Your business isn't showing up in ChatGPT for one of five fixable reasons: your site blocks AI crawlers, your content is invisible without JavaScript, you're missing structured data, you have no llms.txt file, or your business has weak "entity" signals across the web. Most sites fail at least two of these — and most owners have no idea, because there's no error message and nothing shows in analytics. Here's how to diagnose and fix each one.
First, understand how ChatGPT picks businesses to recommend
When someone asks ChatGPT "best wedding photographer in Austin" or "boutique hotel near Lake Tahoe," the answer comes from two paths:
- Live search. ChatGPT fetches current pages (leaning heavily on Bing's index) and reads them on the spot. If its crawler can't access or can't parse your site, you cannot appear — no matter how good your business is.
- Training knowledge. What the model already learned about your business from its training data: your site, reviews, directories, articles, Reddit threads.
Both paths fail silently. A customer asks, your competitor gets named, and you never know the question was asked. That's the uncomfortable part: zero-click loss is invisible in your analytics.
Reason 1: Your site is blocking AI crawlers (the most common one)
AI systems use their own crawlers — GPTBot (OpenAI), ClaudeBot (Anthropic), PerplexityBot, Google-Extended (Gemini). Two things block them:
- robots.txt rules. Many sites (or their SEO plugins, or a developer in 2023) added
Disallowrules for AI bots and forgot about them. - Firewalls and bot protection. Cloudflare's "Block AI bots" toggle, aggressive WAF rules and some hosting defaults block AI crawlers at the network level — robots.txt looks fine, but the bots get a 403.
How to check: open yoursite.com/robots.txt and look for GPTBot, ClaudeBot or PerplexityBot under a Disallow rule. Firewall blocks are harder to spot by eye — a scanner that actually fetches your site the way a bot does will catch it.
Reason 2: Your content only exists with JavaScript
Most AI crawlers read your raw HTML — they don't run JavaScript. If your site is built on a JS framework that renders content client-side, or a website builder that loads text dynamically, an AI crawler may see a nearly empty page: no prices, no services, no address.
How to check: in Chrome, right-click → "View page source" (not Inspect) and search for a sentence from your homepage. If your actual text isn't in the source, AI can't read it either.
Reason 3: No structured data (schema markup)
Schema.org JSON-LD tells machines unambiguously what you are: a Hotel, a LocalBusiness, a Product with a price. Sites with clean schema are far easier for AI to represent accurately — sloppy or missing schema is why AI gets your city, prices or offering wrong.
How to check: paste your URL into Google's Rich Results Test, or view source and search for application/ld+json.
Reason 4: No llms.txt file
llms.txt is a plain-text file at your site root that gives AI models a clean summary of your business — what you do, what you offer, how to reach you. It's the newest signal on this list and takes five minutes to add. We built a free llms.txt generator that writes the file for you.
Check all five reasons in 30 seconds
The free RankInAI scan tests your robots.txt, firewall behavior, JavaScript-blindness, schema and llms.txt — and gives you a 0–100 AI Visibility Score with exact fixes.
Scan My Site Free — No SignupReason 5: Weak entity signals across the web
AI models learn about your business from everywhere, not just your site. If your name, address and description are inconsistent across Google Business Profile, Yelp, directories and social profiles — or you're simply mentioned almost nowhere — the model has low confidence you exist, and low-confidence businesses don't get recommended.
- Make your business name, location and one-line description identical everywhere.
- Get mentioned in places AI reads heavily: local news, niche directories, Reddit threads where people genuinely ask for recommendations.
- Keep reviews flowing on the platforms relevant to your niche — AI models cite businesses with visible review consensus.
Quick reference: symptom → cause → fix
| Symptom | Likely cause | Fix |
|---|---|---|
| ChatGPT says it "can't find information" about your business | Crawlers blocked (robots.txt or firewall) | Remove AI-bot Disallow rules; turn off "block AI bots" in Cloudflare |
| AI describes your business wrong (city, prices, offering) | Missing/invalid schema, no llms.txt, inconsistent listings | Add JSON-LD schema + llms.txt; align name/address/description everywhere |
| Competitors get named, you don't — despite better reviews | JS-hidden content or weak entity signals | Ensure text is in raw HTML; build consistent mentions |
| You appear in Google but never in AI answers | Blocked AI crawlers with Googlebot allowed (very common) | Check robots.txt for GPTBot/ClaudeBot specifically |
How long do fixes take to show up?
ChatGPT's live-search path can reflect fixes within days to weeks — it fetches current pages. The model's built-in knowledge updates more slowly, over months. Fix crawler access and structured data first: those control the live-search path, which is where most buying questions get answered.
Start by finding out exactly where your site fails: run the free scan, get your score, and apply the fixes in an afternoon.