Status quo bias in legacy systems

With older systems, there are two opposing temptations.

The first is:

We’ve always done it this way. It works, doesn’t it?

The second:

This is all outdated. We need to rewrite the whole thing.

Both can be biased.

The status quo receives an unjustified vote of confidence simply because it already exists. At the same time, a new technology may be overrated because it feels more modern, elegant, or interesting to developers.

An old system often contains knowledge that has never been documented anywhere. A new system does not automatically eliminate the organisational and domain problems of the old one. Conversely, “it still runs” is not a sufficient argument for carrying every piece of technical debt indefinitely.

The better question is not:

Old or new?

It is:

Which specific part creates which risk, and what change would actually improve that relationship?

That often leads neither to preserving everything nor to a grand rewrite, but to gradual modernisation.