Table of content
On 15 April 2026, Meta told its advertisers about AI enrichment of their data, switched on automatically after thirty days unless they explicitly opted out.
Three months earlier, Shopify had moved its App Pixels to “optimised” mode by default.
And in September 2025, Safari 26 added anti-fingerprinting protection that degrades part of what the browser still transmits.
Three decisions. None of them yours. All of them changed what your company measures.
The common thread: your measurement is not yours
As long as your data collection lives in the visitor's browser, it depends on three parties you do not control: the browser vendor, the advertising platform, and your CMS. Any of them can change a setting, one day, without asking you.
And you do not find out when the change happens. You find out three months later, when a curve drops in a report and nobody can say why.
That is the real issue. Not the law, not the GDPR, not the revised FADP. The question is who decides what you measure.
What changes when the server is yours
Server-side tracking means routing collection through your own server before it goes anywhere else. Technically, it is a container you host, on a subdomain of your own, and which you can place in Switzerland.
In practice, it moves three decisions from the browser to your side.
What leaves. You choose the fields sent to Meta, Google or LinkedIn, and you strip out what should not go out.
How it leaves. Pseudonymisation, IP truncation, event deduplication: the rules are written once, on your side, and applied to every channel.
Where it is hosted. Your server can stay on Swiss soil, which settles a large part of the data transfer question in a single decision.
When Safari changes its rules or a platform switches an option on by default, your server remains the mandatory point of passage. You observe the change instead of absorbing it.
Competitive and compliant, not one against the other
The two are often set against each other. American companies are said to move without constraints while European ones fill in registers. It is a false debate, and an expensive one.
One figure shows why. In Western Europe, the headline acceptance rate hovers around 75%, but the real opt-in rate falls to 56%, and one visitor in four makes no choice at all. Nearly half of your audience therefore sits in a zone where the browser will never hand you anything.
A client-side setup is blind to that zone. A well-built server-side setup recovers a significant share of it, through conversion modelling and enrichment of consented signals, without ever going beyond what consent allows.
The same piece of work therefore produces both outcomes: better fed campaigns, and compliance that holds. These are not two projects. They are one.
The Swiss argument, in numbers
The 2026 Digital Barometer offers a data point few companies use: 46% of Swiss residents name technological dependence as the country's main weakness, and 83% would trust a service more if it carried a guarantee of Swiss origin.
Hosting your measurement in Switzerland is therefore not only a legal precaution. It is a commercial argument your clients already understand.
You cannot stop Meta, Google or Apple from changing their rules. You can decide that those rules apply after your server, not before it.
Where to start
Three questions are enough to know where you stand.
Which domain does your data travel through today? If the answer is a third-party domain, your measurement is not yours.
Who can change your collection without warning you? Make the list. It is longer than you expect.
What do you do with visitors who never consented? If the answer is nothing, you are paying for acquisition you will never be able to attribute.
Server-side is not one more technical optimisation. It is the moment you take back control of what your company knows about itself.
