Cleotic

Source Tracking

Sources combines two related traffic signals:

  • AI referrals -- real visitors arriving from AI assistants
  • AI crawlers -- crawler visits from AI systems indexing your content

This page focuses on tracking setup in Project Settings: how to connect your site, record crawler visits, and manage tracking sites. Sources shows the resulting AI traffic evidence.

Why track AI crawlers?

Understanding AI crawler activity on your site helps you:

  • Confirm your content is being indexed. If GPTBot visits your blog regularly, ChatGPT likely has access to that content for its responses.
  • Identify gaps in indexing. If certain AI crawlers never visit your site, those models may have limited knowledge of your brand.
  • Prioritise content placement. Pages that AI crawlers visit frequently are more likely to influence AI responses.
  • Correlate crawling with visibility. If crawler visits increase and your visibility scores rise, your content strategy is working.

Two tracking methods

A small JavaScript snippet you add to your website's HTML. When a page loads, it pings Cleotic to record the visit.

Best for: AI-powered browsers that execute JavaScript, such as Arc Search and Perplexity Comet.

Limitation: Most AI training crawlers (GPTBot, ClaudeBot, CCBot, etc.) only fetch HTML and don't run JavaScript. They won't trigger the pixel.

For most comprehensive tracking, choose the server-side integration that matches your platform. See the WordPress setup guide or the Vercel setup guide.

Connecting a site

Open Project Settings -> Tracking setup and click Connect a site. The connection panel keeps the work in three steps:

  1. Platform -- Choose WordPress, Vercel, Bunny CDN, or HTML snippet / Other.
  2. Site details -- Enter the owned domain and an optional name. Continuing from this step creates the tracking site, so you can close the panel and resume later without losing it.
  3. Connect -- Follow the instructions for the selected platform, save or generate its credential, and click Finish setup.

Use Back or the completed step labels to revisit an earlier step. Changing a configured platform clears incompatible credentials and starts verification again; historical analytics stay intact.

After you finish, the site shows Waiting for signal. Cleotic changes it to Connected automatically after the first valid, matching traffic signal arrives. Use Finish later if the provider-side work is not complete yet.

  1. Click Connect a site, or Continue setup on an unfinished site
  2. Choose HTML snippet / Other
  3. Copy the HTML code from the Connect step
  4. Paste it into your website's HTML, just before the closing </body> tag
  5. Deploy your site
  6. Confirm the snippet is installed and click Finish setup

The site becomes Connected after Cleotic receives its first same-domain beacon. The pixel records visits from JavaScript-capable AI browsers.

Managing tracking sites

From Project Settings -> Tracking setup, you can:

  • Toggle active/paused -- Pause a tracking site to stop recording visits without deleting configuration. Incoming server-side tracking data is silently dropped while paused.
  • Continue setup / Review setup / Manage connection -- Open the same connection panel at the Connect step. You can copy endpoints or snippets, generate or rotate provider credentials, and revisit platform or site details.
  • Delete -- Remove the tracking site and its data

Connection status means:

  • Needs setup -- The site record exists, but setup has not been finished.
  • Waiting for signal -- Required configuration is saved and Cleotic is waiting for qualifying traffic.
  • Connected -- Cleotic has received a valid signal matching the configured site.

Only active tracking sites count towards plan capacity. Starter includes 1 active tracking site; higher-tier capacity is shown in the public Pricing page and the live Settings -> Plan & Usage meter.

One tracking site per domain

Cleotic matches incoming requests to tracking sites by domain, so only requests matching the configured domain are recorded.

What gets tracked

The tracking pixel and beacon record visits from identified AI crawlers. Server-side tracking stores eligible production pageview events with AI crawler and referral flags, while filtering out assets, API calls, framework internals, redirects, preview traffic, and failed requests. Each recorded crawler visit includes:

  • The URL and page path that was visited
  • Which AI crawler was identified (e.g., GPTBot, ClaudeBot)
  • The tracking method that detected it (pixel, beacon, or server-side tracking)
  • The page title at the time of the visit
  • The timestamp

Sources also includes:

  • An AI citations explorer for source evidence from model responses
  • An AI traffic surface for page-level AI visits, crawler evidence, referral evidence, and crawler-to-referral gaps

See Crawler analytics to learn how to read crawler evidence inside the AI traffic explorer.