English
Deutsch
PolskiTag Management Systems: From Hardcoding to Server-Side, and How to Choose One
Contents
Anyone working in digital marketing or data analytics today probably uses a Tag Management System (TMS) daily. But it wasn’t always this easy. To truly understand the power of modern platforms like Google Tag Manager, Tealium, or Jentis, we have to look back at how data collection began.
The Early Days: The Hardcoding Era
In the early 2000s, whenever a marketing team wanted to add a tracking pixel (like Google Analytics, Facebook, or an affiliate script) to a website, they had to write a ticket to the IT department. Developers had to manually “hardcode” these JavaScript snippets directly into the website’s source code.
This process was slow, prone to errors, and heavily dependent on IT release cycles. As digital marketing exploded, websites became bloated with dozens of third-party scripts, drastically slowing down page load times and creating a data governance nightmare.
The Evolution: The Birth of TMS
Between 2008 and 2012, the first Tag Management Systems emerged (Tealium was founded in 2008, Google Tag Manager launched in 2012). A TMS provided a single “container” code placed on the website once. From then on, marketers could deploy, modify, and manage tags through a user-friendly web interface without touching the site’s code.
Today, the industry is undergoing another massive shift: Server-Side Tracking (SST). Due to ad blockers, privacy laws (GDPR), and Intelligent Tracking Prevention (ITP) in browsers, client-side tracking (in the user’s browser) is losing data. TMS platforms are evolving to route data through a first-party server before sending it to vendors.
Let’s break down the biggest players on the market, their pros and cons, and their Server-Side capabilities.
The Big Players in Tag Management
1. Google Tag Manager (GTM)
The undisputed king of market share, primarily because its standard version is completely free and seamlessly integrates with the Google ecosystem.
- PROs: Free, massive community, endless tutorials, incredibly user-friendly for beginners.
- CONs: The free version lacks enterprise-level SLAs and advanced data governance. Data goes directly into the Google ecosystem, which raises privacy concerns for some enterprises.
- Server-Side Tracking: Available (Server-Side GTM). Setup is relatively easy on Google Cloud, but it requires technical knowledge to map client-side data to the server container.
- When to choose: When the marketing stack relies heavily on Google (GA4, Google Ads) and the requirement is a fast, cost-effective solution.
2. Tealium (Tealium iQ)
A pioneer and enterprise heavyweight. Tealium is designed for massive corporations with complex, multi-brand data architectures.
- PROs: Incredible tag marketplace (over 1000 native integrations), stellar data governance, privacy controls, and a highly stable environment.
- CONs: Very expensive. The interface is highly technical, and the learning curve is steep.
- Server-Side Tracking: Excellent. Tealium EventStream handles server-side data routing flawlessly, but it requires architecture planning and developer skills to deploy.
- When to choose: For a large enterprise with a huge budget, complex data routing needs, and a requirement to send data to hundreds of non-Google vendors.
3. Adobe Experience Platform (Tags / formerly Adobe Launch)
Adobe’s answer to tag management, deeply integrated into the Adobe Experience Cloud.
- PROs: Deep, native integration with Adobe Analytics, Adobe Target, and AEM. Highly customizable through extensions.
- CONs: Clunky interface compared to GTM. It is absolute overkill without other Adobe products in play.
- Server-Side Tracking: Achieved via Adobe Experience Platform Edge Network (Event Forwarding). It is highly powerful but very complex to implement.
- When to choose: Exclusively when the company’s digital ecosystem is built around the Adobe Experience Cloud.
4. Commanders Act (formerly TagCommander)
A European powerhouse, highly focused on GDPR compliance and consent management.
- PROs: Built with European privacy laws in mind. It seamlessly combines consent management (CMP) with tag management.
- CONs: Smaller global footprint, meaning a smaller community and fewer third-party tutorials.
- Server-Side Tracking: Solid native server-side capabilities designed around strict EU privacy requirements. Implementation is moderate in difficulty.
- When to choose: For a European enterprise where GDPR compliance, data sovereignty, and strict consent management are the absolute top priorities.
