how i group traffic sources without cookies
every dashboard has a "top traffic sources" chart. almost none of them will tell you how it got made. i'm om yaduvanshi. i'm 18, an applied ai developer building cloudline, a cookieless google analytics alternative. one
every dashboard has a "top traffic sources" chart. almost none of them will tell you how it got made.
i'm om yaduvanshi. i'm 18, an applied ai developer building cloudline, a cookieless google analytics alternative. one script tag, under 2 KB, no cookies, no stored ips. day 9 of my 90-day build in public, and part of my series on tracking the web without tracking people. so far: referrers, scroll depth, funnels, button clicks, engaged time, live visitors.
a referrer is a url, not an answer
a pageview in cloudline stores four things: page, referrer, screen size, time. the referrer is the interesting one for this piece. it's a raw string. https://www.google.com/search?q=cookieless+analytics. https://t.co/abc123. or nothing at all, when someone typed the url directly or the browser stripped it.
a raw string is not a traffic source. so the dashboard groups them.
the four buckets
cloudline sorts referrers into four buckets: search, social, newsletters, ai assistants. everything else is websites. no referrer at all reads as direct.
a google search lands in search. a link from x or reddit lands in social. a substack link lands in newsletters. a chatgpt or claude referrer lands in ai assistants, in its own bucket instead of hiding inside websites.
that last bucket is the one i'd have laughed at three years ago. people find websites through chatbots now. it's a traffic source worth naming.
the honest limit
referrers lie. day 2 of this series was about exactly that. in-app browsers strip them. some apps never send them. a chunk of real traffic will always read as direct when it isn't.
the only way to fix the gaps is identity tracking: stitching visits together across sessions. that's the thing i refuse to store. no cookies, no ip storage, no fingerprints. so the buckets stay approximate, and the dashboard doesn't pretend otherwise.
if you need exact attribution per visitor, this isn't your tool. if you want to know roughly where your readers come from without a consent banner, four buckets get you there.
one line, under a minute
if you haven't installed yet:
<script src="https://trycloudline.space/tracker.js" data-site-id="YOUR-SITE-ID"></script>
takes under a minute. removing it is deleting the line.
cloudline is free for early users while i keep building it. cookieless analytics in one line of code.
— om yaduvanshi
Originally published by Dev.to WebDev. Aggregated on AIWithGhost for educational purposes — full credit and traffic to the original publisher.