How visible is your brand in ChatGPT, Gemini and co.? Free GEO potential analysis for AI search until the end of December Request your analysis

Tracking guide

What is server-side tracking and when is it worth it?

Browsers shorten cookies, ad blockers stop scripts, and part of your enquiries and purchases never reaches Google Ads or Meta. Server-side tracking sends the data through a server of your own instead of straight from the browser. We explain how it works, what it costs, what data protection law says about it and at what point the effort pays off for your company.

Malek Arab
Malek Arab Managing Director, Pageworkers
Updated on 13 min read
Contents10 sections
  1. 01What it means
  2. 02Why data goes missing
  3. 03Benefits
  4. 04Three routes compared
  5. 05Costs
  6. 06When it is worth it
  7. 07GDPR and consent
  8. 08Setup in 7 steps
  9. 09Typical mistakes
  10. 10Conclusion and first steps

What is server-side tracking?

Server-side tracking means that the measurement data of your website first goes to a server you control, and only from there to services such as Google Analytics 4, Google Ads or Meta. With classic tracking in the browser, every platform loads its own script and collects the data directly from the visitor.

The server usually runs under a subdomain of your website, for example data.your-company.com. To the browser, the measurement then looks like part of your own website. On the server you define which details go to which platform and which never leave the building.

Paper website card with a lime ribbon leading to a dark blue block as the own server, from where it branches out to three paper boxes as recipients of the data

In day-to-day terms this mainly changes the following:

  • The browser loads fewer third-party scripts because one data stream to your server is enough
  • Cookies are set by your own server and last longer in Safari and Firefox
  • You decide in one place which data Google, Meta or LinkedIn receive
  • Events from the shop, CRM or booking system can be reported straight from the server, without any browser

You will also come across the term server-side tagging. Server-side tagging is Google’s name for the implementation with the server-side Google Tag Manager. Server-side tracking is the umbrella term for every way of sending data through a server.

Why does classic browser tracking lose more and more data?

Because browsers and visitors have been restricting tracking bit by bit for years. Three things come together: Safari and Firefox shorten cookie lifetimes, ad blockers and privacy extensions block known tracking addresses, and without consent in the cookie banner nothing may be measured anyway.

  1. February 2019

    Safari cuts cookies to 7 days

    With ITP 2.1 Apple limits all cookies set by a script in the browser to 7 days. Anyone who returns after 8 days counts as a new visitor.

  2. April 2019

    Only 24 hours after an ad click

    ITP 2.2 shortens the lifetime to 24 hours when the visitor arrives through a link with a click identifier from a domain classified as a tracker. This hits ad clicks in particular.

  3. November 2020

    Cloaked subdomains get noticed

    WebKit detects subdomains that point to a third-party tracking service via CNAME and limits their cookies to 7 days as well.

  4. March 2023

    Safari checks the server’s IP address

    Since Safari 16.4, according to Stape and Simo Ahava, 7 days also apply to cookies set by a server when its IP address does not match the website. A tracking server at any provider is no longer enough for long lifetimes.

  5. March 2024

    Consent Mode v2 becomes mandatory

    For visitors from the European Economic Area Google requires the consent signals ad_user_data and ad_personalization. Without them Google Ads no longer builds audiences.

Sources: WebKit Blog (21 Feb 2019, 24 Apr 2019, 12 Nov 2020), Stape and Simo Ahava on Safari 16.4, Google Ads Help on Consent Mode.

You can see the result in your numbers: the shop or the CRM shows more orders and enquiries than Google Ads and Meta report. The campaigns are then optimised on too little and on the wrong data. The gap is especially large for audiences with many iPhones and for technical audiences, who use ad blockers more often.

Clay balls lie stopped in front of a paper wall while lime balls roll through a dark blue tunnel under the wall into a tray, as an image of data arriving through your own server

What are the benefits of server-side tracking?

The most important benefit is more complete conversion data for your advertising accounts. The other benefits are real, but smaller than many providers promise.