5. Jentis (The Innovator)
Jentis is an Austrian tech company that approached tag management differently. Instead of taking a client-side TMS and adding server-side features later, Jentis was built from the ground up as a native Server-Side Tracking platform.
- PROs: Ultimate data quality, built-in legal compliance, and extreme performance benefits since almost no JavaScript is executed in the user’s browser.
- CONs: It requires a complete paradigm shift. It is a premium product with a higher entry cost, and it means unlearning “client-side” habits.
- Server-Side Tracking: It is the core of the product. Implementation is highly streamlined compared to setting up custom Google Cloud servers for GTM.
The “Synthetic Users” Feature by Jentis: This is where Jentis truly innovates. When a user denies cookies in the consent banner, traditional systems stop tracking entirely, leaving huge gaps in the analytics (sometimes up to 30-40% of traffic). Jentis solves this with Synthetic Users. Instead of dropping a cookie in the browser, the Jentis server uses advanced statistical modeling and anonymized server-side data to generate a “synthetic” session identifier. This allows businesses to retain critical data regarding traffic volume, user journeys, and conversions, fully anonymized and 100% compliant with GDPR, without ever relying on client-side cookies.
- When to choose: When data quality is the highest priority, consent banners are swallowing massive amounts of data, and a turnkey, highly advanced server-side solution is what is needed.
Conclusion: How to Choose?
There is no “best” TMS—only the best TMS for a given ecosystem. A Google-centric stack calls for GTM, an Adobe-centric one for Adobe Tags. Enterprise flexibility points to Tealium. Where privacy and restoring lost consent data are the main goals, Jentis is the future.
Questions and answers
Is the difference between Jentis Synthetic Users and Google’s conversion modeling technical, or only a matter of where it runs?
The difference is technical: the two sit on separate steps, and the confusion comes from treating those steps as one.
The first step is identification: whether anything is written to or read from the visitor’s device. The second is estimation: filling gaps in reporting from aggregate patterns. Conversion modeling is purely the second; no identifier is created, and missing conversions are estimated from observed ones. A synthetic session identifier belongs to the first: the server assigns something that stands in for a session, without asking the browser to keep it. Whatever statistical modeling Jentis applies on top of that belongs to the second.
That distinction is also where the questions belong. Accuracy is a property of the estimation step and is answered with a holdout comparison, not with a vendor claim. Whether consent is required under the ePrivacy rules is a property of the storage step, because those rules attach to storing or reading information on the device rather than to inference made on a server afterwards; processing on the server still needs its own legal basis under the GDPR. Any evaluation that keeps those two questions separate will get useful answers; one that asks “is this compliant” about the whole feature will not.
What does “data governance” in enterprise systems mean in concrete terms, compared with a well-run free container?
At its core, three verifiable things, and the difference lies in whether they are enforced or merely agreed.
First, approval: who may publish, and does a second person have to sign off? The free container has roles and versions, but no enforced approval path. The agreement “nobody publishes alone” holds as long as everyone remembers it. Second, access to data: which values may a tag read at all? In the free container, a custom HTML tag can read every variable and every element of the page; such a tag type is blocked through a list in the page code, not in the container interface. Third, destinations: where may data be sent? A maintained list of permitted recipients is something different from the ability to insert a custom script.
Whether that is worth the price depends on a single number: how many people have access. With three people in a team, agreement reliably replaces technology. With thirty people from five agencies, no agreement replaces a technical boundary, and that is exactly where these systems sell.
Can an exported container in JSON format actually be moved to another vendor’s system?
In practice, no. It is better to assume that from the start than to count on a converter. The export describes tags, triggers and variables in one vendor’s format; their counterparts in another system have different concepts and different limits.
What is portable is the data layer, and it decides the cost of a migration. If events and their parameters are described in a document and the page pushes them regardless of who receives them, replacing the system comes down to rebuilding the rules on the same data. If the logic sits in dozens of variables reading the DOM, the migration is a rewrite from scratch.
Hence a practical conclusion for today, whether or not a switch ever happens: the data layer is the part most worth specifying in writing. The rest is tool configuration, and tools change more often than online shops do.
Related articles
