Turning Things Off Is Part of Building

We understate the value of turning things off.

Technology organisations are naturally good at celebrating launches. A new service, feature or platform gives us something visible to point at. People can see what changed.

Decommissioning is different.

If it goes well, an old system disappears and most people notice almost nothing.

But the organisation may now have less cyber exposure, fewer failure modes, less support burden, simpler architecture and less technical debt.

It may also have more freedom.

Legacy technology does not only cost money to run. It constrains future decisions.

An old integration has to keep working. Specialist knowledge has to be retained. Data has to remain accessible. Security weaknesses have to be managed. The next change has to accommodate another dependency.

Removing a system can therefore make things possible long after the decommissioning work itself has finished.

The awkward part is that much of this value is invisible.

A launch has a moment.

Decommissioning often creates value through things that no longer happen: incidents avoided, contracts no longer required, obsolete components no longer patched, work no longer repeated and future changes no longer constrained.

Those benefits are harder to photograph.

That can affect what organisations prioritise.

The visible thing gets attention while removal is treated as housekeeping, even when removing something may create more long-term value than adding another feature.

This becomes even more interesting if software continues to get cheaper to create.

If AI and better tooling mean we can build more things, we will also need to become better at deciding which things should stop existing.

Otherwise software abundance becomes software accumulation.

Decommissioning is not the opposite of delivery.

It is part of the lifecycle of delivery.

Sometimes the valuable thing a technology organisation ships is a new capability.

Sometimes it is the absence of an old one.