Radovan Slíž · WorkSys.Space · 08/09/2026

We want it on-prem, so we stay in control

Your own server sounds like control. Then I ask who patches it and when a backup restore was last tested. Control is not where the machine stands, but whether someone demonstrably looks after it.

We want it on-prem, so we stay in control. I have been hearing this sentence in meetings for years and I understand it. Your own server, your own data, no outsider involved. It sounds like control.

Only then, in practice, I ask further. Who patches that server? Who checks the backups, and when was a restore last tested? Who handles certificates, access rights after an employee leaves, the hole in a library that security sites write about on a Friday evening? And very often it turns out that control actually means: a server somewhere in a server room, looked after by a person who has thirty other things on their plate.

What control really is

Control is not where the machine physically stands. Control is that someone demonstrably looks after the system, every day, even when nothing is happening. And that is work which, with on-prem, does not end with the purchase. It just becomes an invisible item nobody budgeted for. A regular payment for a service can be planned. A failure of your own server always arrives outside the budget.

Let me show the difference on a situation that happens now and then to everyone who runs software. On a Friday evening, news breaks of a serious vulnerability in a library half the internet uses. With a service, this happens: it has people whose job is exactly this, the fix gets deployed, and the customer learns about the whole thing at most from a change record. With a server in a server room, the news waits until Monday for the person with thirty other duties, and whether they get to it this week is decided by whatever lands on their desk in the meantime. Neither is bad faith. It is the difference between work that is someone's main job and work that is an extra burden.

The price SaaS has

To be fair, SaaS has its price too, and I will say it out loud, since I run the WorkSys.Space platform myself: you pay regularly, not once. You have to trust the supplier, and you should demand that trust be substantiated, with certifications, a data agreement, an answer to the question of what happens to your data when you leave. If a supplier cannot answer that in one sentence, it is a bad sign regardless of where the system runs.

So in my view the question is not on-prem or cloud. It is: who will look after it, and is that their job, or an extra burden? The answer to that decides the fate of the system more than the place where the server stands.

Related on this site:What looking after a running system means with us