How to set up a Cookieless Google Analytics website

February 1, 2021 • 5 minutes read • privacy, googleanalytics, cookieless

Note 1: Some might argue that if you are concerned about cookies you shouldn’t be using Google Analytics in the first place. That’s a valid point in which I agree, but on this post I want to leave that discussion out. The focus here is to explain how to implement Google Analytics without cookies while still keeping an acceptable unique user metrics.

Note 2: This post is using Universal Analytics from Google and it DOES NOT WORK with Google Analytics 4. I tried many times and I could not make it work with GA4, if you know how, please share!

It might be useful for us to start with why cookies are needed in Google Analytics. Without cookies the script is not be able to identify if the user visited that site before, which would cause all page views to be counted as unique.

To solve that, Google Analytics scripts creates a cookie on every user machine during the first visit. On a subsequent visit, the scripts can use that information to know if that’s a new or returning visitor.

Thankfully there are solutions that do not rely on Cookies.

First step: No more cookies

We need to tell GA to not store anything client side. We can do so via some custom params on the gtag function.

gtag('config', TRACKING_ID, { client_storage: 'none', anonymize_ip: true })

That’s it! We’re done. No more cookies.

Second step: Unique user tracking

The code above is enough to remove cookies, but it doesn’t solve the unique user tracking problem. As mentioned before, every visitor will be counted as a new user.

The same gtag function above supports an additional property called client_id. Instead of letting GA generate a client id and storing on a cookie, we can generate one ourselves and send them.

But after generating an ID, where do we store it? Most visitors onpublic websites are not authenticated, so it’s not possible to have a unique id per user.

One option is to use an algorithm that generates an ID based on public information from the user that does not change often and could potentially identify them. The following example is a function written in Node.js that generates a user ID based on their Browser, IP Address and Language. We can also add the current week to get a distinct ID every week even if all other properties have not changed.

A week is probably enough to get meaningful unique user visits like return users, but depending on your needs, it’s possible to use other values, like the Month or Day.

import crypto from 'crypto';

const currentWeekNumber = () => {
  const today = new Date();
  const firstDayOfYear = new Date(today.getFullYear(), 0, 1);
  const pastDaysOfYear = (today - firstDayOfYear) / 86400000;
  return Math.ceil((pastDaysOfYear + firstDayOfYear.getDay() + 1) / 7);

const md5 = (input) => {
  return crypto.createHash('md5').update(input).digest("hex");

export default (req, res) => {
  const clientIP = req.socket.remoteAddress
  const week = currentWeekNumber()
  const userAgent = req.headers['user-agent']
  const acceptLanguage = req.headers['accept-language']
  const host =
  const clientIDSource = `${clientIP}|${host}|${userAgent}|${acceptLanguage}|${week}`

  res.setHeader('Content-Type', 'text/javascript')

The function above generates a unique ID based on some information available to the server and write it to a global variable named user_id. It’s important to note that the function above MUST execute on a server, because JavaScript on the browser does not have access to its network IP address. Although the example above is running on Node.js, the same can be achieved in any language and framework.

Last step: Connecting the dots

Something quite easy to do (which doesn’t mess up with Content-Security-Policy) is to add a script tag right before GA tag:

<script src="/api/user"></script>
<script async src=""></script>

The code above assumes that a GET /api/user maps to that Node.js function above. The user_id variable can then be accessed later when calling gtag.

gtag('config', YOUR_TRACKING_ID, { client_storage: 'none', anonymize_ip: true, client_id: window.user_id })

Live Example

Here’s a live example using the method above which is also open source

That’s it!

Discuss on Twitter Edit on GitHub
Aptakube Aptakube for Kubernetes
Aptakube is a modern, lightweight and multi-cluster Kubernetes desktop client.
Available on Windows, macOS and Linux. Learn more →