch/comptes-annuels): Swiss GAAP financial statements, statutory audit coordination

## Resources
- [Swiss VAT Guide](https://dupont-fiduciaire.ch/ressources/tva-suisse): How Swiss VAT works for foreign companies entering the market
- [AHV Contribution Rates 2026](https://dupont-fiduciaire.ch/ressources/ahv-2026): Current social insurance contribution rates

## About
- [About us](https://dupont-fiduciaire.ch/a-propos): Founded 1998, 12 staff, member of EXPERTsuisse

What to include and what to avoid

Include: factual descriptions with verifiable specifics (dates, locations, certifications), your most important pages with clear descriptions of what question they answer, and any unique credentials or coverage areas that differentiate you.

Avoid: marketing claims without substance ("the leading provider of..."), duplicate content from your homepage, broken links, and pages that are behind authentication (AI systems cannot access them).

Technical implementation

Serve llms.txt as a static file from your domain root with MIME type text/plain; charset=utf-8. In Next.js (App Router), place it in /public/llms.txt — it will be served automatically at /llms.txt. In WordPress, upload via the File Manager or add a route in your server config. The file should be accessible without authentication and not blocked in robots.txt.

Optionally, you can also serve llms-full.txt at the same level — a more detailed version that includes full page content for AI systems that request depth over brevity.

Measuring the impact

There is no direct analytics event when an AI system reads llms.txt. The measurable impact comes indirectly: track your AI citation frequency before and after adding the file (using Citura's monitoring or manual queries to Perplexity/ChatGPT). Sites with well-structured llms.txt files typically see more accurate AI descriptions of their business within 2–4 weeks of publication.