In this guide
- Three kinds of AI traffic exist: crawlers that read pages, assistants that fetch a page for a user and people who click a link in an AI answer.
- Referrals from AI tools can be measured in GA4 by source. Some come with
utm_source=chatgpt.comor a clear referrer. - Self-identifying bots can be separated with a header. Agentic browsers can look like normal Chrome, so treat them as a separate unknown.
A few years ago "bot traffic" meant search engines and scrapers. Now there is a new layer: AI assistants that browse for people, AI crawlers that gather pages and a growing stream of real visitors who arrive from an AI answer. Each has a different effect on your numbers. This page explains how to recognise them, how to measure the ones you want and how to keep the others out of your reports. The details change fast, so check the vendors' documentation. See filtering spam in GA4 with Bot Detection and bot filtering.
Three kinds of AI traffic
| Kind | What it is | How it shows up |
|---|---|---|
| Crawlers | Bots that read public pages to index or train, for example search and training crawlers | Self-identifying user agents such as OAI-SearchBot or GPTBot, no JavaScript analytics |
| Assistants fetching on a user's behalf | A tool fetches a page when a user asks a question | A named user agent in some cases, often no analytics |
| Visitors from AI answers | A person clicks a link in an AI answer | A referrer or UTM such as chatgpt.com; behaves like any visitor |
| Agentic browsers | A browser that acts for a person | May look like ordinary Chrome; hard to distinguish |
Check your tracking in one pass
This is an example. Run the check on your own site to see where yours stands.
Measure the referrals you want
Find AI sources in GA4
In the traffic acquisition report, look at source and medium for entries such as chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com and claude.ai.
Create a custom channel group
A channel called "AI assistants" that matches those sources with a regular expression.
Track conversions from it
Compare engagement and conversion with other channels.
Add a UTM if you control the link
In places where you place links, such as your own profiles.
Some AI tools add utm_source=chatgpt.com to the links they send. Not all do, and some strip the referrer. Expect an undercount.
Keep non-human traffic out
- Self-identifying crawlers can be marked with Open Tracked's Bot Detection power-up in Header mode. Add a trigger exception so GA4 does not count them. See bot filtering.
- Do not block crawlers you want. If you want to appear in AI search results, let their search crawlers read your pages. Check your
robots.txtand your CDN. Blocking the crawler of a search tool means that tool cannot show you. - Agentic browsers cannot be reliably separated. Watch for patterns: very high engagement rates, odd hours, identical paths.
Raw events in. Clean events out.
Open Tracked collects each event on your own domain, removes the noise and sends what is left to Meta, Google Ads, TikTok.
What it means for tracking
Reports
Expect sessions from AI tools to grow, and check their quality, not only their volume.
Consent
Assistants that act for a person do not give consent for them. A visit by an agent should be treated cautiously.
Conversions
Server-side events help when the page runs in an unusual browser environment.
Content
Clear, factual pages that answer questions are more likely to be cited. See Digital marketing analytics.
A routine
- Monthly, look at AI referral sessions and conversions.
- Check your server logs for AI crawlers and what they request.
- Check that your
robots.txtand firewall allow the crawlers you want. - Check the share of sessions with no engagement.
Works with your stack
Grab a logo and throw it. These are the tools Open Tracked already talks to.
Frequently asked questions
How do AI browsers affect analytics?
They add sessions that may be bots or agents, and sometimes real visitors referred from AI answers. Measure by source and engagement.
How do I measure traffic from ChatGPT or Perplexity?
Look at source and medium in GA4 and create a channel group that matches the AI sources.
Can I block AI crawlers?
You can through robots.txt and firewall rules. If you want to appear in AI search answers, allow the relevant search crawlers.
Can I tell an agentic browser from a person?
Not reliably. Watch for unusual patterns.
Does Open Tracked help?
Bot Detection marks self-identifying bots so you can keep them out of GA4, and the logs show requests.
Do I need spreadsheets to read my results?
No. Once tracking is live, the Open Tracked analytics shows spend, revenue, ROAS and profit, and you can ask Ari, the built-in AI assistant, questions about your own numbers in plain words. Ari's allowance depends on your analytics plan.
Is Open Tracked different from Open Track or Open Space Track?
No. Open Tracked is the official name of our server-side tracking platform, and the website is opentracked.io. Many people search for it as Open Track, OpenTrack or Open Space Track. All of them lead to the same product: server-side Google Tag Manager hosting, first-party analytics and a CRM pipeline.
Fix your tracking at the source
Open Tracked hosts your server-side Tag Manager, keeps first-party cookies alive and shows what each ad is really worth.
The Open Tracked team
We build Open Tracked: server-side tracking hosting, analytics and the tools around them, for stores that spend on ads. Questions about this guide, or an idea for the next one?
Comments
Loading comments...
Guides · 5 min read
