Cross Domain Tracking

What it is

We use the cross domain tracking solution provided by Google for GA4 across many of our sites.

This solution works by having a list of domains added to the settings of a GA property. When a site uses that GA property (i.e. they are sending data to the measurement ID for that property), if the user clicks on a link the points to the domain specified in the settings, a set of query parameters will be added to the URL, containing information like the client ID.

The site that is linked to, must be using the same GA property, and provided that it is, the data from the query parameters in the URL will be used to set / override the GA cookie that contains the client ID.

Example

www.nature.com and submissions.springernature.com use the same GA property. The property has submissions.springernature.com listed as a domain for cross domain tracking. This means that we expect users to click links that go from www.nature.com to submissions.springernature.com.

When a link to submissions.springernature.com is clicked, the url is appended with some query parameters. When you land on submissions.springernature.com, a GA cookie is set containing the client id that was stored in the query parameter.

If you already had a GA cookie set on submissions.springernature.com, it will be overwritten.

What it is not

A common misunderstanding is that cross domain tracking means we can identify when a single user (or machine) is visiting many of our websites organically.

This is not true. Cross domain tracking only works with direct links from one site to another.

Example

A user searches on Google and clicks an article on www.nature.com. A user then searches on Google and clicks an article on www.biomedcentral.com.

We do not know that this is the same user. There is nothing connecting them.