Design revisions that finish: running a round so it closes

A revision is a round, and a round is one collected list, one named approver, one closing decision; projects die when the rounds stop being rounds.

Ask around any studio and you will hear the same ghost story: the project that never shipped. Not because the designer ran out of skill, but because the rounds stopped being rounds. Feedback arrives in a slow drip, one message at a time, from five people who never spoke to each other. The designer adjusts the heading, and a new voice prefers the old one. The logo grows a shadow, loses the shadow, regains a smaller shadow. Month three looks exactly like month one with everyone wearier, and the death was not dramatic; it was administrative.

The honest mechanism is that a revision is a unit of work, not a mood. A round has a shape: all feedback is collected as one list, the conflicts inside that list are resolved before a single note is sent, the named approver makes the calls, the designer executes the whole list, and the round closes with a decision. A design process is a small number of such rounds, and everyone can see how many remain. What kills projects is not the third round; it is the round that never closes and quietly becomes a way of working.

Feedback arriving as a slow drip from five people is not a revision process at all; it is a second briefing spread across weeks, delivered one sentence at a time, in a tone the chat app flattens. The project was scoped against the first briefing; nobody scoped it against the second one, because nobody can see it happening. This is also where the money argument hides: unpriced rounds are not free, they are paid in the one resource nobody invoices for and everybody runs out of, attention.

Then there is the voice that says make it pop. That sentence has not made a decision; it has described a feeling and left the decision to the designer's guess. The guess will be wrong, because the feeling was never specified, and the next round will be spent discovering that. Decisions have authors, dates, and reasons; feelings have none of these. A round survives on decisions, so the habit of translating feelings into decisions before the list is sent is the highest-value hour in the whole process.

The practice has a sorting rule that does most of the work: separate wrong from different. Wrong contradicts the agreed brief: the wrong name, the missing language, the color that violates the guidelines, the mockup that shows a service the company does not offer. Wrong gets fixed, no debate, the brief is the contract. Different is a new preference: the heading would look nicer larger, maybe blue would be friendlier. Different is allowed, but it is a decision to reopen the brief, made by the approver, paid in a round, never smuggled in as an error.

Count the rounds openly and agree on them at the start: two included, a third if needed, and everyone knows where the meter is, in money where contracts allow and in attention everywhere else. Then close every round with a decision, even a small one: approved, approved with the noted changes, or one more round with a named reason. A closed round is a rung; the project can stand on it. An unclosed round is a door left open that everyone keeps walking back through.

The honest limit sits before the rounds begin: good revisions require the brief to have been real. A project revising forever is very often a project that skipped alignment, and the rounds are the discovered cost of a first meeting that ended in politeness instead of agreement. Where we sit is the process: we run design in counted, closable rounds and help clients tell wrong from different before the list is sent. Guidance, not guarantees: rounds do not make a design good; they make a design finish, and finishing is where good gets its chance.