Self-hosted vs SaaS CMS

Self-hosted and SaaS describe two ways of running a CMS. A self-hosted CMS runs on infrastructure you control, whether your own servers or a cloud account in your name, and your team is responsible for operating it. A SaaS CMS, software as a service, is delivered by a vendor as a managed online service: you use it over the internet and the vendor runs the machinery behind it. The choice is less about the software’s features than about who does the work of keeping it running.

Who patches, who scales, who owns uptime

In the self-hosted model those responsibilities sit with you. When a security patch is released, your team applies it. When traffic grows, your team provisions capacity. When something breaks at two in the morning, it is your team that owns getting it back. The gain is control: you decide the hosting, the upgrade timing, the configuration, and no third party can change the service underneath you.

In the SaaS model the vendor carries those duties. Patching, scaling, and uptime are theirs to manage, and updates arrive without your involvement. The gain is that you are not operating infrastructure; the cost is that you accept the vendor’s choices about when things change and how the service runs, with less ability to alter them.

Where the data lives

The models also differ in where content is stored. Self-hosting keeps the data in infrastructure you control, which lets you decide its location, relevant when rules such as the GDPR constrain where personal data may be held. With SaaS the data sits in the vendor’s systems, so those questions become a matter of the vendor’s terms and regions rather than your own configuration.

The honest capacity question

The deciding factor is often not preference but capacity. Self-hosting removes licence dependence and gives full control, and it only works if you have the people and the discipline to operate infrastructure reliably; without them, control becomes a liability. SaaS removes that operational burden and, in return, hands the vendor authority over the service and the terms. Open-source software can be self-hosted with no licence fee, and it still needs a competent team to run it securely, keep it maintained, and avoid the lock-in that comes from neglecting either. Which model fits depends on how much you need to control against how much you are equipped to operate. The forms of dependence each creates are covered under vendor lock-in, and the way the choice shifts costs between lines is part of CMS TCO. This choice is separate from where the presentation layer lives, which is the subject of headless vs traditional CMS. The commerce version of the same choice is an engine like Medusa.js, self-hosted with the order data in infrastructure you control — how WAYF builds on it.