"Cookieless" analytics doesn't mean the tracking script does nothing, it means it doesn't store a persistent identifier on the visitor's device to recognize them the next time they show up. No cookie, no local storage entry, no fingerprint built from screen size and fonts and timezone. Each page view is recorded on its own, with just enough signal to group it into a session.
How do you count "sessions" without an identifier, then?
This is the part that trips people up. A cookieless tool can still generate a short-lived, non-persistent identifier for the duration of a single visit, the difference is that identifier doesn't survive across visits, isn't shared across sites, and isn't tied to anything personal. It exists just long enough to know that three page views ten minutes apart were one visitor browsing around, not three separate people.
What you keep
- Accurate visitor and session counts
- Page-by-page navigation within a visit
- Referrer, campaign, device, browser, and location (usually derived from IP address without storing the IP itself)
- Conversions and goal completions
What you lose
The main thing: reliably recognizing the same person across separate visits days or weeks apart. If your business depends on precise "new vs. returning" tracking over long stretches of time, or stitching together a single visitor's behavior across many sessions, a cookieless tool will estimate that less precisely than a tool that persistently tags visitors. For most sites, the trade-off is worth it: accurate day-to-day, week-to-week numbers, without the compliance overhead a persistent identifier brings.
Why it usually skips the cookie banner requirement
Consent requirements are triggered by storing non-essential data on a visitor's device, not by "having analytics" as a category. No persistent storage, no consent needed for the analytics itself, in most cases. We cover the specifics in do you need a cookie banner for analytics?