Implementation of server-side measurement

Analysts across the globe agree that server-side measurement technology (or Server-side Tagging) is key to the future of analytics. And so do we. However, for most companies and digital projects, it’s a challenging technology step that often requires setting up and maintaining a cloud infrastructure that you probably have not been using so far. In that case, professionals like us are here to help you implement server-side measuring from scratch.

Server-side measurement concept

The concept of server-side measurement (as opposed to the client-side measurement that you use nowadays) is nothing new. In principle, it is an attempt to ease the user’s browser and to solve most of the measurement requests on the server. 

Today, it is common that most analytics and marketing measurement on the website is handled by embedding pixels from advertising/marketing and analytics vendors into your page and they do the work for you. Often you don’t even know exactly what the pixels are doing, what data they are collecting, and you don’t have much control over them. You see the numbers in the tools and somehow it works. 

In a simplified way, what happens in the user’s browser at such a moment is that when the page loads, a number of domains of the providers of each tool are queried and the browser is enormously loaded, the page load time is increased and a lot of JavaScript code is processed.

This also implies a lot of risks and limitations that you are exposing your digital project and users to.

Speed and 100% data control

With a complete server-side measurement implementation, this number of requests is greatly reduced. Instead of data going from the browser to dozens of different places, it all goes to your server. Validation and cleaning take place there, then everything is sent to the vendors’ servers (ad networks, etc.). So you have 100% control over what data is sent and where it goes.

Implementation via Google Tag Manager Server-Side container

In the past, you have been forced to use your own solutions and your own servers, and set everything up by yourself. Now we can use a ready-made solution, Google Tag Manager Server Side Container, which will help with the deployment of the infrastructure as well as with the actual setup, debugging, and long-term maintenance. 

Primarily, the server-side GTM is hosted on Google Cloud App Engine, but we can set it up in Cloud Run (for testing purposes), or you can host it in your own cloud (or another cloud, e.g. AWS, or private cloud) using Docker container (which ensures portability of the whole server-side measurement technology from Google – it’s a kind of cloud flash drive).

Who is server-side measurement suitable for?

Google Tag Manager server-side measurement is especially suitable for:

  • Large digital projects that care about the protection and security of their customers’ data,
  • Projects where complete control over first-party data collection (and ownership of processes and infrastructure) from collection to visualization matters (branches like banks, insurance companies, healthcare, major e-commerce participants, etc.),
  • Projects that want to take advantage of all the resources to reduce frontend JavaScript on the browser side, while ensuring a reduction in the number of requests to third party libraries and endpoints,
  • Projects that want to be technologically in the lead of their industry and invest in new data collection technologies,
  • Projects that want to ease content security policy (CSP) site requirements and prohibit their projects from communicating directly with advertising vendor sites,
  • Projects that need to enrich the collected data with real-time data from internal databases (user orders, product margins, user ids across systems, etc.),
  • Projects that need to clean data at the moment of collection (removal of personal data, IP anonymization, removal of digital fingerprints, etc.),
  • Projects that want to reduce individual library conflicts on the frontend, prevent potential data layer abuse, or protect their measurement IDs and API keys

And also for:

Projects that, even with a properly implemented consent policy, lose data due to ad blockers (paradoxes like the user giving consent but using AdBlock or strict private browser settings).

However, the purpose of server-side measurement is not to circumvent national legislation on electronic communications or marketing or to circumvent ad blockers. On the contrary, when using Server Side Measurement, we as a company need to be much more transparent and state exactly what is happening with the data, because from the average user perspective  (and the controlling entity), the data will go to a place that is not transparent at all and it is not easy to verify what is happening with the data without a clear declaration. 

You need exactly the same legal basis for processing data (mostly personal data) via server-side measurements as for client-side measurements (common browser measurements). The main motivation for server-side measurement must therefore definitely not be to circumvent the above. Unfortunately, we as Archetix cannot participate in such a project.

Server-side measurement costs

Another aspect to consider is cost. Currently, in the Czech market, we are used to the idea that collecting data from your websites and applications is “free”. This is not the case with server-side measurement. From the first moment you start using this technology, you need to take into account the costs of infrastructure and its maintenance. These can start at units of hundreds of crowns per month for smaller projects, and you can normally expect costs in the lower thousands of crowns for projects with 10 million hits per month. 

However, these costs are marginal to the above benefits, often demanding one hour of a senior analyst’s time, of which dozens are spent on server-side measuring. If the above amounts are an issue for your project, server-side measurement is probably not yet a suitable technology for you and you need to focus on growing your marketing and especially analytics maturity.

The most common tasks we solve for clients via server-side measurement

Data stream consolidation

Ideally, all data goes from the browser to one place (and only once) and is then sent to individual marketing vendors (note: not all major vendors support server-side measurement in December 2021 yet, but it is expected that number of them will gradually increase), often using a unified data model based on Google Analytics 4 as a modern standard for collecting data from websites and applications.

Measurement in the First Party Context

The measurement is performed on a subdomain hosted by you (e.g. server.domain.com), which is the same as the primary domain of the project (domain.com). This allows, among other things, to create secure 1st party cookies and helps in ensuring the maximum number of measured interactions

Data cleaning

The ability to clean data in real-time at collection, saving resources, time and analysts’ mental energy needed to clean data retrospectively.

Server conversion measurement

For major advertising platforms, it is already possible to measure conversions using API (e.g. Facebook Conversion API, Google Ads conversion tracking, etc.)

User geolocation

A side-tracking server can be one way to determine the approximate location of users if it is relevant for your project and you have the legal basis to do so.

Real-time data enrichment

Completing customer data before storing it in the database, returning the margin back to the browser, calculating the current value of the customer, etc. are topics we often discuss with our customers.

Are you considering server-side measurement for your project? Do you need more information or an initial consultation? Don’t hesitate to get in touch and get a technological edge over your competitors today.

Scroll to Top