Comparison · Where the measurement runs
Measuring in the browser or through a server of your own? The foundation in the browser first. The server comes when ad budget and gap justify it.
Both routes measure the same conversions: orders, enquiries, bookings. They differ in who sends the report to Google and Meta and what happens when the browser blocks it. This comparison sorts the differences by coverage, cookie banner, cost and operation.
The tracking check shows within two working days which route fits you. Credited if you commission within 30 days.
The two routes
- In the browser: The visitor’s browser reports directly to Google and Meta
- Through your server: The report goes first to a server that belongs to you
How we compare
-
Both routes can be right
After the check we pick the simplest route that meets your need, not the most expensive one.
-
No percentage promises
How much a route brings back depends on your visitors. Anyone promising a fixed number does not know your data.
-
The foundation first
No route repairs wrong numbers. If amount or reference are wrong in the shop, they arrive wrong everywhere.
-
Technical, not legal
Whether a setup is permissible is for you and your privacy advisers. We build the technology.
In short
A server of your own is not a substitute for correct measurement, it is a different transport route. If amount, reference or cookie decision are wrong in the browser, the server delivers exactly those errors more reliably. That is why the foundation comes first and the server follows once ad budget, remaining gap and number of platforms justify the extra effort.
In detail
The differences in detail
| Point | In the browser | Through your server |
|---|---|---|
| Who reports | The visitor’s browser, directly to Google, Meta or others. | Your server, which checks the report and then forwards it to the platforms. |
| Ad blockers | Known addresses of foreign providers are often blocked. The report never arrives. | Runs through your own address and therefore rarely sits on the block lists. Not a free pass, but far more stable. |
| Cookie banner | The visitor’s decision controls the Google scripts directly in the browser. | The decision still has to be obtained in the browser and passed to the server. The server creates no permission of its own. |
| What goes out | Whatever the platform’s template provides. | The server can shorten details, make them unreadable or not forward them at all. |
| Double counting | Prevented through the order number in the browser. | Additionally through a shared reference between browser and server. |
| Entry cost | No extra hosting. The effort sits in setup and testing. | Server, own address, hosting and operation are added. |
| Running cost | Practically none. | Depends on hosting and the volume of reports. |
| Operation | Tag Manager versions, platform changes, theme and plugin updates. | Additionally server, certificates, reachability and monitoring. |
| Troubleshooting | Directly in the browser, with the usual tools. | In two stages: first the browser, then the server. More work, but cleanly loggable. |
| The typical first step | Put reports, amounts, cookie decision, ad account and references in order. | Move an already verified measurement to a server of your own. |
The decision
What speaks for which route.
A server of your own pays off when
- a noticeable ad budget depends on the quality of the reports
- the gap after the repair stays noticeable in the business
- several platforms are served, such as Google and Meta
- data should stay in Germany and you decide what goes out
- browser and server should prevent double counting together
- details should be shortened or made unreadable before forwarding
The browser is enough when
- · the measurement is being rebuilt or repaired right now
- · a moderate ad budget runs
- · one platform and one account are in use
- · the gap after the repair is small and explainable
- · nobody wants to run a server
- · it is about right numbers, not more numbers
Whether a specific setup is permissible depends on the data, the recipients and the legal basis. This is a technical assessment, not legal advice.
Where it goes wrong
The most common misconceptions
- A server makes the cookie banner unnecessary. Wrong: consent stays mandatory.
- A server repairs wrong amounts or reports. It does not, it only forwards them.
- A server bypasses every ad blocker. It reduces the losses, it does not remove them.
- A server replaces the browser. In practice both run in parallel, with a shared reference.
- A server is always more expensive. With a lot of ad budget the better numbers justify the operation, with little they rarely do.
Common questions
Can I start with the server without repairing the browser measurement?
Technically yes, sensibly rarely. The server takes over what was defined in the browser. Wrong reports, duplicate purchases or a missing order number travel along unchanged.
How much more does Google see through the server?
That depends on consent rate, ad blockers, browsers and platform and can only be quantified after a baseline measurement. Anyone promising a fixed percentage does not know your data.
Do I need both routes at once?
In most cases yes. Browser and server report in parallel, a shared reference prevents double counting. The browser stays the source for behaviour on the page, the server takes care of reliable delivery.
What does running a server of my own cost?
That depends on hosting. A server of your own at Hetzner is predictable by size, rented services start low and cloud billing grows with the volume of reports. The decision is made after the check, not before.
Read on
Where to go next.
Start the tracking check
Four required answers, the rest is optional. You get a reply by email within one business day, no phone call.
Next step
Which stage do you really need?
The tracking check shows how big your gap is and whether a server of your own pays off. With a fixed price.
The tracking check shows within two working days which route fits you. Credited if you commission within 30 days.