  1. 1

    More conversions arrive

    The data stream runs through your own domain and is blocked less often. Bidding strategies in Google Ads and Meta learn from more real conversions and steer the budget better.

  2. 2

    Cookies last longer

    A cookie set by your own server survives more than 7 days in Safari if the server matches the IP address of your website. Returning visitors and long decision paths in B2B are attributed correctly again.

  3. 3

    Control over the data

    On the server you remove or shorten details such as the IP address before anything goes to Google or Meta. In the browser every third-party script has access to everything on the page.

  4. 4

    Data from your own systems

    The server can report events that never happen in the browser: the paid invoice, the qualified lead from the CRM, the confirmed appointment. That way you optimise for revenue instead of submitted forms.

  5. 5

    Slightly less load in the browser

    Fewer third-party scripts mean slightly shorter loading times. The effect is measurable but small, and on its own no reason to switch.

How much more arrives depends on your audience and the route you choose. Reliable figures only come from your own account, with old and new tracking running in parallel. Be wary of blanket promises such as “30% more conversions”.

Which routes are there: Google tag gateway, server GTM or server-to-server?

There are three levels, from a quick start to connecting your systems directly. They build on each other, and for many companies the first or second is enough.

Google tag gateway

Effort: hours, no server costs

  • Google tags load through your own domain
  • Set up through Cloudflare in a few clicks and free of charge
  • No server of your own, no maintenance
  • Only covers Google, Meta and LinkedIn stay in the browser
  • No filtering or enriching of data

A good first step if you mainly use Google Ads.

The usual route

Server-side Google Tag Manager

Effort: days to weeks, ongoing server costs

  • One data stream from the browser, distributed to GA4, Google Ads, Meta, LinkedIn and others
  • Own cookies with a longer lifetime
  • Data can be filtered before it is passed on
  • The trigger still comes from the browser and can be blocked
  • Server, updates and monitoring need someone responsible

Fits most companies with several advertising platforms.

Server-to-server

Effort: development per system

  • Shop, CRM or booking system report events straight to the platform
  • No browser involved, nothing can be blocked
  • Ideal for purchases, confirmed appointments and qualified leads
  • Needs programming and upkeep per interface
  • The link to the ad has to be passed along completely

The most accurate solution for the few events that really count.

The third route includes the platforms’ interfaces: Meta’s Conversions API, enhanced conversions and offline import in Google Ads, the GA4 Measurement Protocol and LinkedIn’s Conversions API. Often both run side by side. The browser reports the event, the server reports it again, and the platform merges the two.

So that nothing is counted twice, every event needs a unique identifier. Meta matches browser and server through event_id and event_name and discards the second event if both arrive within 48 hours. If the identifier is missing in the browser, Meta counts every purchase twice.

What does server-side tracking cost?

For small and medium-sized websites the running costs are usually between €20 and €150 a month, plus the one-off setup. Google Tag Manager itself remains free, what costs money is the server the container runs on.

  1. 1

    Server at Google Cloud

    Google quotes around 45 US dollars a month for a Cloud Run instance with 1 vCPU and 0.5 GB of memory and recommends at least 2 instances so that no data is lost during an outage. That makes around 90 US dollars a month as a minimum, more with a lot of traffic.

  2. 2

    Managed hosting

    Providers such as Stape run the server for you. Stape is free up to 10,000 requests a month, paid plans start at around 20 US dollars a month. For most company websites this is the cheapest way in.

  3. 3

    Google tag gateway

    Free through Cloudflare. All you need is your domain at Cloudflare.

  4. 4

    Setup

    Typical market prices are a one-off €750 to €2,000 for a setup with GA4, Google Ads and Meta, depending on how many events and platforms are connected. Direct interfaces to the shop or CRM come on top.

  5. 5

    Upkeep

    Platforms change their interfaces, consent tools get updates, new events are added. Expect a few hours per quarter for checks and adjustments.

Clay scale with a stack of coins in one pan and lime balls in the other, as an image of weighing the cost of the server against the gain in data

Sample calculation: first year with a managed server and 3 platforms

Setup with GA4, Google Ads and Meta, one-off
€1,250
Managed hosting, 12 months at €20 each
€240
Checks and adjustments, 4 quarters
€400
Total in the first year
€1,890

Made-up, rounded example and not a real quote. From the second year on, around €640 remain in this example.

