Full Answer

Ad blockers target client-side JavaScript by design, and analytics tags are among their primary targets. Statista reports 31.5% global ad blocker usage, which means roughly one in three visitors never fires your GA4 tag at all — no pageview, no event, no attribution. For AI referral traffic specifically, this creates a compounding blind spot.

The first layer of loss comes from the AI engines themselves. Google AI Mode attaches noreferrer to every outbound link, so the referrer header is stripped before the click reaches your site. ChatGPT's mobile app is inconsistent with how it passes attribution data. Perplexity passes a referrer but not reliably across all browsers. The second layer is the ad blocker: even when an AI engine does pass referrer data, a blocked analytics script means nobody records it.

Server-side tracking sidesteps both layers. Because events are captured at the server — before the browser decides what to block — the raw request headers, including referrer and user-agent, are available for attribution. A broader look at the scale of the AI data gap is covered in 88% of Small Businesses Have No AI Search Strategy. The practical takeaway: any analytics setup that depends entirely on client-side JavaScript is structurally incapable of measuring AI search traffic accurately.