Your CMS catches fire every week? Our three-layer model hasn't seen a fire in years.

We operate over 100 production CMS systems for mid-market organisations. Looking back today, it wasn't always this calm. We too had phases when updates were risky, when a Friday-evening deployment became a nerve test, and when we had to talk too often with clients about outages instead of about content.

That's different today, and not because of more luck but because of a clear architecture. We've moved our operation onto a three-layer model that is built the same way for every client. No client project is a technical island, no installation is “somehow different”. That decision was hard work and it has paid off. Since then our CMS doesn't catch fire every week, it almost never does.

Why classic setups are so fragile

The standard route in the industry is well-meant and still ends up in chaos regularly. Every client gets their own installation, their own modules, their own server configuration, their own upgrade paths. After three years, 50 projects are 50 different machines. Security updates become a risky drive, because no-one really knows any more what was loosely patched in where. Major upgrades become quarter-long projects.

You probably know the result: weekly fires. Tickets nobody can prioritise any more. A development team that spends 80 percent of its time tending old wounds instead of creating new value.

The three layers at a glance

Our model separates three layers cleanly, each with its own responsibility, its own release cadence and its own contracts to the other layers.

Layer 1: Platform

The lowest layer is our platform. That's where hosting, network, databases, observability, secret management and the CI/CD pipelines live. This layer is identical for every client, run centrally and hardened centrally. Updates at this level reach all clients at once, because they all use the same foundation.

Layer 2: CMS standard

On top sits our CMS standard, today on TYPO3. Core extensions, quality gates, Content-Security headers, backend performance optimisations, monitoring hooks: all of that is anchored in this layer. It too is the same for all clients and is moved forward by us in standardised upgrade windows. In practice that means: when a new TYPO3 LTS appears, we plan the rollout cleanly once and apply it under control to every client.

Layer 3: Client individuality

Only in the top layer lives what makes you unique: your templates, your custom extensions, your content, your interfaces, your workflows. This layer can and should be individual, but it sits on a standardised foundation. That keeps it maintainable while doing justice to your brand.

Why this means quiet for you

The three layers lead to operating properties you'll notice in everyday life.

First: security updates are no longer special events. If a patch is needed in the platform or CMS layer, we roll it out for all clients at once. What is a month-long project elsewhere is a documented routine here.

Second: major upgrades become predictable. We carry them out on weekdays, because we're not flying blind into an individual installation each time but repeating an approach for a platform layer.

Third: you get genuine innovations faster. Our “Level 5 Agent-Native” package, for example, reaches the field quickly precisely because the clients sit on the same foundation. Without the model, rollouts like that simply wouldn't be economically viable.

What this means for editorial work

The second effect our clients name again and again is the calm in their editorial day. Editors who had to work for years in unstable installations almost always say the same sentence to us after moving onto our model: “It's gone quiet.” Fewer tickets, fewer “why isn't this working?” moments, less anxiety before every update.

That calm isn't a side effect, it is the actual promise. You aren't buying the most technically complex solution, you're buying the most stable everyday operation.