Blog
8 min read
A revision round is billable the moment it stops matching what the proposal scoped. That is the test. Not how long it took. Not how patient the client has been, and not how many favours the account lead feels are owed.
Most agencies reach that question far too late. The work is already done, the round was the fifth, and someone now has to write an awkward email about it. Deciding how to charge for extra revision rounds is two jobs, not one, and they happen at different moments. You define what one round covers when you write the proposal. You price the additional ones from what delivering a round costs you.
A revision is a change inside the creative direction the client already approved. A new brief is a change to that direction. Size does not decide it.
Mediabistro's guidance for agencies draws the line between a revision and a new direction with two paired examples. Rewording a headline is a revision. Switching the campaign's target audience is not, because the concept underneath the headline has moved.
Easy to state. Hard to hold at half past four on a Friday, which is why the distinction belongs in writing rather than in someone's judgement. A proposal that promises "reasonable revisions" has defined nothing at all, because reasonable means whatever the client believes it means on the day they are disappointed.
One test survives contact with a real project. Ask whether the request could have been made before the concept was approved. If it could, and nobody raised it, the client has changed their mind rather than refined the work. That is a chargeable event, and saying so in the first week costs you a fraction of the goodwill it costs to say so on the day you send the final invoice.
A round is a unit you invented. It means what your proposal says it means, and nothing more.
Five things turn it from a word into a boundary.
| What to define | The vague version | The version you can bill against |
|---|---|---|
| Who gives feedback | "the client" | one named approver, consolidating comments |
| How feedback arrives | email, calls, corridor chats | a single written list per round |
| What sits inside a round | "amends" | changes within the approved concept |
| Turnaround on feedback | "promptly" | five days from delivery |
| What closes a round | nothing | written sign-off, or the next list arriving |
The last row earns its place. Without a defined close, round one never actually ends, and a trickle of comments arriving across three weeks gets treated by everyone in the room as a single round nobody thought to price. That is where the fee goes.
This is the same discipline as designing profitable projects for a creative agency. The margin is set by what the proposal defines. Not by what the team absorbs later and never mentions.
Start from what a round costs you to deliver. Then add margin. A flat figure lifted from a freelancer's blog post tells you nothing about your studio, your rates or your approval chain.
The arithmetic runs on effective hourly rate. Runn defines effective hourly rate on a fixed-price project role by role. Take the share of the fee a role earns and divide it by the hours that role books. Set the answer against the rate you would have charged for the same time under time and materials. Fixed fees hide that comparison until the job closes.
Five steps get you a number you can defend.
Step three is the one agencies skip. Design time is visible and gets logged. Account time is neither. Nobody books the afternoon spent decoding a contradictory feedback list, and it still gets spent.
Step one deserves a warning too. If your recorded time is patchy, the first pass at this will be rough. That is fine. Build the number from three jobs you remember clearly, use it on the next proposal, then correct it once you have cleaner data. An imperfect figure you can explain still beats a confident guess.
Whatever it costs you to deliver, plus your target margin. The point is that you can show the working, which is what turns an awkward negotiation into a rate card.
Take a brand identity project priced at $12,000 with three rounds scoped. Your team books 80 hours across the whole job. Divide one by the other and the effective rate is $150 an hour before anyone asks for anything more.
Now the fourth round lands. It takes nine hours of design, two hours of account handling and one hour of project management. Twelve hours at an internal cost of $65 an hour is $780. Apply a 45% margin and that round prices at roughly $1,420.
Guessing would have produced a lower figure. A guess anchors on what feels polite to ask for, not on what the round consumed in design, account handling and project management.
Run the same sum for a second and third job before you settle on a rate card. Two numbers close together give you a price. Two numbers wildly apart tell you the rounds are not comparable. Worth knowing in itself. The usual culprit is an approval chain with three people on it instead of one.
The same logic works on a site rather than in a studio. Watching labour burn and unpriced variations before a fixed-price job closes is the same discipline: price the change from what it actually cost, not from what feels fair to ask.
Into unlogged amends, mostly. Retainers hide additional rounds far better than project fees do, because the hours disappear into a monthly allocation that nobody reconciles against the scope schedule.
TrinityP3, which advises marketers on agency remuneration, works through a retainer where the hours run well past what the fee covers. An account director contracted at 0.6 of a full-time equivalent books around 1,498 hours a year against an expected 1,248. The consultancy puts that at roughly 0.72 FTE against the 0.6 the retainer covers.
A retainer clause promising "design support" with no round allowance has sold unlimited amends at a fixed monthly fee.
The fix is a monthly allowance rather than a per-project count. Say six rounds a month across all live work, with a named rate for anything beyond that. Then report the running total each month, in the same note as everything else. A client who has watched the allowance drop from six rounds to one over the course of three weeks does not argue about the extra charge when it lands, because they watched it coming.
Scoro's rundown of the five agency pricing models makes the matching point about fixed fees. Without clearly defined boundaries, clients keep requesting deliverables and the fee never moves to match.
Before the work starts, in writing, with the fee named. Never afterwards.
The conversation is short when the proposal has done its job. You are not reopening a negotiation. You are pointing at a clause the client already agreed to and applying it the way you said you would.
Three sentences cover almost every case:
The third sentence does more work than the first two. A client who is offered a choice has not been told off, and the fee moves without the relationship taking any damage.
What sinks these conversations is timing, not tone. Raise the charge before the work starts and you sound organised. Raise it afterwards and you sound like you are recovering a loss, whatever words you dress it in, because the client can no longer decline work that has already been done. Same fee, same clause. Completely different reception.
The proposal is the one document a client reads properly. So the round definition, the round count and the price of a further round all belong there, on the same page as the deliverables.
Korrel builds proposals with deliverables and risk flags, then tracks time and cost against those estimates as the project runs. That comparison is the useful part. When the fourth round arrives, you already know what rounds one to three consumed. The price comes from your own records, not from instinct.
Start with the last job that overran. Pull the recorded time by round, work out what round three cost you, and put that number in your next proposal as the price of a fourth. One job's history is enough to stop guessing.
COMMON QUESTIONS:
RELATED READING:
Turn implementation unknowns into testable scope exclusions, then route each failed assumption into discovery or a priced change request.
Design the boundary of each stage before work starts. Define its output, included review, dependencies, and trigger for reopening the fee.
Price discovery as a defined engagement. Cost each output, name the unknowns it will test, and charge only for the risk that remains.
Korrel turns briefs into structured proposals, then tracks what each job actually costs.
START FREE TRIAL