AI Referral Traffic

AI referral traffic is the slice of your website visits that started inside an AI answer. Someone asked ChatGPT for a family dentist in Fresno, the answer named your practice with a link, and they clicked. That click is an AI referral. It's tracked the same way any referral is, by the referrer header or a tracking parameter on the URL, and it shows up in analytics as its own source when the tool knows how to classify it.

The totals are small but rising quickly. According to Similarweb-family data, AI platforms sent about 0.32% of all website traffic in 2026, up from 0.24% in 2025 and just 0.02% in 2024. Within that AI slice, ChatGPT dominates at 74.78%, followed by Gemini at 11.56%, Perplexity at 7.23%, Copilot at 3.51%, and Claude at 2.62%. If you're going to watch one engine's referrals first, ChatGPT is the one.

Measuring it is harder than it should be. Some engines strip the referrer, so a visit from an AI answer can land in your analytics as direct traffic. ChatGPT appends a utm_source=chatgpt.com parameter to many outbound links, which helps, but not every engine does the same. A generic analytics setup will undercount AI referrals unless someone configures it to recognize each engine's referrer domain and parameter pattern.

For a small business, the number of AI-referred visits matters less than what those visitors do. A visitor who arrives from an AI answer has already read a recommendation naming you. They tend to land on a specific page, the one that got cited, and they're usually further along than a visitor who clicked a generic search result. A weak or irrelevant landing page wastes a warm lead.

Every Tibly site includes analytics that attribute visits from ChatGPT, Perplexity, Gemini, Copilot, Claude, and Meta AI to their engine, classified server-side with a UTM fallback when the referrer is stripped. Paired with the visibility scans, you can see both halves of the loop: how often an engine names you, and whether being named turns into people on your pages.