Whether this pays off is decided by your advertising budget. If you spend €5,000 a month on ads and better data makes you just 5% more efficient, you have more than recovered the running costs. With an advertising budget of €500 a month the sums rarely add up.

Who is server-side tracking worth it for, and who not?

It is worth it if you steer advertising budget by conversions and your platforms visibly see fewer conversions than your own system. For a website without ads that only wants to count visitors, it is usually too much effort.

  • You regularly spend money on Google Ads, Meta or LinkedIn and use automated bidding strategies
  • There is a clear gap between orders or enquiries in your own system and the conversions in the advertising accounts
  • Your customers decide over weeks, for example in B2B, and return several times before they enquire
  • The actual conversion happens outside the browser, for example in the CRM, on the phone or in the booking system
  • You want to control which data goes to US platforms

In these cases it adds little

  • Hardly any advertising budget and no campaigns that optimise for conversions
  • The figures from the advertising account and your own system are already close together
  • The basics are still missing: no legally sound cookie banner, no Consent Mode, no defined conversions
  • Nobody in-house or at the agency feels responsible for the server

The order matters. First come a legally sound banner, Consent Mode v2 and a few clearly defined conversions. Then enhanced conversions in Google Ads, which work without a server of your own. Only then the server. If you reverse the order, you measure the same mistake at a higher price.

Is server-side tracking GDPR-compliant and does it need consent?

Yes, it can be run in a GDPR-compliant way, and yes, consent remains necessary. The server changes the route the data takes, it does not change the legal situation. Anyone who sets or reads cookies that are not strictly necessary for the website needs the visitor’s prior consent under section 25 of the German TDDDG. This applies regardless of whether a script in the browser or your own server sets the cookie.

Server-side tracking is therefore not a way around the cookie banner. Providers advertising “tracking without consent” usually mean a heavily reduced measurement without cookies and without personal data. Whether that is permissible in a specific case is something your data protection officer should check. For advertising purposes such as remarketing and conversion measurement in Google Ads and Meta, there is no way around consent.

The server still has advantages for data protection if you configure it properly:

  • The consent from the banner is passed on to the server, and it only sends what was allowed
  • IP addresses and other details can be shortened or removed before data goes to third parties
  • Third-party scripts can no longer read form contents because they are not loaded in the first place
  • You choose the location of the server, for example a data centre in the EU

The privacy policy should name the server, its operator as a processor and the platforms it passes data on to. You conclude a data processing agreement with the hosting provider. This section does not replace legal advice.

How is server-side tracking set up?

In 7 steps, from taking stock to switching off the old scripts. A simple setup with Google Tag Manager is ready in 2 to 4 weeks, with direct interfaces to the shop or CRM it takes longer.

  1. 1

    Taking stock

    Which events are measured today, which platforms receive them, and how large is the gap to your own system. This is often where it turns out that conversions are counted twice or not at all.

  2. 2

    Check consent

    The cookie banner must support Consent Mode v2 and pass on its signals reliably. Without this step everything else is worthless.

  3. 3

    Set up the server

    Create the container in Google Tag Manager and start it at Google Cloud or a managed provider, ideally in an EU region.

  4. 4

    Set up your own subdomain

    The server gets a subdomain of your website via DNS. For long cookie lifetimes in Safari it should be reachable through the same infrastructure as the website, for example your CDN.

  5. 5

    Move the events

    The browser sends one data stream to the server. There, tags distribute the events to GA4, Google Ads, Meta and other platforms, each with a unique identifier against double counting.

  6. 6

    Run in parallel and compare

    Old and new tracking run side by side for 2 to 4 weeks. Only when the figures are plausible do you switch over.

  7. 7

    Remove old scripts and monitor

    The browser pixels go, as far as they have been replaced. An alert reports when the server fails or conversions suddenly stop arriving.

