AI assistants have quietly become a real acquisition channel for B2B sites. When someone asks ChatGPT, Perplexity, or Google’s AI Overviews to recommend a tool or explain a concept, the answer often links out — and those clicks are high-intent, late-funnel visitors who already trust the recommendation. The problem is that GA4 does almost nothing to surface them by default. This guide shows how to find AI traffic you are already getting, give it its own channel, and measure whether it actually converts.
Why AI traffic hides in GA4
GA4 classifies traffic into channel groups based on source and medium. AI assistants do not fit the built-in definitions, so they scatter into two unhelpful buckets. Some sessions land in Referral, mixed in with every other linking domain. Many more land in Direct, because the AI app or in-app browser strips the referrer before the click reaches your site — and GA4 treats any session with no referrer as direct. The result is that a genuinely new, growing channel looks like background noise. Before you can grow AI traffic, you have to make it visible.
Step 1: Find the AI referrals you already have
Start in Reports, then Acquisition, then Traffic acquisition, and switch the dimension to Session source / medium. Add a filter or search for the hostnames AI tools use: chatgpt.com and chat.openai.com for ChatGPT, perplexity.ai for Perplexity, gemini.google.com for Gemini, and copilot.microsoft.com for Microsoft Copilot. Anything that appears as one of these with a referral medium is confirmed AI traffic. Do not confuse this with /glossary/referral-spam, which is fabricated traffic from domains that never actually sent a visitor — these are real sessions from real assistants.
Step 2: Give AI traffic its own channel group
A one-off filter is not enough; you want AI traffic isolated everywhere, permanently. In Admin, open Data settings and create a custom channel group (or edit a copy of the default). Add a new channel called AI Assistants and define its rule as Source matches a regex containing chatgpt|openai|perplexity|gemini|copilot. Order it above Referral and Organic Search so those sessions are claimed by the AI channel first. From then on, every report that uses the default channel grouping can be re-pointed at your custom group, and AI traffic shows up as a first-class line item instead of vanishing into Direct and Referral.
Step 3: Measure whether AI traffic converts
Volume is the vanity metric here; conversion is the real question. Once AI traffic has its own channel, compare its /glossary/conversion-rate against organic search and direct. In practice, AI-assistant referrals often convert better than cold organic because the visitor arrives pre-qualified by the assistant’s recommendation — but you cannot claim that without the data. Add your key events (demo requests, trial starts, signups) to a comparison and segment by the AI Assistants channel. If the channel is small but converts at two or three times your site average, it deserves disproportionate attention. Use /features/ai-analytics-assistant for ad-hoc channel comparisons, and /features/reports for weekly workspace digests when email delivery is enabled.
Step 4: Watch the trend, not the snapshot
AI traffic is growing month over month for most B2B sites, and the interesting signal is the slope, not a single number. Because referrer stripping pushes some AI clicks into Direct, also watch for unexplained jumps in Direct traffic to deep content pages — a spike in direct visits to a niche blog post is frequently AI traffic in disguise. Set up an alert so a meaningful change in the AI channel reaches you the week it happens rather than the quarter you finally check. Automated /features/automated-alerts turn this from a manual audit into a standing watch.
What good looks like
A healthy setup has three properties: AI assistants appear as a named channel in your reporting, you know that channel’s conversion rate relative to organic and direct, and you get notified when it shifts. Teams that reach this point can make the case to invest in generative engine optimization — structuring content so assistants cite it more often — because they can finally prove the channel drives pipeline. If you are weighing whether your current analytics stack can answer questions like this without a data analyst, the /compare comparisons walk through how plain-English analytics differs from dashboards you have to build by hand.
The takeaway is simple: AI traffic is already arriving, GA4 just is not showing it to you. Spend twenty minutes isolating it into its own channel, confirm how it converts, and put an alert on the trend. You will know within a month whether ChatGPT and its peers are a rounding error or your next best acquisition channel.