The instinct is to offer more rounds than the competition, because it feels generous and removes an objection. It does remove the objection. It also reliably produces worse work.
Unlimited revisions do not reassure a client into deciding faster. They remove the reason to decide at all.
The short answer
Two. More rounds feel generous and reliably produce worse work, because unlimited revision removes the reason to decide at all rather than reassuring anyone into deciding faster. Define what a round is — one consolidated set of comments, from one decision-maker, within a stated window — and say plainly what falls outside it: a changed brief, a new room, a selection reopened after approval. Then write the clause into the fee proposal, where it can be read before it is needed.
What a round actually is
Define it before you count it, or the number is meaningless.
A round is one consolidated set of feedback and the work done in response. Consolidated is the operative word. Three notes on Monday, two more on Wednesday and a change of mind on Friday is not one round — it is three, because the file gets reopened three times and things that were settled drift each time.
Say this explicitly. Most clients trickle feedback not to be difficult but because that is how noticing works. Told that batching is how they get the best result, most will batch.
Two rounds, and why
For renders and presentations, two consolidated rounds is the sensible standard.
The first round catches the substantive things — a wrong finish, a camera that does not show what matters, lighting that reads colder than intended. The second catches refinement. A third round, in our experience, almost never contains anything that could not have been said in the second. It contains second thoughts, which are a different thing and worth charging for.
If you routinely need more than two, the problem is upstream. Either the brief was thin, or the client is discovering their own taste in public, and neither is fixed by more rounds.
Two is the render answer — the others differ
“Two rounds” is a sensible default for images. It is the wrong number for some of the other things you deliver, because the failure mode is different in each.
| Deliverable | Rounds | Why |
|---|---|---|
| Concept presentation | 1 | You are asking for a direction, not polishing one. A second pass usually means the concept was presented before it was ready. |
| Renders | 2 | One substantive, one refinement. A third almost always contains second thoughts rather than corrections. |
| Drawing packages | 2, plus corrections | Coordination errors are corrections and must never be counted, or people stop reporting them. |
| FF&E schedule | Continuous, with a freeze date | It is a living document until procurement starts. Count from the freeze, not from issue. |
Notice that the schedule is the odd one out. Rounds are the wrong instrument for a document that is meant to change daily — what it needs is a date after which changes are priced, not a count of edits.
What counts, and what does not
Distinguish three categories and price them differently.
Corrections — something does not match the brief. Always free, always immediate, never counted. This matters: a client who fears using a round to report your error will stop reporting errors.
Refinements — adjustments within the agreed direction. This is what the rounds are for.
Changes — a new direction, a different scheme, an added room. Not a revision at all. New work, quoted separately.
Most disputes about revisions are actually disputes about which category something falls into. Naming the categories up front converts an argument into a classification.
The window, and the second decision-maker
Two things quietly break a round count, and both belong in the same clause.
A feedback window. Without one, a round has no end: comments arrive in week two, then week five, and the file is reopened each time with everything around it already moved on. Five working days is generous and workable for images; ten for a drawing package. Say what happens when it lapses — the round is treated as closed, and later comments are quoted — because that is the sentence that stops the drift.
One decision-maker. Most trickle feedback is not indecision, it is two people commenting independently. Ask on day one who consolidates and whose view is final; a project with an invisible second approver is where late reversals come from. Where a couple or a committee is genuinely deciding together, that is fine — but they nominate one person to send the set, and the set arrives once.
Both belong in the proposal rather than in a conversation, because you will not want to have this conversation in month three, when it will sound like a complaint rather than a process.
Writing the clause
Short, in the proposal, not buried in terms:
"Includes two rounds of revisions. A round is one consolidated set of feedback. Corrections to anything not matching the approved brief are not counted. Changes of direction are quoted separately."
Four sentences that prevent most of the conversations you would otherwise have.
What to do when they run out
Do not silently absorb it, and do not make it a confrontation. Say what it is, offer the price, and let them choose:
"That's a change of direction rather than a refinement, so it sits outside the two rounds. It's about half a day — I can quote it properly if you'd like to proceed."
Clients accept this far more easily than designers expect. What they react badly to is discovering a limit they did not know existed — which is a documentation failure, not a pricing one.
The counterintuitive part
Constrained rounds produce better work. When feedback is finite, it gets considered before it is sent. The client reviews properly, consolidates, and tells you the things that matter instead of thinking aloud.
Unlimited rounds produce the opposite: reactive notes, contradictory instructions, and a final image that has been sanded down by twelve small compromises rather than shaped by two good decisions.
The limit is not a restriction you impose on the client. It is the mechanism that makes their feedback worth acting on.
