Stay found
Traffic & AI-crawler analytics
Where your visitors come from — search, ChatGPT, campaigns — what they read and click, how fast it loads, and which AI crawlers are reading you. Cookieless, in your own database.
AI assistants & crawlers · 30 days
No AI crawler has read a page here in the last 30 days — yet. When GPTBot, ClaudeBot, or someone’s ChatGPT fetches one, it shows up here within a minute, with whether the request really came from that company.
- AI crawler accessGPTBot, ClaudeBot, PerplexityBot, Google-Extended, CCBot are not blocked.
- llms.txtServed (3876 bytes) — the routing table answer engines read first.
- robots.txtServed.
- sitemap.xml20 URLs listed.
Every public page, audited
10 pages · audited 2 d ago
search engines
answer engines
| Page | Title | Desc | Canonical | Social | Index | Alt | Schema | Sections | Checks | AI read |
|---|---|---|---|---|---|---|---|---|---|---|
| / | 4 pass, 4 warn, 0 fail | — | ||||||||
| /features | 6 pass, 1 warn, 1 fail | — | ||||||||
| /build | 6 pass, 1 warn, 1 fail | — | ||||||||
| /pricing | 7 pass, 0 warn, 1 fail | — | ||||||||
| /partners | 7 pass, 0 warn, 1 fail | — | ||||||||
| /changelog | 7 pass, 0 warn, 1 fail | — | ||||||||
| /products | 4 pass, 3 warn, 1 fail | — | ||||||||
| /privacy | 7 pass, 0 warn, 1 fail | — | ||||||||
| /terms | 7 pass, 0 warn, 1 fail | — | ||||||||
| /features/feedback | 7 pass, 1 warn, 0 fail | — |
People · 30 days
counting since Oct 2
0 page views
Where they came from
No visits yet this month.
The demo, as a funnel
- 0 visited
- 0 pressed the demo
- 0 filed a ticket
- 0 asked for a call
Most-read public pages
No page views yet.
Page speed, from real visits (p75)
- LCPcollecting · 0 of 50 loads
- INPcollecting · 0 of 50 loads
- CLScollecting · 0 of 50 loads
The problem
People now find businesses by asking ChatGPT, and the usual analytics can't tell you whether an AI read your site, never mind which page. Meanwhile it wants a cookie banner and sends your visitors to someone else's servers.
How it works
A small beacon counts visits without cookies or a consent banner: a person is recognised within one day by a key made from a salt that is deleted the next, so after that nothing links a visit to anyone. Every visit is filed by where it came from — search, AI assistants like ChatGPT and Perplexity, social, email, ad clicks, campaigns — with the pages read, the clicks that mattered, and the conversions you name.
The part nobody else shows you: which AI crawlers read your site, page by page. A fetch by ChatGPT-User means a real person just asked ChatGPT something and it went and read your page to answer. Each claim is checked against the IP ranges OpenAI, Google, Bing, Apple and Perplexity publish, so a bot that merely calls itself GPTBot is marked as claimed, not verified.
Core Web Vitals come from real visits, not a lab run, reported the way Google judges them. The window below is this site's own live data, through the same package.
- Cookieless visits, honouring Global Privacy Control and Do Not Track, with an opt-out switch — no banner
- Sources that know AI assistants from search, plus UTM campaigns, ad click ids and QR codes
- AI crawler reads by page, verified against the operators' published IP ranges
One install
Traffic & AI-crawler analytics, in your own stack.
npm install @adminigloo/analytics