Skip to main content

​
What Traceten does

When someone reads about your product in ChatGPT, Perplexity, or Google AI Overviews and then visits your site, most analytics tools miss it. The AI assistant doesn’t pass a referrer header, so the visit shows up as “direct” traffic. You have no idea whether that conversion came from an AI recommendation or a bookmark. Traceten solves this. The tracking snippet collects behavioral and contextual signals from every visit and runs them through a four-stage detection pipeline:
  1. Rule-based matching: known AI referrer hostnames, UTM sources, and user-agent patterns. Runs in under 1ms.
  2. ML classifier: a machine-learning model trained on behavioral signals, for traffic with no referrer at all. Runs in under 10ms.
  3. Cross-customer pattern matching: anonymized signal sharing across consenting customers. Rolling out.
  4. Cryptographic verification: signed requests from AI assistants that support them. Rolling out as adoption grows.
The result is a confidence score (0.0 to 1.0) and a named AI source for each detected visit.

​
What you can measure

  • Which AI sources send traffic to your site (ChatGPT, Claude, Perplexity, Gemini, Copilot, and more)
  • How that traffic converts compared to search and direct
  • Which pages AI assistants recommend most
  • Revenue attributed to AI-referred traffic via Stripe and Shopify

​
How to get started

Quickstart

Install the snippet and see your first AI traffic event in under 5 minutes.

Install guide

Platform-specific instructions for HTML, WordPress, Webflow, Shopify, and Next.js.

How detection works

The four-stage pipeline and what each stage catches.

Privacy

What data we collect, what we never collect, and how we handle it.