What is Rexxet?
Rexxet is an AI referral tracking and visibility analytics platform that shows which AI assistants like ChatGPT, Gemini, and Perplexity are recommending your brand by capturing referrer data before it is lost in tools like Google Analytics.
Most analytics tools tell you what happened after a visit is logged. Rexxet works at a different layer — it intercepts the moment a visitor arrives, reads where they came from before that information disappears, and builds a picture of your AI visibility that no other tool can see.
Here is exactly how it works, from the moment you paste your snippet to the moment you see your first AI referral in your dashboard.
The core problem: When an AI assistant recommends your brand and a user clicks through, most platforms strip the referrer header. Google Analytics sees "Direct." The visit is logged but the source is lost forever. Rexxet reads the referrer in the browser before it gets stripped.
Step 1 — The Snippet
Everything starts with one line of code. You paste it inside the <head> tag of your website. It loads asynchronously — meaning it has zero impact on your page speed or Core Web Vitals.
<script src="https://app.rexxet.com/track.js?pid=YOUR_PROJECT_ID" async></script>
The pid parameter is your unique project ID, generated when you add your site to Rexxet. It's how the backend knows which account to attribute the hit to. Every site you track has its own ID — your data is completely siloed from other accounts.
Step 2 — The Detection
Every time a visitor lands on your site, the Rexxet script fires. In under 50 milliseconds, it checks document.referrer — the browser's built-in memory of where the user came from.
It compares that referrer against a list of known AI domains:
- chatgpt.com and chat.openai.com → ChatGPT
- gemini.google.com → Gemini
- perplexity.ai → Perplexity
- grok.com → Grok
- claude.ai → Claude
- copilot.microsoft.com → Microsoft Copilot
- meta.ai → Meta AI
- search.brave.com → Brave Leo
- And more added as new AI platforms emerge
If the referrer matches a known AI source, the script captures the source name, the page path the visitor landed on, their device type, and a timestamp. If the referrer is not from an AI source, the script does nothing — no data is sent, no overhead, no noise.
Step 3 — The Signal
Once an AI referral is detected, the script sends a lightweight signal to the Rexxet backend using the browser's navigator.sendBeacon API. This method is specifically designed for analytics pings — it fires after the page loads and does not block any user interaction. The visitor never knows it happened.
The signal contains:
Which AI sent the visit
ChatGPT, Gemini, Perplexity, Grok, Claude, Copilot — the exact platform that referred the user.
Which page they landed on
The exact URL path — /pricing, /about, /blog/post-name — so you know which content AI is citing.
The device type
Mobile, tablet, or desktop — helping you understand how AI-referred users are accessing your content.
The timestamp
Precise to the second, so you can see traffic patterns, spikes, and trends over time.
Step 4 — The Dashboard
Once hits start coming in, your Rexxet dashboard shows you the data in real time. Above your brand perception analysis you see a live metrics strip showing:
This Week's Count
Total AI referrals detected in the last 7 days, with a week-on-week comparison.
Top AI Source
Which AI platform is sending you the most visitors — ChatGPT, Perplexity, Gemini, or others.
Most Cited Page
Which page on your site AI recommends most often. Usually your pricing or about page.
7-Day Sparkline
A visual bar chart showing day-by-day referral volume for the past week.
The dashboard also shows a live feed of recent hits — the last 10 AI referrals with source, page, and time. When you see ChatGPT → /pricing → 4 minutes ago, you know exactly what is happening in real time.
Step 5 — The Brand Perception Audit
Tracking referrals tells you what is happening. The brand perception audit tells you why — and what to fix.
Rexxet uses the same AI models that are sending you traffic to audit how those models currently perceive your brand. You fill in your "Who Am I" — your brand name, what category you operate in, and your unique value proposition. Rexxet then asks the model:
"As an AI language model, what is your trained perception of this brand? How would you describe it to a user who asks about it? What do you know, what do you get wrong, and when would you recommend it over competitors?"
The result is an honest audit of your brand's AI perception — your AI Trust Score, your AI label, the pros and cons the model associates with you, which queries you win and lose, what the model gets wrong about you, and a ready-to-paste "Feed Me Formula" that corrects the gaps.
Step 6 — The Competitor Emulator
For any niche you analyze, Rexxet shows you the top three brands that AI recommends most in that space — and builds you a complete playbook for each one:
- Why AI recommends them — specific reasons based on their content depth, schema markup, and authority signals
- What content formats they use that AI loves — comparison guides, data reports, how-to guides
- What schema types they have that you don't
- Which topic clusters they own
- Your specific opportunity — the angle they haven't covered that you can own
- A step-by-step emulation plan — specific articles to write, formats to use, schema to add
This turns AI visibility from a passive insight into an active content strategy.
The Baseline Period
When you first add your tracking snippet, Rexxet shows you a baseline building progress bar. For the first two weeks it collects data and gives you early signals — how many hits detected, which AI is already sending you traffic. After two weeks you have enough data for meaningful patterns and the full metrics dashboard activates.
This is intentional. AI referral data needs at least a week of baseline to be meaningful. A single spike on day one could be noise. Two weeks of consistent data tells you something real.
Ready to See Your AI Traffic?
Paste one script. See your first AI referral within seconds of your first visit.
Add Your Site Free →What Rexxet Does Not Do
Rexxet does not track individual users. It does not use cookies. It does not store personally identifiable information. The only data it captures is the AI source, the page path, the device type, and the timestamp. There is no fingerprinting, no session tracking, and no cross-site tracking of any kind.
Your visitors' privacy is fully preserved. Your data is fully yours — it stays in your account, it is never sold or shared, and it is stored permanently so you have a growing historical record of your AI visibility over time.