AI Visibility Lab · Switzerland
How to Get Your Business Cited by ChatGPT
Getting cited by ChatGPT, Claude, and Perplexity is not luck. It comes down to specific technical configurations and content decisions that AI models use to decide whether to extract and cite your content. Here is exactly what to do.
Check your AI citation score free →Step 1 — Allow AI crawlers in robots.txt
This is the single most common mistake. Every major AI model uses a named crawler. If your robots.txt doesn't explicitly allow them, you don't exist to those models.
User-agent: GPTBot
Allow: /
User-agent: ClaudeBot
Allow: /
User-agent: PerplexityBot
Allow: /
User-agent: Google-Extended
Allow: /
Add these to your /robots.txt. Without them, ChatGPT's GPTBot, Anthropic's ClaudeBot, and Perplexity's crawler are blocked.
Step 2 — Add Organisation schema
AI models use structured data to identify real-world entities. Without an
Organization schema block, your business is an anonymous website — not a
citable entity. Add this to your homepage's <head>:
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "Organization",
"name": "Your Business Name",
"url": "https://yoursite.ch",
"description": "One sentence: who you help, what problem you solve.",
"areaServed": { "@type": "Country", "name": "Switzerland" }
}
</script>Step 3 — Write answer-first
AI models extract the first substantive paragraph from your pages. If that paragraph is "Welcome to our website" or a mission statement, it has nothing citable. Every page should open with a direct answer to the question it targets.
Don't
"Welcome to Zürich Nutrition Clinic. We are passionate about helping you achieve your health goals through personalised nutritional guidance..."
Do
"Zürich Nutrition Clinic provides evidence-based nutritional therapy for adults managing chronic conditions, based in Zürich-Oerlikon. Appointments available in German and English."
Step 4 — Add FAQ sections with schema
FAQ content is the most directly citable format for AI models. A question with a clear, factual answer is exactly what AI assistants want to surface. Mark them up with FAQPage schema and they become prime citation candidates. Target the exact questions your customers ask before calling you.
Step 5 — Create an llms.txt file
/llms.txt is an emerging standard:
a plain-text file at your domain root that tells AI models directly who you are,
what you do, and what your key pages cover. Think of it as a briefing note for AI.
# Your Business Name
One-sentence description of who you serve and what problem you solve.
## Services
- Service 1: brief description
- Service 2: brief description
## Location
City, SwitzerlandWhat a citable business page looks like
ChatGPT is more likely to cite a page when the page contains concrete, attributable facts. The model needs to identify the business, understand the service, and extract a useful answer without guessing. Vague positioning copy is weak evidence; specific service and location statements are stronger evidence.
| Page element | Weak version | Citable version |
|---|---|---|
| Opening sentence | We help businesses grow online. | AI Visibility Lab audits Swiss SME websites for citation readiness across ChatGPT, Claude, Perplexity, and Google AI Overviews. |
| Service description | We offer AI consulting. | The audit checks schema, AI crawler access, answer-first content, FAQ markup, entity clarity, and /llms.txt coverage. |
| Location signal | Serving clients everywhere. | Based in Switzerland, focused on Swiss SMEs and multilingual Swiss market signals. |
| Proof point | We get great results. | Each engagement records a baseline audit, a fix pack, and a re-measurement so the before/after change is visible. |
| FAQ answer | It depends. | For browsing-based AI systems, robots.txt and schema changes can be visible after the next crawl; base-model knowledge changes more slowly. |
A practical citation-readiness test
Before asking whether ChatGPT cites your business, test whether a human can extract the same answer from the page in under one minute. If the answer is not explicit on the page, an AI model has no reliable reason to cite it.
- Open the target page and copy the first paragraph into a blank document.
- Check whether it names the business, service, audience, and geography directly.
- Find one FAQ answer that matches the exact question a buyer would ask ChatGPT.
- Inspect the page source for Organization, Article, and FAQPage schema.
- Open
/robots.txtand confirm GPTBot, ClaudeBot, and PerplexityBot are allowed. - Open
/llms.txtand confirm the page is listed or described.
- If the page cannot pass this test, rewrite the page before expecting AI citations.
- If the page passes this test but is not cited, compare it against the pages AI models already mention.
- If competitors are cited instead, add clearer examples, stronger entity signals, and more specific Swiss-market context.
Frequently asked questions
Does ChatGPT crawl my website? +
ChatGPT uses a crawler called GPTBot. If your robots.txt blocks it, your site won't appear in ChatGPT's browsing results. Add an explicit Allow directive for GPTBot.
How long until I get cited after making changes? +
For real-time browsing (ChatGPT with web search, Perplexity): days to weeks after your next crawl. For base model training data: longer, as that updates with new training runs. Perplexity tends to index fastest.
What content does ChatGPT prefer to cite? +
Answer-first paragraphs, clear factual statements, FAQ sections, and content from sites with consistent topical coverage. Marketing language and vague descriptions are not cited.
Is this the same for Claude and Perplexity? +
The same principles apply: allow the crawler, add entity schema, write answer-first. Each model has different crawler names (ClaudeBot, PerplexityBot) and slightly different weighting, but the core technical requirements are identical.