See what Claude sends to your site—and what it requests.
Chad reports attributable human visits from Claude separately from Anthropic crawler activity, so a page fetch never masquerades as a customer visit.
Claude traffic analytics includes human browser referrals from Claude plus server-side activity from Anthropic's user, search, and training crawlers. These are separate signals and should be analyzed separately.
Three crawler purposes, one important distinction
Anthropic publishes different user agents for user-requested retrieval, search indexing, and model training. A request from any of them is machine activity—not a human session.
- Claude-User: retrieves a page in response to a user's request.
- Claude-SearchBot: discovers content for Claude's search experiences.
- ClaudeBot: crawls content for model development.
- Claude referral: a person arrived in a browser with an attributable Claude referrer.
Questions Chad can answer
- How many attributable visitors came from Claude this week?
- Which pages did those visitors enter and which goals did they complete?
- Which paths did Claude-SearchBot or ClaudeBot request?
- Did Claude referral traffic change from the previous comparable period?
Use Claude Code to install and query Chad
Your coding agent can add Chad's browser snippet and server adapter to a supported codebase. After setup, Chad's read-only MCP server makes the resulting analytics available inside Claude Code alongside Slack or Teams reports.
Common questions
Does ClaudeBot traffic equal Claude referrals?
No. ClaudeBot is a crawler. A Claude referral is an attributable human browser visit.
Can Chad track Claude on a client-only website?
The browser tracker can measure attributable human referrals. Known crawler requests require server-side visibility, such as a supported adapter or proxy.
Can I ask about Claude traffic from Claude Code?
Yes. Chad provides a private, read-only hosted MCP connection for supported coding agents after you create a credential in Slack.