From our practice

How we measure appointment bookings on pageworkers.com

On our own website the most important conversion is the booked appointment. It was never reliably measurable in the browser, because the booking happens in an embedded calendar from another provider and visitors with tracking protection do not show up there at all.

So we moved the event to the server. The booking system reports every confirmed booking directly to our server, which passes it on to GA4 through the Measurement Protocol. From there it arrives in Google Ads as a conversion. The event in the browser keeps running, but now only counts for attribution to the campaign.

All other events still run through the normal Tag Manager on our site. For the one conversion the evaluation depends on, the direct route was the better solution than a complete rebuild.

1conversion that comes straight from the server on our site
0bookings swallowed by an ad blocker
Since 09/2026in operation on pageworkers.com
Malek Arab Malek ArabManaging Director, Pageworkers

Want to know how many conversions your campaigns are missing?

We compare your advertising account with your real enquiries and tell you whether server-side tracking pays off for you.

View Google Ads agency

Which mistakes happen most often with server-side tracking?

Most mistakes arise where browser, server and consent interact. They often only show up after weeks, when the figures in the advertising account look too good or too bad.

The mistakes we see most often

  • Conversions are counted twice because browser and server send the same event without a shared identifier
  • The server also sends without consent because the signal from the banner does not arrive
  • The subdomain points to a third-party server, and Safari still cuts the cookies to 7 days
  • The ad’s click identifier gets lost on the way, and the conversion cannot be attributed to any campaign
  • After the switch nobody checks any more whether the server is running
  • Suddenly far more conversions are celebrated as a success although they are duplicates

What is the best way to start with server-side tracking?

With a comparison of your figures. Put the conversions from Google Ads and Meta next to the real enquiries or orders of one month. If the gap is small, first get banner, Consent Mode and enhanced conversions in order. If it is large, start with the Google tag gateway or with a single conversion through the server.

Server-side tracking is not a project you set up once and forget. Browsers and platforms keep changing their rules, and those who compare regularly notice early when something no longer adds up. If you want campaigns and measurement from one source, talk to our Google Ads agency.

FAQ

Frequently asked questions about server-side tracking

What is the difference between server-side tracking and server-side tagging?

Server-side tagging is Google’s name for the implementation with the server-side Google Tag Manager. Server-side tracking is the umbrella term and also includes direct interfaces such as Meta’s Conversions API or the GA4 Measurement Protocol.

Does server-side tracking bypass ad blockers?

Partly. The data stream runs through your own domain and is blocked less often. As long as the trigger comes from the browser, strict blockers can still stop it. Only events that your shop or CRM reports straight from the server are independent of that. The visitor’s consent applies in every case.

Does the Google tag gateway replace the server-side Google Tag Manager?

No. The Google tag gateway only delivers the Google tags through your own domain. It does not filter data and does not serve other platforms such as Meta or LinkedIn. The two can be combined.

Does server-side tracking work with WordPress, WooCommerce and Shopify?

Yes. For WordPress and WooCommerce there are plugins that send the data stream to the server-side Google Tag Manager, Shopify offers its own interfaces and apps. On custom-built websites the data stream is built straight into the code.

What happens if the tracking server goes down?

Then measurement data is lost for that time, but your website keeps running normally. That is why Google recommends at least 2 instances. In addition, an alert should report when the server does not respond or conversions stop arriving.

Why does Google Ads show more conversions after the switch?

In the best case because conversions now arrive that were blocked before. Check against the real orders or enquiries anyway. If Google Ads is above that, events are being counted twice, usually because browser and server send without a shared identifier.

Do I need server-side tracking if I only use Google Analytics?

Usually not. Without advertising campaigns that optimise for conversions, the economic lever is missing. A correctly configured GA4 with Consent Mode is generally enough for analysing the website.

Authenticity
07/15/2026
100%
Recommended on ProvenExpert.com
Excellent
4.93 / 5.00
39
Reviews on ProvenExpert.com
128
Reviews from 3 other sources
View profile on ProvenExpert.com 
Your contacts
Pageworkers team
Available now