At a glance
- With server-side tracking, the browser sends measurement data to a server of your own under your own domain; only that server forwards it to GA4, Google Ads or Meta.
- Benefits: less data loss, control over which data is passed on, fewer scripts in the browser.
- It does not replace consent. What may not be collected without consent may not be collected through your own server either.
- It pays off above all with a substantial ad budget and when closed deals are meant to flow back to the ad platforms as conversions.
How server-side tracking works
With classic tracking, the browser loads scripts from Google, Meta or LinkedIn and sends measurement data directly to their servers. With server-side tracking, every measurement first goes to a server of the company, usually under a subdomain of its own such as data.example.co.uk. That is where it is decided which data goes on to which service.
A common setup is Google Tag Manager in its server version, the so-called server container. It runs in a cloud environment, with a specialised provider or on a server of your own, ideally located in the EU.
What it improves
- Less data loss: ad blockers mainly block requests to known tracking addresses. Requests to your own domain are affected less often.
- Longer recognition: Safari limits cookies set via JavaScript to seven days. A server of your own can set cookies with a longer lifetime; Safari limits those as well, though, if the server visibly runs with a third-party provider.
- Control: before forwarding, data can be trimmed or removed, such as the IP address or parameters containing personal details.
- Faster pages: fewer scripts in the browser take load off smartphones in particular.
- Feedback to ad platforms: conversions can be passed on reliably, including those that only happen in the CRM.
What it does not do
Server-side tracking is no way around consent. Under Section 25 of Germany’s TDDDG, which implements the EU ePrivacy rules, any access to information on the end device, such as setting or reading a cookie, requires consent unless it is strictly necessary. The GDPR applies on top for the processing of personal data. Your own server changes nothing about that; it only makes sure that permitted measurements arrive more completely.
Nor does it deliver complete data. Visitors who decline are not tracked; with Google, Consent Mode can partly estimate such gaps through modelling. Then there are the costs of running the server and the maintenance effort when platforms change their interfaces.
GA4, Google Ads and Meta
- Google Analytics 4
- Events go to GA4 through the server container; Consent Mode controls what is sent when visitors decline
- Google Ads
- Conversions through the server, plus enhanced conversions (hashed contact data after consent) and offline conversions from the CRM
- Meta
- Conversions API alongside the pixel; duplicate events are recognised through a shared event ID
- Conversions API for forms and events from the CRM
The biggest difference usually comes from sales feedback: when Google and Meta learn which leads became deals, they steer delivery towards deals instead of clicks. How this fits together is shown in the Campaigns module.
Consent Mode v2 in brief
Since March 2024, Google has required Consent Mode v2 for users in the European Economic Area so that remarketing, personalised advertising and parts of measurement keep working. Version 2 transmits two additional signals: whether data may be sent to Google for advertising purposes (ad_user_data) and whether personalised advertising is allowed (ad_personalization).
In basic mode, Google tags only load after consent. In advanced mode, they send cookieless signals when visitors decline, from which Google estimates missing conversions.
When it pays off
- An ad budget at which missing conversions lead to noticeably worse decisions.
- Many visitors using Safari on iPhone and Mac.
- Conversions that only happen in the CRM and are meant to flow back to the ad platforms.
- The aim of controlling exactly which data goes to which services.
For small websites with little advertising, a clean cookie consent with Consent Mode v2 is usually enough. The server container is then a second step.
Frequently asked questions
Is server-side tracking allowed without a cookie banner?
Only for measurements that do not require consent. Whether that is the case depends on whether the end device is accessed and which data is processed. The server alone does not make a measurement consent-free.
Does it work with Shopify, Shopware or WordPress?
Yes. These systems send events to the server container, either through Tag Manager in the browser or through their own interfaces and extensions.