A Creative Approval Process That Ends at Round Two
Revision rounds have risen from three to five since 2007. How to run creative reviews that end sooner, with feedback agencies can actually use.
A Creative Approval Process That Ends at Round Two
It's round six. The client's CEO has just seen the work for the first time and would like to "go back to something closer to the first version". Your designer has quietly stopped using the file names and started adding dates. Nobody can remember which logo placement was agreed.
Everyone in the room is capable. Everyone wants the work to be good. The process is what's failing.
And it's getting worse. According to data from the IPA, reported by Marketing Week, the average number of revisions before approval has risen from three rounds in 2007 to five today.
Both Sides Think the Other Side Is the Problem
The same report, The BetterIdeas Project, found something more revealing. 52% of marketers said they give clear and constructive feedback. Only 30% of agencies said they receive it.
That's a gap of 22 points on the same conversations. Clients believe they're being clear, and agencies don't agree. Neither side is lying. "Can it feel a bit more premium?" sounds clear to the person saying it, because they can picture what they mean.
So the fix isn't asking clients to try harder. It's designing a process where vague feedback has nowhere to go.
Most Revisions Start Before the First Draft
When work goes round and round, the review usually isn't the real problem. The brief is.
If nobody agreed what the work has to do, every reviewer judges it by their own taste, and taste has no finish line. Before any design starts, get three things in writing:
- The objective. What should someone do or feel after seeing this?
- The audience. Who, specifically? "Everyone" can't be reviewed against.
- The mandatories. Logo, legal lines, formats and sizes, plus any claim that will need evidence under rules like the ASA's advertising codes.
Then get the brief signed off by the same person who'll approve the final work. That single step removes a surprising amount of pain, because every later comment can be checked against something both of you agreed.
The Process, Step by Step
1. Name one approver
One person gives the client's final answer. They can gather views from others, but only their consolidated feedback counts. If a CEO will have the last word, you need to know that now, and ideally they should approve the brief too.
2. Agree the rounds before you start
Write it in the scope: two rounds of amends included, and further rounds quoted. This isn't about being difficult. It makes feedback more careful, because each round now has a value.
3. Review internally first
Nothing goes to the client until someone who didn't make it has checked it against the brief. That includes the boring checks: spelling, correct logo, correct format. A client who finds a typo stops looking at the idea.
4. Present the work, don't just send it
A file in an email invites a gut reaction. Walk through it on a call, or record a two-minute explanation: this is the objective, this is how the work meets it, this is the decision we need from you.
5. Ask for feedback in a usable shape
Send a simple framework with every review:
| Ask | Why it helps |
|---|---|
| Does it meet the objective in the brief? | Moves the discussion from taste to purpose |
| What's factually wrong? | Separates must-fix from preference |
| What's the problem, not the solution? | "This feels cold" helps; "make it orange" often doesn't |
| Which comments are from the approver? | Stops contradictions travelling to the designer |
6. Consolidate before you act
If comments conflict, send them back to the approver to settle. Don't guess, and don't quietly try to satisfy both. Resolving conflicts is the client's job, and this is the moment when having one approver pays off.
7. Number the versions and log the decisions
Use a clear naming convention, such as ClientProject_Poster_v3, and never "final_FINAL_2". Keep a one-line record of what was decided at each round. When someone asks why the logo moved, you can answer in ten seconds.
8. Get sign-off in writing
An email saying "approved" is enough. It protects both of you, and it marks the point at which further changes become new work.
How Long Should Reviews Take?
Agree the turnaround up front, for both sides. Something like this is a reasonable starting point for a small campaign, though your own timings will vary:
- Agency: first concepts in the agreed time, amends within two working days
- Client: consolidated feedback within two working days
- If no response: a reminder, then the timeline moves by the same number of days
That last line matters most. It turns a late review into the client's own decision about the deadline, rather than your problem to absorb.
When Rounds Are Still Running Away
Even with all this, some projects go long. It's worth working out why before you just add more rounds.
New reviewer at round four. Someone senior appeared late. Pause and go back to the brief with them, because more amends won't solve a disagreement about the objective.
Comments that reverse earlier ones. Show the decision log, politely. Often the client doesn't remember agreeing it.
Endless small tweaks. This sometimes means nobody wants to be the person who says yes. Ask directly: "What would need to change for you to approve this today?"
Late feedback from the wrong person is expensive, and it's a commercial problem as well as a creative one. In Setup's 2023 Marketing Relationship Survey, dissatisfaction with value and delivery were clients' top reasons for leaving. It's worth tracking how many rounds each client takes, because a client who always needs seven rounds on a two-round scope is telling you something about your pricing. We looked at the broader question of how to charge for AI work, and the same logic applies here: if you can't see the cost, you can't price it.
The Honest Trade-off
Tight processes can squeeze out good surprises. Sometimes the best idea shows up in round four, when someone sees the work on a wall and has a thought they couldn't have had earlier. If you need to explore rather than refine, a structured format like the five-day design sprint may suit you better. If you're doing big-idea brand work, leave room for that, and don't treat a strict round count as sacred.
Clients can also bristle at rules. So present this as what it is: a way to get them better work, faster, with less of their time. Most marketing teams are as tired of round six as you are.
Make It Your Standard
Once this is written down, it stops being a difficult conversation you have with each client and becomes simply how you work.
Build your creative review SOP, free
Round six is rarely a creative problem. It's a decision that nobody made at the start. Agree the objective and the approver before the work begins, and most of the rounds never need to happen.
