The real cost of your legacy system isn't the support contract. It's everything you can't build on top of it.

The maintenance cost is visible. It's on an invoice.

What's not on an invoice:

The integration you couldn't build because your system has no API. The automation you can't run because the data is locked in a proprietary format. The reporting dashboard that doesn't exist because extracting the data takes three days of manual work. The remote access that isn't possible because the system only runs on one machine in one office.

These aren't hypothetical. They're real constraints that shape decisions every week.

One of the recurring conversations in Austrian banking and manufacturing IT - not 'should we migrate?' but: 'What have we stopped even trying to do because we assumed the system couldn't support it?'

The answer was usually more than anyone had realised.

The legacy system becomes the ceiling. Not because anyone decided it would. Just because, over time, nobody could see past it.

A rough way to think about the real cost:

What would you build tomorrow if this constraint wasn't there? What would your team automate? What data would you actually use?

That gap - between what you're doing and what you'd do - that's the number nobody puts on the invoice.