Useful, Not Done

A useful first version gives people something to react to. They can try it on real work, show us what we misunderstood, and decide whether it deserves more attention. Waiting for a complete answer can postpone exactly the encounter that would improve it.

“Done” is still a useful word. A commitment needs an end, a handoff needs to be usable, and people should not inherit an experiment disguised as a finished service. We want to shorten the distance to useful work without leaving the unfinished part for somebody else to discover.

Be clear about the invitation

A rough tool can be ready for its maker to try on familiar material. More work may be needed before another team depends on it. The person using it should understand what it does, where they need to check it, and whom to turn to when it goes wrong.

Those arrangements should fit the use. A personal experiment does not need the ceremony of a production service. A service does not become ready because its demo went well.

Improvement includes knowing when to stop

Repeated use can teach us what to improve. It can also tell us that the tool is useful enough, that an existing product would serve people better, or that the problem was not worth solving this way.

Small improvements do not automatically multiply into a transformed company. They compete for attention, interact with one another, and create upkeep. We keep the changes that help and reconsider the ones that make more work elsewhere.

The aim is something people can use and continue. Sometimes that means another iteration. Sometimes it means a clear handoff and leaving it alone.

Original illustration generated with AI for this belief.